Objectives: Chronic aspiration of gastric fluid contents can decrease long-term survival of pulmonary transplants due to development of obliterative bronchiolitis. However, little is known about the early immune response and the cascade of events involved in the development of obliterative bronchiolitis.
Materials and Methods: We utilized a rat orthotopic pulmonary transplant model and a single aspiration of either gastric fluid or normal saline to investigate the histologic, cellular, and cytokine changes associated with an acute gastric fluid aspiration event compared with normal saline at 2 and 10 days after aspiration.
Results: Our observations included a decrease in pulmonary compliance and increased airway inflammation and acute rejection of the transplanted lung, as well as increases in macrophages, granulocytes, and proinflammatory cytokines such as interleukin 1β, transforming growth factor β1 and β2, and tumor necrosis factor α in bronchoalveolar lavage fluid from the transplanted lung of gastric fluid-aspirated rats compared with normal saline-aspirated rats.
Conclusions: The acute inflammatory response observed in the present study is consistent with changes found in chronic models of aspiration-associated injury and suggests a potentially important role for mast cells in the development of obliterative bronchiolitis.
Key words : Gastroesophageal reflux, Obliterative bronchiolitis, Pulmonary allograft dysfunction
Lung transplantation is the criterion standard for patients with end stages of lung diseases, including emphysema, cystic fibrosis, and pulmonary fibrosis. However, long-term survival of pulmonary allografts remains poor compared with other solid-organ transplants. Obliterative bronchiolitis (OB), which is a major cause of chronic lung allograft dysfunction (CLAD), is the leading cause of death in lung transplant recipients after the first posttransplant year. Obliterative bronchiolitis is characterized histologically by submucosal fibroproliferation of the small airways, leading to luminal compromise and respiratory failure.1 Mounting evidence suggests a potential role of gastroesophageal reflux in the development of OB after transplant.2,3 In one study of lung transplant patients with gastroesophageal reflux disease (GERD; n = 43), it was concluded that OB could be decreased and lung function could be improved if patients underwent Nissen fundoplication surgery before the late stages of OB syndrome.4
To further investigate the relationships between GERD, OB, and CLAD, we utilized a rat model of orthotopic lung transplantation using an external cuff technique5 and experimentally induced chronic gastric fluid aspiration.6,7 This model has been used to recapitulate the clinical observations associated with the development of OB in pulmonary grafts after 8 weeks of weekly aspiration. A mildly histoincompatible donor and recipient strain combination has been used, with immunosuppressive agents administered to recipients to avoid severe acute rejection.6-8 Of note, this model has been utilized to explore components of gastric fluid aspirate, which may induce the development of OB. For example, both soluble and insoluble components of gastric fluid induced the development of OB.9 Also, low pH gastric fluid (pH 2.5) and neutralized gastric fluid (pH 7.4) have similar numbers of OB lesions after 8 weeks of aspiration.10
The exact mechanism(s) by which gastroesophageal reflux disease contributes to the development of OB is not yet completely understood. Chronic aspiration after pulmonary transplant is associated with increases in proinflammatory cytokines, which include interleukin (IL)-1α, IL-1β, IL-6, tumor necrosis factor (TNF)-α, granulocyte-macrophage colony-stimulating factor (GM-CSF), and transforming growth factor (TGF)-β, in both bronchoalveolar lavage fluid and serum.8 In addition, mast cells may also play a role in the development of OB and are associated with OB in human allografts.11 In a rat orthotopic lung transplant model, cromolyn administration decreased perioperative mortality, graft failure, and the number of OB lesions.12 These data suggest that early donor allograft inflammation likely contributes to the development of OB-associated CLAD.
The acute inflammatory response to gastroesophageal reflux after pulmonary transplant has not yet been well described. Diminished mucociliary clearance and other posttransplant complications could potentially alter both the physical characteristics of an aspiration event (length of time before clearance) and the subsequent immune response.13,14 Although the development of OB in patients is thought to take at least 2 months with a mean of 11 months,15 examination of the acute inflammatory and immune response to early aspiration events may provide insight into the processes that initiate development of OB-associated CLAD. In this study, we utilized an established rat orthotopic single-lung transplant model to investigate the histologic, cellular, and cytokine changes associated with a single aspiration event that would occur within 7 days of the lung transplant.
Materials and Methods
Male Wistar Kyoto (RT-1I) and Fisher 344 (F344; RT-1lv1) rats were purchased from Harlan Laboratories (Indianapolis, IN, USA). All rats were housed in specific pathogen-free conditions in the animal care facilities at Duke University Medical Center (Durham, North Carolina, USA) in accordance with institutional guidelines. All animal care and use procedures were approved by the Duke Institutional Animal Care and Use Committee. The rats weighed 210 g to 300 g at the time of transplant.
Gastric fluid collection
Gastric fluid was collected from F344 rats as previously described.16 The pooled gastric fluid had a pH of 2.5. The pooled gastric fluid was aliquoted and stored at -80°C until needed.
Transplant and aspiration procedures
Left lungs from Wistar Kyoto rats were orthotopically transplanted into F344 rats using a modification of the previously reported nonsuture external cuff technique6; mean (SD) cold ischemic time was 46.6 ± 5.2 minutes, and mean (SD) warm ischemic time was 13.7 ± 2.8 minutes. For postoperative analgesia, transplanted rats received 5 mg/kg subcutaneous ketoprofen (Fort Dodge Animal Health LLC, Fort Dodge, IA, USA) once daily for 3 days and 0.05 mg/kg subcutaneous buprenorphine (Reckitt Benckiser Pharmaceuticals, Richmond, VA, USA) twice daily for 2 days. For antimicrobial prophylaxis, transplanted rats received 5 mg/kg subcutaneous Baytril (Bayer Healthcare, Whippany, NJ, USA) once daily for 3 days. For immunosuppression, transplanted rats received 5 mg/kg subcutaneous cyclosporine A (Novartis Pharm Stein AG, Stein, Switzerland) once daily every other day. The first dose of cyclosporine was administered at the time of graft reperfusion.
Transplanted rats received a one-time aspiration of 0.5 mL/kg of fluid (0.9% normal saline or gastric fluid) into the left (transplanted) lung 1 week posttransplant, as described previously.6
Study design and grouping
A total of 32 pulmonary allografts were performed. Transplanted rats were randomly assigned into 2 groups: rats receiving aspiration with normal saline (control group, n = 16) and rats receiving aspiration with gastric fluid (n = 16). The study duration (2 days vs 10 days) was also randomly assigned before the aspiration event. Half of the rats (n = 8) in each group were killed 2 days after the aspiration event for tissue collection, whereas the remainder (n = 8) of each group were killed 10 days after the aspiration event for tissue collection.
Procurement of heart-lung block and measurement of compliance
After death, the hearts and lungs of transplanted rats were removed en bloc as previously described.10 Compliance data were collected as previously described.9 In brief, pulmonary compliance of the allograft was measured using a custom-made pressure volume device, from which the volume obtained after every 50 mm H2O rise in pressure was recorded up to 350 mm H2O and then at 380 mm H2O.
Collection of bronchoalveolar lavage fluid
The right and left main bronchi were sequentially lavaged with phosphate-buffered saline as previously described.9 One milliliter of bronchoalveolar lavage (BAL) fluid was centrifuged, and the supernatant was stored at -80°C for cytokine analysis. The remaining BAL fluid was centrifuged at 480g for 5 minutes at 4°C, and the cell pellet was used for flow cytometry analysis.
Lung tissue was fixed using 10% neutral buffered formalin and stained with hematoxylin and eosin as well as Masson trichrome stain for collagen. Sectioning and staining were performed by the Substrate Services Core and Research Support Services at Duke University. All fields from 2 sections of the left lung (one from the upper lung portion and another from the lower lung portion) were analyzed by coauthors Chang and Sanders for airway inflammation and acute rejection. The extent of airway inflammation and the degree of acute rejection were graded using previously described scoring systems.1 All histologic sections were graded in a blinded fashion. Based on consensus, an airway inflammation grade and acute rejection grade were reported for each section.
Multiplex suspension cytokine array
Bronchoalveolar lavage fluid was analyzed for multiple analytes using Procarta rat cytokine immunoassay kits (eBioscience, Inc., San Diego, CA, USA) except for TGF-β1, TBF-β2, and TBF-β3, which were analyzed with Fluorokine multiplex kits (R&D Systems, Minneapolis, MN, USA). Samples were acid activated for TGF measurements, and all samples were run in duplicate and as directed by the manufacturer. Assays were analyzed using a BioPlex 100 and BioPlex Manager software 4.1 (Bio-Rad Laboratories, Hercules, CA, USA).
Cells were processed as described in detail previously.17 All antibodies were obtained from BD Biosciences (Franklin Lakes, NJ, USA) unless otherwise noted: PE anti-CD3 (G4.18), APC anti-CD3 (1F4), APC anti-CD4 and PE-Cy5 anti-CD4 (OX-35), PerCP anti-CD8a and Biotin anti-CD8a (OX-8), PE-Alexa Fluor 647 anti-CD11 (OX-42; AbD Serotec, Raleigh, NC, USA), PE anti-CD25 (OX-39), PE anti-CD28 (JJ319), Alexa Fluor 488 anti-CD45RA (B cell only marker) (OX-33, AbD Serotec), fluorescein isothiocyanate (FITC) anti-CD59 (TH9), PE anti-CD62L (HRL1), PE anti-CD68 (ED1, AbD Serotec), PE anti-CD81 (Eat2), PE anti-CD86 (24F), Alexa Fluor 488 anti-CD90 (OX-7, AbD Serotec), FITC anti-CD134 (OX-40), Alexa Fluor 647 anti-CD161a (10/78, AbD Serotec), Alexa Fluor 647 anti-CD163 (ED2, AbD Serotec), FITC anti-CD172a (ED9, AbD Serotec), Alexa Fluor 488 anti-FoxP3 (FJK-16s, eBioscience, Inc.), FITC anti-granulocyte (HIS48), PE anti-CD200R (OX-102, AbD Serotec), and PerCP anti-MHCII (RT1B, I-A) (OX-6).
Streptavidin without biotin-labeled primary antibodies and properly labeled isotype antibodies were used as controls, and “fluorescence minus one controls”18 were used for FoxP3 and CD68 gating.
Cells were analyzed within 24 hours in the Duke Human Vaccine Institute Flow Cytometry Core Facility using an LSR II cytometer (BD Biosciences) and FlowJo software (FlowJo LLC, Ashland, OR, USA).
Continuous data are presented as mean and SEM, and categorical data are presented as counts. All comparisons of quantitative data across 3 or more groups utilized the 2-way analysis of variance, with Bonferroni posttest where required. The D’Agostino-Pearson omnibus K2 test was used to determine normality. Comparisons between 2 groups utilized the t test. Histologic grading data were analyzed using Cochran-Mantel-Haenszel statistics. To address the hypothesis that gastric fluid aspiration worsens airway inflammation and contributes to more severe forms of acute rejection compared with saline aspiration, one-tailed P values for Cochran-Mantel-Haenszel statistics were reported. The analysis of variance and t tests were performed using GraphPad Prism 5.01 software (GraphPad Software, La Jolla, CA, USA). Cochran-Mantel-Haenszel statistics were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). All P values reported are 2-sided unless otherwise stated, and a P < .05 was considered significant.
Perioperative complications and death
A total of 32 rat orthotopic left lung transplants were performed. Neither intraoperative nor perioperative morbidity or mortality occurred in this study. However, what appeared to be severe aspiration-induced pneumonia was observed in 2 animals euthanized 2 days after aspiration with gastric fluid. Because of the extent of airway necrosis and consolidation (n = 1) and airway necrosis and parenchymal necrosis (n = 1), histopathology scores for these 2 animals could not be obtained.
Lung compliance of pulmonary allografts
The compliance of the pulmonary allografts was measured ex vivo either 2 or 10 days after aspiration of assigned fluid. At applied pressures greater than 50 mm H2O, the average compliance of lungs from the gastric fluid aspiration group was significantly lower than that of the control (saline) group (Figure 1). Similar findings were observed between the groups at 2 and 10 days after aspiration. However, the average compliance of lungs in either group was greater 10 days after aspiration than 2 days after aspiration, although these observations were not statistically significant.
The severity of airway inflammation was graded using a discrete scale from 0 to 2, where 0 represents normal lung and 2 represents granulomatous pneumonitis with prominent giant cells. Scattered areas of disruption of the epithelial barrier and/or cell shedding were observed in all allografts graded as 2. As shown in Figure 2, pulmonary allografts from the gastric fluid aspiration group exhibited increased airway inflammation compared with the normal saline group. At 10 days, there was significantly increased airway inflammatory changes in the gastric fluid aspiration group compared with the normal saline group (P = .03). In addition, airway inflammation was significantly increased at 10 days compared with at 2 days after gastric fluid aspiration (P = .04).
The severity of acute rejection was graded using a continuous scale from 0, representing normal pulmonary parenchyma, to 4, representing diffuse perivascular, interstitial, and air-space infiltrates of mononuclear cells with prominent alveolar pneumocyte damage and endothelialitis. As shown in Figure 3A, pulmonary allografts from the gastric fluid aspiration group exhibited higher acute rejection grades than the normal saline group at both 2 days (P = .01) and 10 days (P = .01) after aspiration. Unlike the pattern seen for airway inflammation grade, no significant difference was observed between 2 days and 10 days in rats who underwent gastric fluid aspiration. Representative histologic sections from each group, stained with hematoxylin and eosin, are shown in Figure 3, B-E.
Cytokine analysis of bronchoalveolar lavage fluid
The results of the cytokine analysis of BAL samples collected from left lung allografts are shown in Table 1. Compared with the normal saline group, inflammatory cytokines IL-1α, IL-1β, IL-2, IL-13, TGF-β1, TGF-β2, and TNF-α were significantly higher in BAL samples from rats in the gastric fluid aspiration group at 2 days or 10 days after aspiration. Macrophage chemoattractants CCL2, CCL3, CCL5, and CCL7 were also significantly higher in the gastric fluid aspiration group than in the normal saline group at 2 days or 10 days after aspiration. Within the gastric fluid aspiration group, we observed that cytokines IL-1α, IL-1β, IL-2, IL-13, TGF-β1, and TNF-α were all significantly reduced at 10 days after aspiration compared with at 2 days after aspiration.
Cellular analysis of bronchoalveolar lavage fluid
Analysis of cells from BAL samples collected from left lung allografts are shown in Tables 2 and 3. Compared with aspiration with normal saline, a significant increase in the percentage of regulatory T cells (FoxP3+CD4+ T cells), CD86+ T cells, CD134+ T cells, natural killer cells, granulocytes, and activated alveolar macrophages were observed 2 days after gastric fluid aspiration. However, only the percentage of CD134+ T cells and granulocytes remained significantly increased 10 days after gastric fluid aspiration compared with normal saline aspiration.
The exact mechanism(s) by which gastroesophageal reflux disease contributes to the development of OB is not yet completely understood. Animal models used to assess OB-associated CLAD have proven useful in circumventing the difficulties of studying patients during the perioperative period. These animal models include the mouse and rat heterotopic tracheal transplant model,19-21 the rat orthotopic single-lung transplant model,6,22 and large animal models.23,24 However, most studies investigating OB-associated CLAD have explored later stages of the disease process.
In this study, significant airway inflammation in the transplanted lung was seen 2 days after a single aspiration with gastric fluid compared with aspiration with normal saline. The airway inflammation persisted at 10 days after aspiration. Interestingly, airway inflammation was evident at 2 days regardless of the type of fluid aspirated. This inflammation could be due to the transplant procedure, performed 7 days before aspiration. However, another possible explanation is that the aspiration event alone produced some degree of inflammation in the allograft.
Evaluation of BAL samples revealed a significant increase in cytokines and chemokines involved in macrophage activation and migration in rats treated with gastric aspirate compared with normal saline. These included CCL3, CCL7, and GM-CSF. All were elevated 2 days after gastric fluid aspiration compared with aspiration with normal saline. Alveolar macrophages with activated phenotypes increased more than 5-fold 2 days after gastric fluid aspiration, and total macrophage numbers increased as well, compared with normal saline aspiration. Two cytokines that affect endothelial cell permeability and proliferation, sVCAM1 and vascular endothelial growth factor, increased more than 2-fold at day 2 and remained elevated at day 10 in gastric fluid-aspirated samples compared with normal saline-aspirated samples. Cytokines involved in the activation and tissue extravasation of T and B cells such as IL-2, IL-12, and sRANKL were similarly elevated. Another cytokine that was elevated after gastric fluid aspiration compared with normal saline aspiration was nerve growth factor (NGF)-β, a proinflammatory mediator of innate immunity that is secreted by mast cells and that promotes differentiation and maturation of mast cells, suggesting a potential early and self-perpetuating inflammatory role for this cell type.25 Cytokines with anti-inflammatory effects such as IL-10 were not significantly increased either at 2 or 10 days after gastric fluid aspiration compared with normal saline aspiration. A significant increase in regulatory T cells was evident at 2 days after gastric fluid aspiration compared with normal saline aspiration. Increased numbers of regulatory T cells have been associated with GERD in esophageal tissues.26 However, it is unknown whether a transient increase in BAL-associated regulatory T cells might play a role in the chronic aspiration OB model.
Interestingly, granulocyte migration to the lung increased more than 50-fold 2 days after gastric fluid aspiration compared with aspiration with normal saline in BAL samples. Additionally, granulocytes accounted for nearly 80% of the total cell count at day 2 in BAL samples that received gastric fluid aspiration compared with aspiration with normal saline. Although granulocyte numbers decreased at 10 days, they were still 8-fold higher in samples from gastric fluid aspirated rats. Increased migration and subsequent degranulation of granulocytes were affected by CCL2, CCL3, CCL5, eotaxin, GM-CSF, IL-1β, and IL-13, all of which were elevated after gastric fluid aspiration compared with aspiration with normal saline. Granulocyte migration and activation could be important in the mechanisms leading to the development of OB. Granulocytes can migrate to sites of inflammation within minutes and contribute to inflammatory processes in the lung in several ways. An acute increase in granulocyte number has also been observed after a single gastric fluid aspiration in a nontransplanted rat lung.27 However, 7 days after aspiration, granulocyte numbers had returned to normal values in the nontransplanted lungs, whereas a significant increase in granulocyte number was still evident at 10 days in the transplanted lungs in this study.
Other inflammatory cytokines that were elevated in this study have been well documented in the literature for their suspected involvement in acute and/or chronic pulmonary diseases. For example, IL-1 is thought to be involved in chronic obstructive pulmonary disease and asthma, and overexpression of IL-1β in mouse lung epithelium causes distal air space enlargement, mucus metaplasia, and airway fibrosis.28 In our study, IL-1β was elevated more than 20-fold in BAL samples 2 days after gastric fluid aspiration and remained 15-fold higher at 10 days after gastric fluid aspiration compared with aspiration with normal saline. Another cytokine implicated in pulmonary diseases such as asthma, chronic bronchitis, chronic obstructive pulmonary disease, acute lung injury, and acute respiratory distress is TNF-α. This cytokine is often detected early in the course of inflammation29 and can be released quickly on degranulation of mast cells. Here, we detected a 2.5-fold increase in TNF-α 2 days after aspiration with gastric fluid compared with aspiration with normal saline. However, elevated levels of TNF-α were not detected at day 10 between the 2 aspiration groups. A third cytokine, TGF-β, has an active role in regulating inflammatory processes when produced in low levels.30 However, when expression of TGF-β is upregulated, it is thought to be linked to the development of fibrosis in chronic inflammation.31 Both TGF-β1 and TGF-β2 were elevated 2 days after aspiration with gastric fluid (9-fold and 4-fold, respectively) and 10 days after aspiration with gastric fluid (10-fold and 3-fold, respectively) compared with aspiration with normal saline.
The acute inflammatory response observed in the present study is consistent with changes found in chronic models of aspiration-associated injury and indicate the involvement of both Th1 and Th2 cytokines in the development of OB as well as other pulmonary diseases.32,33
Volume : 17
Issue : 1
Pages : 84 - 92
DOI : 10.6002/ect.2017.0152
From the the 1Department of Surgery and the 2Department of Pathology, Duke
University Medical Center, Durham, North Carolina, USA; the 3Division of
Thoracic and Cardiovascular Surgery, Hualien Tzu Chi Hospital, Hualien, Taiwan;
the 4Department of Surgery, Tzu Chi University, Hualien, Taiwan; and the
5Department of Immunology, Duke University Medical Center, Durham, North
Acknowledgements: This study was performed at Duke University Medical Center, Durham, NC 27710, USA. This work was funded in part by the Parks Protocol Memorial Fund and the Fannie E. Rippel Foundation for general scientific research support. The authors declare no financial or other potential conflicts of interest. The authors thank Julie Fuller (Substrate Services Core & Research Support Services in the Department of Surgery at Duke) for expert assistance with histology.
Corresponding author: Andrew S. Barbas, Duke University Medical Center, DUMC Box 3512, Durham, NC 27710, USA
Phone: +1 248 202 1687
Figure 1. Lung Compliance of Pulmonary Allografts After a Single Aspiration
Figure 2. Histologic Analysis of Airway Inflammation in Pulmonary Allografts
Figure 3. Histologic Analysis of Acute Rejection in Pulmonary Allografts
Table 1. Cytokine and Chemokine Levels in Bronchoalveolar Lavage Fluid
Table 2. Phenotypic Analysis of Cells in Bronchoalveolar Lavage Fluid (By Percentage)
Table 3. Phenotypic Analysis of Cells in Bronchoalveolar Lavage Fluid (By Cell Count)