Donor B-cell alloantibody deposition and germinal center formation are required for the development of murine chronic GVHD and bronchiolitis obliterans

Abstract

Chronic GVHD (cGVHD) poses a significant risk for HSCT patients. Preclinical development of new therapeutic modalities has been hindered by models with pathologic findings that may not simulate the development of human cGVHD. Previously, we have demonstrated that cGVHD induced by allogeneic HSCT after a conditioning regimen of cyclophosphamide and total-body radiation results in pulmonary dysfunction and airway obliteration, which leads to bronchiolitis obliterans (BO), which is pathognomonic for cGVHD of the lung. We now report cGVHD manifestations in a wide spectrum of target organs, including those with mucosal surfaces. Fibrosis was demonstrated in the lung and liver and was associated with CD4(+) T cells and B220(+) B-cell infiltration and alloantibody deposition. Donor bone marrow obtained from mice incapable of secreting IgG alloantibody resulted in less BO and cGVHD. Robust germinal center reactions were present at the time of cGVHD disease initiation. Blockade of germinal center formation with a lymphotoxin-receptor-immunoglobulin fusion protein suppressed cGVHD and BO. We conclude that cGVHD is caused in part by alloantibody secretion, which is associated with fibrosis and cGVHD manifestations including BO, and that treatment with a lymphotoxin-β receptor-immunoglobulin fusion protein could be beneficial for cGVHD prevention and therapy.

Figure 1

Allogeneic transfer of BM and low concentrations of splenic cells in hosts treated with Cy/TBI-caused cGVHD and BO. B10.BR mice treated with 120 mg · kg⁻¹ · d⁻¹ Cy (day −3, −2) and lethally irradiated (day −1) were transplanted with BM alone or BM plus 0.75-1 × 10⁶ splenocytes from C57Bl/6 mice. n = 10 mice/group. Weights of the animals (A) and survival (B) were tracked up to 60 days after transplantation. PFTs were performed on anesthetized animals at day 60 (C) and day 28 (D) after transplantation. Animals were artificially ventilated, and resistance, elastance, and compliance were measured as parameters of distress in lung function in animals receiving low-dose splenocytes or T cells. Representative data from 3 individual experiments. **P < .01.

Figure 2

Multiorgan disease in mice with cGVHD; immune infiltration and collagen deposition. (A) Tissues (lung, liver, tongue, salivary glands [SG], thymus, spleen, ileum, and colon) were harvested at day 60 after transplantation and stained with H&E to determine pathology. Bright-field images were captured at 100× magnification with an Olympus BX51 microscope. (B) Inflammation, immune infiltration, and parenchymal changes were scored with a cumulative scoring system used previously. (C) Collagen deposition was determined with a Masson trichrome staining kit; blue indicates collagen deposition. (D) Collagen deposition was quantified as a ratio of blue area to total area of tissue. Quantification was performed with the analysis tool in Photoshop CS3. Data from 2 individual experiments were pooled to obtain pathology scores; n = 12. *P < .05; **P < .01; ***P < .005.

Figure 3

CD4⁺ T-cell and B220⁺ B-cell infiltration is seen in lungs and livers of transplanted animals. Liver and lung tissues were harvested at day 60 after transplantation from animals receiving BM only and BM plus splenocytes (BM + spleen) and analyzed by immunohistochemistry and methyl blue counterstaining. Representative images from 3 individual experiments are shown for CD4 (A) and B220 (D). Images were captured with a bright-field microscope at 200× magnification. For lung, infiltration was quantified by obtaining a ratio of CD4⁺ (B) or B220⁺ (E) cells to total cells in a 100-mm² field of view under the microscope. Shown is an average of the count from 4 representative fields. For liver, CD4 (C) and B220 (F) cell infiltration was quantified by counting the number of antibody binding–positive cells and obtaining an average of counts from 4 representative fields. n = 6. *P < .05; **P < .01; ***P < .005.

Figure 4

Antibody deposition detected in target areas of lung and liver in diseased animals. Six-micrometer sections of frozen lung and liver tissues harvested at day 60 after transplantation were analyzed by immunofluorescence. Tissues were incubated with FITC-conjugated anti-mouse Ig. (A) Representative images for lung and liver from 3 individual experiments. n = 8. (B) Ig deposition was quantified on a 0-3 scale to determine the amount of antibody in the tissues. (C) Serum from animals given BM only and animals given BM plus splenocytes was collected at day 60 after transplantation and incubated with healthy B10.BR lung and liver tissue followed by FITC-conjugated anti-mouse Ig to detect the presence of tissue-specific antibodies in the serum of diseased animals. White arrows depict areas of Ig deposition. Fluorescence was detected with an Olympus FluoView 500 confocal laser scanning microscope at a magnification of 200×.

Figure 5

Animals receiving B cell–deficient BM show a decrease in pathology. B10.BR recipients were transplanted with WT BM or BM from μMT knockout mice with or without WT T cells. (A) At 60 days after transplantation, mice were anesthetized and artificially ventilated to measure pulmonary function parameters. (B) Collagen deposition was quantified from trichrome-stained samples as a ratio of blue area to total area of tissue. Quantification was performed with the analysis tool in Photoshop CS3. (C) Lung and liver tissues were harvested, and 6-μm frozen sections were stained with FITC-conjugated anti–mouse Ig for antibody deposition within the tissues. White arrows denote areas of Ig deposition. Images shown are representative images of 3 individual experiments; n = 4. *P < .05; ***P < .005; P = .4066 for lung BM (μMT) only vs BM (μMT) + T cells; P = .2860 for liver BM (μMT) only vs BM (μMT) + T cells.

Figure 6

Secreted antibody is required for pulmonary dysfunction in animals with cGVHD. (A) B10.BR mice were transplanted with BM ± splenocytes from WT or (m+s)IgMxJhD BALB/c mice and anesthetized at day 60 after transplantation for PFTs. Resistance, compliance, and elastance were measured; n = 8. (B) Collagen deposition was quantified from trichrome-stained samples as a ratio of blue area to total area of tissue. Quantification was performed with the analysis tool in Photoshop CS3. (C) Infiltration of CD4⁺ cells in the lung and liver of transplanted mice. (D) Lung and liver tissues were harvested, and 6-μm frozen sections were stained with FITC-conjugated anti-mouse Ig for antibody deposition within the tissues. White arrows denote areas of Ig deposition. Representative image from 2 independent experiments; n = 8. *P < .05; **P < .01; ***P < .005.

Figure 7

Disruption of GC formation by LTβR-Ig treatment reduces lung dysfunction and cGVHD organ tissue fibrosis. B10.BR recipients were transplanted as per Figure 1. A cohort of animals receiving BM + splenocytes were treated with 200 μg of murine LTβR-Ig or the control Ab, murine MOPC21, every 3 days beginning on day 28 after transplantation until day 52. (A) Spleen tissue harvested from these animals at day 60 was analyzed by immunofluorescence for GC structures. GCs were detected by colocalization of IgM (green), VCAM-1 (blue), and peanut agglutinin (red); merged images show overlap (white) to discriminate GC. White arrows highlight GC. Images were captures with an Olympus FluoView 500 confocal laser scanning microscope at 100× magnification; n = 5. (B) The size of the GC was quantified by measuring the area of peanut agglutinin staining in Photoshop C3. (C) Frequency of GCs was quantified by counting the number of GCs in 1 mm² of spleen section. (D) PFTs were performed on anesthetized animals on day 60 after transplantation to measure lung function. (E) Animals treated with LTβR-Ig and MOPC21 were examined for fibrosis in the lung and liver. (F) Presence of deposited lung and liver tissue–specific antibodies in animals treated with LTβR-Ig and MOPC21 was determined by immunofluorescence by staining with FITC-conjugated anti–mouse Ig. White arrows depict Ig deposition. Images were captured at 200× magnification and are representative of 2 individual experiments; n = 5. *P < .05; **P < .01.