Dietary Docosahexaenoic Acid Prevents Silica-Induced Development of Pulmonary Ectopic Germinal Centers and Glomerulonephritis in the Lupus-Prone NZBWF1 Mouse

Abstract

Ectopic lymphoid structures (ELS) consist of B-cell and T-cell aggregates that are initiated de novo in inflamed tissues outside of secondary lymphoid organs. When organized within follicular dendritic cell (FDC) networks, ELS contain functional germinal centers that can yield autoantibody-secreting plasma cells and promote autoimmune disease. Intranasal instillation of lupus-prone mice with crystalline silica (cSiO₂), a respirable particle linked to human lupus, triggers ELS formation in the lung, systemic autoantibodies, and early onset of glomerulonephritis. Here we tested the hypothesis that consumption of docosahexaenoic acid (DHA), an ω-3 polyunsaturated fatty acid with anti-inflammatory properties, influences the temporal profile of cSiO₂-induced pulmonary ectopic germinal center formation and development of glomerulonephritis. Female NZBWF1 mice (6-wk old) were fed purified isocaloric diets supplemented with 0, 4, or 10 g/kg DHA - calorically equivalent to 0, 2, or 5 g DHA per day consumption by humans, respectively. Beginning at age 8 wk, mice were intranasally instilled with 1 mg cSiO₂, or saline vehicle alone, once per wk, for 4 wk. Cohorts were sacrificed 1, 5, 9, or 13 wk post-instillation (PI) of the last cSiO₂ dose, and lung and kidney lesions were investigated by histopathology. Tissue fatty acid analyses confirmed uniform dose-dependent DHA incorporation across all cohorts. As early as 1 wk PI, inflammation comprising of B (CD45R⁺) and T (CD3⁺) cell accumulation was observed in lungs of cSiO₂-treated mice compared to vehicle controls; these responses intensified over time. Marked follicular dendritic cell (FDC; CD21⁺/CD35⁺) networking appeared at 9 and 13 wk PI. IgG⁺ plasma cells suggestive of mature germinal centers were evident at 13 wk. DHA supplementation dramatically suppressed cSiO₂-triggered B-cell, T-cell, FDC, and IgG⁺ plasma cell appearance in the lungs as well as anti-dsDNA IgG in bronchial lavage fluid and plasma over the course of the experiment. cSiO₂ induced glomerulonephritis with concomitant B-cell accumulation in the renal cortex at 13 wk PI but this response was abrogated by DHA feeding. Taken together, realistic dietary DHA supplementation prevented initiation and/or progression of ectopic lymphoid neogenesis, germinal center development, systemic autoantibody elevation, and resultant glomerulonephritis in this unique preclinical model of environment-triggered lupus.

Keywords: autoimmunity; ectopic lymphoid structure; lung; omega-3; polyunsaturated fatty acid; silica; systemic lupus erythematosus.

Figure 1

Experimental design. Feeding of groups of female NZBWF1 mice with CON (0 g/kg DHA), Low DHA (4 g/kg DHA) or High DHA (10 g/kg DHA) diets were initiated at age 6 wk. At age 8 wk, mice were dosed intranasally with 25 μl PBS VEH or 25 μl PBS containing 1 mg cSiO₂ weekly for 4 wk. Cohorts (n = 7–8 per group) of animals were terminated at 12, 16, 20, and 24 wk of age which corresponded to 1, 5, 9, and 13 wk post-instillation (PI) of final cSiO₂ dose.

Figure 2

DHA supplementation suppresses cSiO₂-induced elevation of inflammatory cells in BALF. Total cell counts in BALF samples from NZBWF1 mice were conducted using a hemocytometer and then monocytes/macrophages, lymphocytes, neutrophils, and eosinophils quantified after differential staining using morphological criteria. Abbreviations: C, control diet; L, Low DHA diet; H, High DHA diet. Data are means ± SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). Complete statistical analyses can be found in Supplementary Table 5.

Figure 3

DHA consumption prevents cSiO₂-triggered inflammation in lungs of NZBWF1 mice. Representative light photomicrographs depict H&E-stained lung sections from CON/cSiO₂ groups at (A) 5 wk post-instillation (PI), (C) 9 wk PI, and (E) 13 wk PI; (B) CON/VEH at 13 wk PI; (D) Low DHA/cSiO₂ at 13 wk PI; and (F) High DHA/cSiO₂ at 13 wk PI. Black arrows in photomicrographs indicate lymphoid cell infiltration in the peribronchiolar and perivasculature interstitium. Dietary DHA substantially suppressed lymphocytic infiltration (D, F; See also Table 4). Abbreviations: a, alveolar parenchyma; b, bronchiolar airway; v, blood vessel.

Figure 4

DHA consumption delays cSiO₂-triggered B-cell appearance in lungs of NZBWF1 mice. Pulmonary densities of immunohistochemically labeled B-cells (CD45R⁺) in hematoxylin counterstained lung sections were immunohistochemically and morphometrically determined as described in Materials and Methods. Representative photomicrographs depict CD45R⁺ cells (reddish brown chromogen, indicated by arrows) in (A) CON/cSiO₂ at 5 wk post-instillation (PI), (C) 9 wk PI, and (E) 13 wk PI; (B) CON/VEH at 13 wk PI; (D) Low DHA/cSiO₂ at 13 wk PI; and (F) High DHA/cSiO₂ at 13 wk PI. Abbreviations: a, alveolar parenchyma; b, bronchiolar airway; v, blood vessel. (G) Volume densities of CD45R⁺ cells are graphically expressed as means + SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). Complete statistical analyses can be found in Supplementary Table 5.

Figure 5

DHA consumption delays cSiO₂-triggered T-cell appearance in lungs of NZBWF1 mice. Pulmonary densities of immunohistochemically labeled T-cells (CD3⁺), in hematoxylin counterstained lung sections were morphometrically determined as described in Materials and Methods. Representative photomicrographs depict CD3⁺ cells (reddish brown chromogen, indicated by arrows) in (A) CON/cSiO₂ at 5 wk post-instillation (PI), (C) 9 wk PI, and (E) 13 wk PI; (B) CON/VEH at 13 wk PI; (D) Low DHA/cSiO₂ at 13 wk PI; and (F) High DHA/cSiO₂ at 13 wk PI. Abbreviations: a, alveolar parenchyma; b, bronchiolar airway; v, blood vessel. (G) Volume densities of CD3⁺ cells are graphically expressed as means + SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). Complete statistical analyses can be found in Supplementary Table 5.

Figure 6

DHA consumption delays cSiO₂-triggered follicular dendritic cell networking in lungs of NZBWF1 mice. Pulmonary densities of immunohistochemically labeled follicular dendritic cell (CD21/35⁺) networking in hematoxylin counterstained lung sections were morphometrically determined as described in Materials and Methods. Representative photomicrographs depict CD21/35⁺cells (reddish brown chromogen, indicated by arrows) in (A) CON/cSiO₂ at 5 wk post-instillation (PI), (C) 9 wk PI, and (E) 13 wk PI; (B) CON/VEH at 13 wk PI; (D) Low DHA/cSiO₂ at 13 wk PI; and (F) High DHA/cSiO₂ at 13 wk PI. Abbreviations: a, alveolar parenchyma; v, blood vessel. (G) Volume densities of CD21/35⁺ cells are graphically expressed as means + SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). Complete statistical analyses can be found in Supplementary Table 5.

Figure 7

Dietary DHA suppresses cSiO₂-triggered plasma cell influx in lungs and elevation of anti-dsDNA IgG in bronchoalveolar lavage fluid of NZBWF1 mice at 13 wk post-instillation. (A–D) Light photomicrographs of lung sections, immunohistochemically stained for IgG (dark reddish brown chromagen), from mice in the 13-wk post-exposure groups – (A) CON/VEH, (B) CON/cSiO₂, (C) Low DHA/cSiO₂ and (D) High DHA/cSiO₂ mice. A marked infiltration of IgG⁺ plasma cells (arrows) was present in interstitial tissue surrounding blood vessels (v) and bronchiolar airways (BA) in the lungs of CON/cSiO₂ mice (B). Extracellular IgG was also conspicuous in alveolar airspaces (a) of these mice. Densities of IgG⁺ plasma cells and alveolar extracellular IgG were strikingly less in the lungs of both Low DHA/cSiO₂ and High DHA/cSiO₂ mice (C and D, respectively). Only a few widely scattered IgG⁺ plasma cells were occasionally present in the peribronchiolar and perivascular interstitial tissue of CON/VEH mice (A). (E) Morphometrically determined volume densities of IgG⁺ plasma cells are graphically expressed as means + SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). (F) Dietary DHA suppresses cSiO₂ -induced elevation of anti-dsDNA IgG in bronchiolar lavage fluid (BALF). BALF was pooled for each experimental group and analyzed in duplicate. Data represent means ± SEM. ^* indicates significantly different from CON/VEH group (p < 0.05).

Figure 8

Dietary DHA suppresses cSiO₂ -induced elevation anti-dsDNA IgG in plasma of NZBWF1 mice. Anti-dsDNA IgG in plasma of NZBWF1 mice was quantitated by ELISA. Data are means ± SEM (n = 7–8 per group). Symbols: ^* indicates significantly different from CON/VEH group (p < 0.05); # indicates significantly different from CON/cSiO₂ group (p < 0.05). Complete statistical analyses can be found in Supplementary Table 5.

Figure 9

cSiO₂ induced tubular proteinosis in kidneys of NZBWF1 mice at 13 wk PI. Light photomicrographs of kidney sections from NZBWF1 mice at 13 wk PI that were histochemically stained with Periodic acid-Schiff and hematoxylin (PASH) and morphometrically analyzed for tubular proteinosis. Significant lesions were detectable in 0/8, 5/8, 1/8, and 0/8 mice in the CON/VEH, CON/cSiO₂, Low DHA/cSiO₂, and High DHA/cSiO₂ groups, respectively. Representative photomicrographs of the renal cortex in (A) CON/VEH; (B) CON/cSiO₂; (C) Low DHA/cSiO₂; and (D) CON/VEH at 13 wk PI; and (D) High DHA/cSiO₂. Abbreviations: g, glomerulus; rt, renal tubules; ^*, tubular protein; arrows, sclerotic glomerulus. (E) Morphometric assessment of tubular proteinosis. Data are means ± SEM (n = 8 per group). Symbol: ^* indicates significant difference (p < 0.05) between CON/VEH and CON/cSiO_2. Complete statistical analyses can be found in Supplementary Table 6.

Figure 10

Lymphocytic and B-cell accumulation in kidneys of NZBWF1 mice 13 wk following cSiO₂ exposure are prevented by dietary supplementation with DHA. Renal densities of lymphocyte and B-cell (CD45R⁺) were determined by routine light microscopic examination (H&E stained sections) and immunohistochemistry, respectively, as described in Materials and Methods. Representative light photomicrographs of H&E stained kidney sections depict (A) CON/VEH, (B) CON/cSiO₂, (C) Low DHA/cSiO₂; and (D) High DHA/cSiO₂ at 13 wk PI. Light photomicrographs of CD45R⁺ cells (reddish brown chromogen, indicated by arrows) in lung sections from (E) CON/VEH, (F) CON/cSiO₂, (G) Low DHA/cSiO₂; and (H) High DHA/cSiO₂ at 13 wk PI; and (F) High DHA/cSiO₂ at 13 wk PI. Abbreviations: rc, renal cortex; rm, renal medulla; ia, intralobular artery; iv, intralobular vein. Asterisk indicates focal aggregates in the interstitium. (I) Volume densities of CD45R⁺ cells are graphically expressed as means ± SEM (n = 8 per group). No significant differences were identified for comparisons to CON/VEH or CON/cSiO₂. Complete statistical analyses can be found in Supplementary Table 6.

Figure 11

Temporal profile of cSiO₂-triggered ectopic germinal center formation in the lung. Diagrammatic representation of lymphoid cell infiltration and development of ectopic lymphoid structures with germinal centers in the interstitial tissue surrounding a bronchiolar airway and adjacent blood vessel (A) prior to cSiO₂ exposure, (B) at 1 and 5 wk PI, and (C) at 9 and 13 wk PI. Small accumulations of lymphoid cells are present in the perivascular and airway interstitial tissue at 1 wk PI with increasing numbers of lymphocytes (mainly B-cells) in the interstitial tissue at 5 wk PI. Distinct ectopic lymphoid structures characterized by follicular dendritic cell-laden germinal centers and numerous plasma cells were present by 9 and 13 wk after the final cSiO₂ exposure.