Systemic lupus erythematosus (lupus), a chronic autoimmune disease predominantly affecting young women of child-bearing age, is caused by the loss of immunological tolerance resulting from yet poorly understood interactions between an individual’s genome and the environment (1, 2). Initial onset of this disease typically involves unresolved inflammation and incomplete clearance of dead cells, accumulation of self-antigens, and autoantibody production, leading to the formation of immune complexes. Tissue deposition of these complexes fosters cytokine and chemokine production, infiltration of mononuclear effector cells, and ultimately, cell death. Collectively, these pathologic effects promote systemic inflammation and tissue injury, adversely affecting multiple organs including kidney, skin, heart, lung, and brain (3). Frequent recurrent cycles of flaring and remission in lupus leads to permanent organ damage that, without treatment, can manifest as severe glomerulonephritis and end-stage renal disease (kidney failure). Lupus management includes therapies based on immunosuppression, lymphocyte depletion, and cytokine/chemokine neutralization with monoclonal antibodies or receptor antagonists (4). Limitations of existing treatments include heterogeneity of individual patient symptoms, diversity of responses to therapeutics, adverse side effects leading to permanent organ damage, and high expense leading to financial burden on the patient. Thus, there is clear need for alternative safe and low-cost interventions for lupus.

Environment can influence onset and progression of autoimmunity in individuals with a genetic predisposition towards lupus (5). Notably, crystalline silica dust (cSiO₂) exposure in construction, mining, and ceramics industries has been epidemiologically associated with lupus and other autoimmune diseases (6–9). Induction of autoimmunity can be recapitulated at the preclinical level by introduction of cSiO₂ dust particles into the lungs of mice predisposed to lupus (10–14). Our laboratory has found in lupus-prone female NZBWF1 hybrid mice that, following four repeated weekly intranasal instillations with cSiO₂, latency of glomerulonephritis onset is reduced by three months (15, 16). At the mechanistic level, airway instillation of cSiO₂ causes the persistent build-up of particle-bearing monocytes/macrophages, release of proinflammatory cytokines and chemokines, recruitment of neutrophils and macrophages into the alveolar space, programmed and necrotic cell death, and accumulation of nuclear and cytoplasmic debris (17). These events facilitate perivascular accumulation of T-cells, B-cells, and IgG autoantibody-producing plasma cells, leading to formation of ectopic lymphoid tissues (ELT) in the lung. The centrality of the lung in cSiO₂-triggered lupus onset and progression in NZBWF1 mice provides a unique window into potential mechanisms for and interventions against autoimmune pathogenesis in humans exposed to this and other respirable particles in the environment.

Diet is another environmental factor that can influence onset and progression of autoimmunity, with lipids being of particular importance. Consumption of marine ω-3 polyunsaturated fatty acids (PUFAs) can ameliorate chronic inflammatory and autoimmune diseases (18) and extend human lifetime (19). Mechanisms underlying ω-3 attenuating effects include i) influencing membrane function by altering lipid rafts and signaling, ii) moderating gene expression by repressing/activating transcription factors, iii) competition with ω-6 PUFA for metabolic enzyme binding sites and iv) serving as substrates for formation of pro-resolving metabolites [reviewed in (18, 20, 21)]. Unfortunately, Western diets contain large amounts plant- and animal-derived fats, therefore skewing tissue phospholipid fatty acid content overwhelmingly towards the more proinflammatory ω-6 PUFAs and away from the anti-inflammatory ω-3 PUFAs (22, 23). However, this imbalance can be overcome by eating fish or dietary fish/microalgal oil supplements containing the ω-3 PUFAs docosahexaenoic acid (C22:6 ω-3; DHA) and eicosapentaenoic acid (C20:5 ω-3; EPA). Relevant to the present study, dietary supplementation with ω-3 PUFA-rich fish oil inhibits inflammatory gene expression, production of autoantibodies, glomerulonephritis, and/or death in numerous murine preclinical lupus models (24–29), with DHA-enriched fish oil being most efficacious (30, 31). Consistent with these mouse studies, several clinical investigations support the assertion that increased ω-3 PUFA consumption might benefit patients with lupus (21, 32–34).

To ascertain how ω-3 PUFAs influence environment-triggered lupus flaring, Bates and coworkers (16) compared cSiO₂-triggered autoimmune disease onset and progression in autoimmune-prone female NZBWF1 mice fed control diets or diets containing DHA at the realistic human caloric equivalent of 2 or 5 g/d. After 2 wk on these dietary regimens, mice were treated weekly for 4 wk with intranasally instilled cSiO₂ (1 mg) and then sacrificed 1, 5, 9, and 13 wk. DHA supplementation dose-dependently reduced cSiO₂-triggered B- and T-cell, follicular dendritic cell, and IgG⁺ plasma cell accumulation in the lungs along with development of glomerulonephritis. Subsequently, tissues from the aforementioned study were analyzed for 800 immune-related genes using targeted multiplex platform (35). Intriguingly, pulmonary transcriptome signatures of cSiO₂-treated mice fed control diet were indicative of progressive increases of genes associated with inflammation, innate/adaptive immunity, chemokines, type 1 interferon (IFN) responses, and antigen processing. Remarkably, DHA supplementation dose-dependently attenuated these responses, suggesting that consumption of this fatty acid impedes gene responses that play a central role in inflammation, ectopic lymphoid tissue neogenesis, and autoimmunity.

It is not yet known how modulation of cSiO₂-induced gene expression by DHA influences the inflammatory proteome. We therefore conducted a targeted proteomic analysis of archived tissue fluids from the Bates et al. study (16) to test the hypothesis that short-term repeated pulmonary exposures of lupus-prone mice to cSiO₂ induces inflammatory proteins in alveolar fluid and plasma, which are suppressed by DHA consumption. The results establish that cSiO₂ elicited a diverse inflammatory proteome in the lung that was consistent with unresolved inflammation and pulmonary ectopic lymphoid neogenesis and, furthermore, dietary DHA supplementation markedly quelled this robust inflammatory protein response.

Since inhaled cSiO₂ particles are cleared slowly from the lung, they persist and drive a cycle in alveolar macrophages involving phagocytosis, phagolysosome permeabilization, and inflammasome activation that culminates in cell death and reentry of cSiO₂ into the alveolar space (17). Repetition of this cycle evokes chronic unresolved inflammation in the lung and release of self-antigens, leading to eventual loss of immune tolerance (38, 39). In prior studies with lupus-prone NZBWF1 female mice, we discovered that DHA feeding prior to cSiO₂ instillation suppresses progressive increases over wks 1, 5, 9, and 13 PI in inflammatory/autoimmune gene expression, inflammatory cell recruitment, and autoantibody production (16, 35, 36). For example, in mice fed a control diet without DHA, cSiO₂ instillation triggered a mixed inflammatory cell infiltrate in the lung (alveolitis composed of monocytes/macrophages, neutrophils and lymphocytes) with semiquantitative histopathology severity scores of minimal (1 out 4) and mild to moderate (2/3 out of 4) at 1/5 and 9/13 wk PI, respectively (16). Likewise, total inflammatory cells in BALF from these cSiO₂-instilled mice were modestly increased at 1 and 5 wk PI and dramatically increased by 9 and 13 wk PI, as compared to saline vehicle-instilled animals fed the control diet. In contrast, cSiO₂ mice fed DHA supplemented diet (human daily equivalent 5 g/day) had minimal alveolitis at 1,5, and 9 wk PI and only mild alveolitis at 13 wk PI. Total inflammatory cell numbers in BALF of these DHA-supplemented mice were similarly suppressed at 9 and 13 wk PI (16).

The findings presented herein show for the first time how cSiO₂ and DHA influence the inflammatory proteome in BALF and plasma in this preclinical model of environment-triggered autoimmunity. First, cSiO₂ induced a diverse inflammatory proteome in the BALF that persisted and progressed long after particle exposure was ceased. Upregulated proteins included chemokines, enzymes, adhesion molecules, co-stimulatory molecules, growth factors, TNF superfamily members, and signal transduction proteins. Second, the plasma inflammatory protein responses to cSiO₂ were limited to chemokines and much less robust than observed in the BALF, confirming the lung to be the central site for inflammation and autoimmunity in this model. Third, DHA consumption markedly suppressed cSiO₂-induced expression of many of these inflammatory proteins in BALF at wks 1, 5, and 9 PI and, furthermore, DHA’s ameliorative actions against some inflammatory proteins appeared to be lost at 13 wk PI. Fourth, correlation analyses confirmed that DHA’s inhibitory effects on cSiO₂-induced inflammatory protein signatures corresponded with prior measurements of transcriptional responses, ectopic lymphoid structure development, and autoantibody production. Finally, DHA’s ameliorative effects were associated with inhibition of proinflammatory cytokine-driven innate immune pathways and activation of lipid-signaling signaling pathways. Altogether, these new findings support the contention that DHA supplementation might be a promising intervention against cSiO₂-induced inflammation and autoimmune disease onset and flaring in susceptible individuals.

IPA analysis indicated that IL-1, IL-6, and TNF-α were among the chief regulators for predicted pathways mediating cSiO₂-triggered inflammatory proteins and that these pathways were downregulated by DHA. cSiO₂ instillation into the lung is well-known to acutely elicit secretion of proinflammatory cytokines such as IL-1β, IL-1α, IL-6, and TNF-α by direct or indirect action on alveolar macrophages, neutrophils, and epithelial cells (40–43). Therefore, it was surprising that cSiO₂’s effects on these proinflammatory cytokines in BALF were modest across the time cohorts of this study. Nevertheless, these low responses concurred with low expression of corresponding cytokine mRNAs in the parent study (35). The lack of a robust proinflammatory cytokine responses to prolonged cSiO₂ exposure compared to that for chemokines and other proteins likely reflects the strict regulation and relatively short half-lives of proinflammatory cytokine mRNAs and proteins in tissues due to their injurious potential (44).

Pathway analysis further revealed that DHA’s quelling effects were associated with attenuation of ARE‐mediated mRNA decay, TREM1, NF-κB, and STAT3 signaling pathways, and potentiation of PPAR signaling and PPARα/RXRα and LXR/RXR activation. In the parent investigation from which tissue fluids were sourced, we reported that total ω-3 PUFA and ω-6 PUFA represented 6.6 and 33.4% of lung lipids in cSiO₂-treated CON-fed mice, respectively, whereas in the cSiO₂-treated DHA-fed group, they were 23.2 and 16.3%, respectively (16). At the mechanistic level, increasing the ω-3:ω-6 PUFA ratio might hamper lipid raft development and prevent activation of transmembrane receptors that control innate and adaptive immune processes (45). Additionally, DHA and EPA can be liberated from the membrane by intracellular and extracellular phospholipases (46, 47), allowing them to act as ligands for extracellular and intracellular receptors that inhibit proinflammatory signaling pathways (48, 49). Consistent with our IPA pathway analysis here, these receptors include PPAR and RXR family members that downregulate NF-κB- and STAT-dependent transcription of innate immune genes (50, 51). Moreover, DHA and EPA can directly compete with ω-6 PUFAs as enzymatic substrates for downstream bioactive lipid mediators, leading to reduced proinflammatory metabolites (e.g. prostaglandins, leukotrienes) and increased specialized pro-resolving metabolites (e.g. resolvins, protectins, maresins) that promote resolution of inflammation (52, 53). Accordingly, changing the tissue balance by enhancing ω-3 PUFAs and diminishing ω-6 PUFAs likely reduces the cSiO₂-induced inflammatory proteome by several independent or interdependent mechanisms.

The diverse array of inflammatory proteins induced by cSiO₂ in this lupus-prone mouse model provides new insights into how the particle triggers inflammation and autoimmunity and how such triggering might be ameliorated by consumption of DHA. Chemokines were foremost among the proteins induced by cSiO₂. The robust increases in BLC (B-lymphocyte chemoattractant; CXCL13) are particularly pertinent in ELT neogenesis. BLC, which is produced by follicular dendritic cells, T follicular helper cells, and Th17 cells, is essential for B-cell localization into lymphoid follicles. Other CXCL chemokines might also impact cSiO₂-induced autoimmunity. MIP-2 (CXCL3) primarily recruits neutrophils to sites of inflammation that, upon degranulation, could secrete proteolytic enzymes that elicit bystander tissue injury and inflammation, or release neutrophil extracellular DNA traps via NETosis, a source of self-DNA and autoantigens (54). Consistent with the increase in CD3⁺ T cells in lung ELT observed in the original study (16), MIG (CXCL9) and ITAC (CXCL11) recruit and position activated T-cells (55). CCL chemokines can also recruit monocytes, dendritic cells, lymphocytes, and eosinophils, which collectively promote cSiO₂-induced inflammation and autoimmunity in pleiotropic manners (56–58).

cSiO₂-triggered release of extracellular enzymes into the alveolar fluid and suppression of these responses by DHA also have significant implications to inflammation and autoimmunity in the lung. cSiO₂ induction of matrix metalloproteinases (MMPs), a family of enzymes that degrade extracellular matrix and basement membrane components, is consistent with prior in vivo and in vitro investigations (59). cSiO₂ also elicited extracellular granzyme B, which has been linked to lupus pathogenesis by cleaving autoantigens and exposing neoepitopes, as well as ACE and the aspartic acid protease renin 1, two key enzymes in the renin-angiotensin system. The latter regulates many physiological functions, including inflammation, and has been previously shown to play a role in silicosis in mice and rats (60–64).

Elevated soluble adhesion molecules in the BALF following cSiO₂ instillation are highly indicative of persistent inflammation, extracellular matrix breakdown, and potentially, as a means of negative feedback inhibition in the lung. Selectins have a key role in inflammation by mediating leukocyte rolling, adhesion, and transmigration into inflamed tissues. Leukocyte (L)-selectin (CD62L) is expressed on neutrophils, monocytes, and most lymphocytes (65) and is a biomarker of inflammation and autoimmune disease, and notably, occupational exposure to cSiO₂ (62–64). Endothelial (E)-selectin (CD62E) and platelet (P)-selectin are biomarkers in autoimmune and rheumatic diseases (66–69). DHA suppression of adhesion molecules in the alveolar fluid observed in this study is remarkably in agreement with the findings in plasma of prior preclinical and clinical investigations of ω-3 PUFAs (70).

Consistent with extracellular matrix breakdown, cSiO₂ induced release of soluble TNF superfamily members into the alveolar fluid, and their induction was suppressed by DHA consumption. The ligand-receptor pairs CD40L-CD40 and OX40L-OX40 represent critical immune checkpoints for B cell differentiation, antigen presenting cell activation, and T cell modulation (71). High serum levels sCD40L or sOX40L have been reported in lupus and other rheumatic autoimmune diseases (72, 73). CD6, a lymphocyte surface receptor that associates with the T cell receptor/CD3 complex, acts as a co-stimulatory molecule for T cell activation (74). sCD6 is detectable in patients with systemic autoimmune and inflammatory disorders (75, 76). Other TNF receptor superfamily proteins that were upregulated by cSiO₂, downregulated by DHA, and have pertinence to lupus and autoimmunity include sBAFFR (77), sTNFR1 and sTNFR2 (43), and sTRAIL (78–80).

Strengths of this investigation include employment of a widely utilized lupus-prone mouse model, a well-recognized autoimmune trigger, and a human-relevant dose of DHA. However, this investigation had several limitations that require additional consideration in future studies. Since this was a discovery-based approach, we employed a single standard dilution for BALF and plasma and used MFI as a measure of protein signal, thus limiting our ability to perform absolute quantitation of inflammatory proteins identified herein. Nevertheless, proteins of further interest can be quantified by ELISA in future studies employing our preclinical model. Additionally, the dose of cSiO₂ employed here to induce autoimmunity in the NZBWF1 mouse represents one half of a human equivalent lifetime exposure to cSiO₂ at the recommended NIOSH exposure limit (15). The inability to adequately clear particles at this high dose likely unleashes a vicious cycle of inflammation, cell death, and re-release of cSiO₂. This propensity for unresolved inflammation appeared to overwhelm DHA’s capacity to suppress induction of some inflammatory proteins by wk 13, suggesting there is a limit to its ameliorative action. Thus, additional insights could be gained by replicating this study using lower cSiO₂ doses and extending experiment duration.

A further limitation of this study is that since we did not include a non-autoimmune strain in our parent investigation, it could be argued that the observed protein signatures may be a generalized inflammatory response to cSiO₂ that happens to coincide with autoimmunity in this lupus-prone mouse strain. However, in an earlier prior study, we compared the effects to cSiO₂ and DHA in female NZBWF1 mice to those for one of the parental lines, female NZW/LacJ, which does not exhibit the propensity towards autoimmunity (81). At 12 wk PI, the NZBWF1 mice exhibited marked macrophage, neutrophil, and lymphocyte infiltration with robust ectopic lymphoid tissue neogenesis in the lung that were suppressed by DHA feeding. In comparison, inflammatory cell responses in the lung and kidney to cSiO₂ were extremely modest in the NZW/LacJ mice but were nonetheless ameliorated by DHA supplementation. Inflammatory and autoimmune responses to cSiO₂ are likewise much lower in female C57BL/6 mice, another mouse strain with low lupus predilection (15). Thus, it could be speculated that cSiO₂-induced inflammatory proteome would be similarly less robust in NZW/LacJ and C57Bl/6 mice, however, this supposition will need future confirmation using these non-autoimmune strains and extending the study length beyond three months after cSiO₂ treatment.

In conclusion, the demonstration here that DHA prevents or delays cSiO₂-triggered inflammatory protein expression in NZBWF1 mice provides new insights pathways by which ω-3 PUFAs interfere with environmental triggering of chronic inflammation and autoimmune disease. These findings have resonance with one of the four goals of the 2020-2030 Strategic Plan for NIH Nutrition Research which is to develop precision nutrition strategies to reduce the burden of disease in clinical settings (82). The DHA diet used here equates to human intake of 5 g/d, which is a realistic and safe human dose (83). Consistent with our findings, ameliorative effects on inflammatory mediators or inflammatory cell function in human investigations have been found at marine ω-3 PUFA intakes of > 2 g/d (21), and a recent meta-analysis of human clinical trials determined that supplementation with 3 g/d could benefit patients with lupus (33). Taken together, ω-3 supplementation could be a precision nutrition approach for delaying progression of autoimmunity and reducing lupus flaring following respiratory exposures to cSiO₂ or potentially other airborne particles.

Original proteomic datasets from this study are available in a publicly accessible repository: https://doi.org/10.26078/W1G0-KB31. Other existing gene and autoantibody expression datasets used in this study are available at https://doi.org/10.26078/4697-1p77 and https://datadryad.org/stash/share/LLav72h-OvhiPYfqma2-qzCSzeZCJjIts55UWwaXqzA, respectively.

The animal study was reviewed and approved by Institutional Animal Care and Use Committee at Michigan State University (AUF #01/15-021-00).

LR: data curation, data analysis/interpretation, figure preparation, manuscript writing. MB: study design, animal study, necropsy, immunohistochemical analyses and morphometry, manuscript preparation and manuscript editing. AB and KW: data analysis/interpretation, figure preparation, manuscript editing. JH: study design, lung/kidney histopathology, morphometry, data analyses. JP: study design, coordination, oversight, funding acquisition, manuscript preparation and submission. All authors contributed to the article and approved the submitted version.

This research was funded by NIH ES027353 (JP), NIH F31ES030593 (KW), NIH T32ES007255 (KW), Lupus Foundation of America (JP, KW), USDA National Institute of Food and Agriculture Hatch Projects 1020129 (JP) and UTA-01407 and UTA-01456 (AB), and the Dr.Robert and Carol Deibel Family Endowment (JP).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.