The Institutional Animal Care and Use Committee (IACUC) at Michigan State University (MSU; AUF# 201800113) approved all animal experimental protocols for this study. SLE-prone NZBWF1 mice (Jackson Laboratories, Bar Harbor, ME) were housed at MSU’s animal facility, which was maintained at a constant temperature (21-24°C), humidity (40-55%), and a 12-hr light/dark cycle.

After the mice were bred, the dams were euthanized between gestational days 14 and 18. Specifically, dams were placed in Optimice cages and euthanized via CO₂ inhalation (4.7 L CO₂/min) for 10 minutes to ensure death to both the dam and neonates. Death of the dam via CO₂ inhalation was confirmed by paw pinch. Cervical dislocation was used as a secondary form of euthanasia for the dam. Fetuses were promptly dissected from the dam by severing the placental arteries. Loss of access to the maternal blood supply served as the secondary form of euthanasia for the neonates.

Excised fetal livers were further processed to generate fetal liver-derived alveolar-like macrophage (FLAM) cell cultures as previously described (31, 37). Briefly, fetal livers were dissociated in sterile phosphate-buffered saline (PBS) to create a single-cell suspension. Suspensions were filtered through a 70-µm filter and centrifuged at 220 x g for 5 minutes. Two wash steps were performed using sterile PBS before resuspending cells in modified RPMI media (mRPMI, Thermo Fisher), which contained 10% fetal bovine serum (FBS, Thermo Fisher), 1% penicillin-streptomycin (P/S, Thermo Fisher), 30 ng/mL murine granulocyte-monocyte colony-stimulating factor (GM-CSF, PeproTech), and 20 ng/mL recombinant human TGF-β1 (PeproTech). Cells were plated in 10 cm culture plates (one liver per plate). mRPMI media was replaced every ~2 days until cells created an adherent monolayer and exhibited a round AM-like morphology (~1 wk). Cells were then frozen in liquid nitrogen until needed for this study.

FLAMs were thawed and cultured in mRPMI media for this experiment, and cells between passages 10–11 were used for subsequent studies. Upon LPS stimulation, NZBWF1 FLAMs exhibited significantly elevated type I IFN gene responses compared with FLAMs derived from C57BL/6 controls ( Supplementary Figure 1 ), supporting their utility as a model for investigating therapeutic interventions targeting hyperinflammatory SLE-associated macrophages.

GraphPad Prism Version 10 (GraphPad Software, La Jolla, California, USA, www.graphpad.com) was used to visualize fatty acid, quantitative RT-PCR, and cytokine data. These data were subjected to the ROUT outlier test (Q=1%) and then the Shapiro-Wilk test (p < 0.01) to identify outliers and assess normality, respectively. Data failing to meet the assumption for normality were analyzed using the non-parametric Kruskal-Wallis test, followed by Dunn’s post-hoc test. Data that met assumptions for normality and equal variances were analyzed by parametric one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test. Data are shown as mean ± standard error of the mean, with p < 0.05 considered statistically significant. The SynergyFinder 3.14.0 R package (45) was used to evaluate inhibitory interactions between DHA and DEX on ISG expression. Synergy scores were calculated using the zero-interaction potency (ZIP) model (46). Visualizations of RNA-seq differential expression and functional enrichment analyses were generated using GraphPad Prism and R.

LPS stimulation in NZBWF1 FLAMs induced robust upregulation of representative ISGs, including Irf7, Mx1, Ifit1, Isg15, Oasl1, and Rsad2. DHA monotherapy concentration-dependently suppressed this IFN response. Significant ISG suppression was observed at 25 µM (p < 0.05 vs. LPS control) ( Figure 1A ), while lower concentrations (5 to 10 µM) showed a significant decrease ( Figures 1B, C ). DEX monotherapy at 1 µM effectively suppressed ISG expression, whereas lower doses (≤100 nM) demonstrated incomplete and variable transcriptional inhibition ( Figures 1A–C ). Combining subinhibitory DHA (5 to 10 µM) with low-dose DEX (1 to 100 nM) enhanced the suppression of LPS-driven ISGs (p < 0.05 vs. individual treatments), exceeding additive effects ( Figures 1B, C ). This combination amplified DEX’s potency by 10- to 100-fold, achieving near-complete transcriptional silencing. This potentiation was observed across multiple ISGs, indicating broad modulation of type I IFN-regulated genes and signaling pathways.

Using SynergyFinder 3.14.0 and the ZIP synergy model, we quantified the synergy between DHA and DEX related to the inhibition of ISG expression. At all concentrations, DHA and DEX demonstrated significant synergistic interactions in inhibiting Irf7 ( Figure 2A ), Oasl1 ( Figure 2B ), Rsad2 ( Figure 2C ), Ifit1 ( Supplementary Figure 1A ), Isg15 ( Supplementary Figure 1B ), and Mx1 ( Supplementary Figure 1C ). Synergy was most robust when FLAMs were pretreated with 1 nM or 10 nM DEX in conjunction with 5 µM DHA ( Figure 2 and Supplementary Figure 1 ). At higher concentrations of DHA (i.e., 10 µM and 25 µM) and DEX (i.e., 100 nM and 1000 nM), ZIP synergy scores were still greater than 0, indicating a smaller magnitude of synergy, as the pathways were more robustly inhibited by monotherapies at these higher concentrations.

Consistent with gene expression, LPS exposure stimulated secretion of IFN-β and selected ISG products (CCL2 and CXCL10) after 24 hr compared to VEH-treated control cells ( Supplementary Figure 2 ). These responses were significantly attenuated by DHA and DEX monotherapies. When DHA and DEX were administered in combination, the secretion of IFN-β, CCL2, and CXCL10 was further reduced. Accordingly, combining DHA and DEX enhanced suppression of ISG protein expression, further underscoring their potential as a combined therapeutic strategy for modulating TLR4-driven pathogenic gene responses.

Macrophages orchestrate immune responses through gene expression finely tuned by a complex interplay of TFs to balance inflammation, antimicrobial defense, and resolution. Dysregulation of these pathways can contribute to SLE onset and drive pathogenesis, underscoring the importance of macrophages as therapeutic targets in SLE management. While cellular and molecular mechanisms of GC and O3FA treatments have been individually studied for their anti-inflammatory effects on macrophages (13, 31, 72–74), the combined impact of these treatments on transcriptional networks remains unexplored. Here, we hypothesized that O3FAs could be used as adjuncts to reduce GC dosages needed to suppress proinflammatory and type I IFN responses in lupus macrophages. We addressed this question by determining how cotreatment with DEX and DHA modulates critical regulatory hubs and the transcriptional landscape in LPS-stimulated NZBWF1 FLAMs. Following LPS stimulation, FLAMs derived from NZBWF1 mice demonstrated more robust type I IFN responses relative to FLAMs derived from C57BL/6 mice, which do not develop lupus, indicating their suitability as an experimental model for evaluating therapeutic strategies targeting hyperinflammatory macrophages in SLE. Our findings, as summarized in Figure 11 , support the conclusion that DHA+DEX cotreatment synergistically quells LPS-induced changes in transcription and regulatory factor activities, markedly attenuating expression of IFN-stimulated and proinflammatory genes that contribute to SLE pathogenesis. The demonstrated synergy between O3FA and GC in lupus macrophages closely mirrors prior findings on the separate effects of these agents across TLR, NF-κB, AP-1, STAT, and IRF signaling axes. This synergy reveals novel crosstalk mechanisms between O3FAs and GCs, highlighting the potential therapeutic value of combining lipidomic and pharmacological approaches to combat SLE-associated inflammation.

O3FA and GC synergy is highly consistent with previous investigations of the individual effects of these agents on TLRs, NF-κB, AP-1, STATs, and IRFs. O3FAs, notably DHA, disrupt TLR signaling through biophysical and structural mechanisms. DHA’s highly unsaturated conformation prevents stable interaction with MD2, a co-receptor for TLR4, effectively blocking TLR4 dimerization and downstream NF-κB activation (75). Beyond direct receptor interference, DHA increases membrane fluidity by incorporating phospholipid bilayers, which disperses lipid raft microdomains critical for TLR4 colocalization with CD14 (76). These biophysical effects impair receptor clustering and signaling amplification, highlighting a dual mechanism of action via direct structural inhibition and indirect membrane remodeling. GC treatment can significantly reduce expression levels of TLR4 and MyD88 in monocytes (77). GCs also induce the expression of mitogen-activated protein kinase phosphatase-1 (MKP-1), which inhibits p38 MAPK activation downstream of TLR4, dampening cytokine production in macrophages (78). Additionally, GCs regulate TLR signaling through microRNA-mediated mechanisms, such as increasing miR-511-5p expression, which directly targets TLR4 to inhibit the production of proinflammatory cytokines (79).

O3FA suppression of NF-κB is mediated through interference with both canonical and non-canonical inflammatory pathways. DHA and EPA competitively inhibit arachidonic acid metabolism, reducing proinflammatory prostaglandin E2 and leukotriene B4 production, which are known to enhance NF-κB activity (80). Oxidized metabolites of O3FAs, such as 18-HEPE and 17-HDHA, exhibit enhanced potency by activating PPARα, which sequesters NF-κB coactivators and promotes its nuclear export (81, 82). O3FAs also directly inhibit IκB kinase (IKK) phosphorylation, preventing IκB degradation and NF-κB nuclear translocation in macrophages (83). GC interference with NF-κB signaling is multifaceted and central to its anti-inflammatory effects. GCs induce the synthesis of IκBα, which sequesters NF-κB in inactive cytoplasmic complexes, preventing its nuclear translocation and transcriptional activity (84). The GC receptor (GR) physically associates with the p65 subunit of NF-κB, disrupting its DNA-binding and transcriptional activation capabilities (55). GCs also induce the expression of GC-induced leucine zipper (GILZ), which binds to the transactivation domain of activated NF-κB p65, further inhibiting its activity (85).

O3FAs attenuate AP-1 signaling by targeting MAPK cascades. In murine macrophages, O3FAs suppress p44/42 and JNK/SAPK phosphorylation—steps preceding AP-1 activation—which leads to reduced AP-1 activity and subsequent downregulation of proinflammatory cytokine genes in macrophages, confirming transcriptional-level anti-inflammatory effects (86). EPA suppresses phosphorylation of p38 MAPK and JNK/SAPK in human monocytic THP-1 cells, reducing c-Fos/c-Jun heterodimer formation and AP-1 DNA-binding activity (87). O3FA inhibition of AP-1 may be linked to the upregulation of MAPK phosphatase-1 (MKP-1), which dephosphorylates and inactivates JNK (88). GC-mediated suppression of AP-1 signaling occurs through several mechanisms. GRs physically interact with c-Jun and c-Fos, components of AP-1, inhibiting their transcriptional activity without requiring GR binding to DNA (89, 90). GCs also inhibit the phosphorylation and activation of JNK, thereby reducing AP-1 activity (91, 92). The induction of MAPK phosphatase-1 (MKP-1) by GR activation further suppresses AP-1 activity by dephosphorylating and inactivating JNK (93).

O3FAs attenuate STAT activation. We have previously found that DHA inhibits the expression of STAT1/STAT2-target genes in LPS-treated macrophages (31). RvD2, a pro-resolving metabolite of DHA, suppresses phosphorylation of STAT1 in bone marrow-derived macrophages (94). DHA and its metabolites also inhibit STAT3 phosphorylation in cancer cells (95–98). GCs primarily inhibit STAT1 through the induction of SOCS1, which inhibits STAT1 activation by degrading phosphorylated JAK2 (74). Furthermore, GCs suppress TLR-mediated STAT1 phosphorylation at Ser727 and Tyr701 during later phases of activation, impairing its nuclear translocation and transcriptional activity. Recruitment of GR to DNA-bound STAT3 is associated with trans-repression or transcriptional antagonism (99).

O3FAs indirectly regulate IRFs and IFNAR signaling by inhibiting ISG expression in LPS-treated macrophages (31). DHA also attenuates IFNAR signaling by reducing STAT1 phosphorylation, thereby blunting IFN-driven inflammatory gene expression (100). Meanwhile, GCs have been shown to interfere with IRF signaling by suppressing STAT1 mRNA transcription, leading to reduced activation of IRF-dependent pathways and diminished IFN-inducible gene expression, particularly in macrophages (101). DEX inhibits IRF3 phosphorylation and nuclear translocation in macrophages by suppressing TBK1, a kinase crucial for IRF3 activation (102). The GC receptor sequesters GRIP1, a coactivator for IRF3 and IRF9, preventing their transcriptional activity and disrupting the activation of ISGs (72). GCs antagonize the co-recruitment of IRF3 and NF-κB subunit p65 to ISRE-containing promoters, thereby reducing ISG transcription (72). GCs interfere with IFN receptor signaling by inhibiting the assembly of the STAT1-STAT2-IRF9 (ISGF3) transcription complex, essential for type I IFN signaling, and also prevent the nuclear translocation of IRF9, a critical step for triggering IFN-responsive gene expression (103). Furthermore, GCs block IFN-induced IRF1 mRNA levels, disrupting transcriptional activation of interferon-responsive genes regulated by IRF elements in the GC receptor promoter region (104). These findings suggest a strong synergistic potential between O3FAs and GCs that are amenable to in vitro and in vivo investigations.

Figure 11 illustrates how DHA+DEX co-therapy hypothetically impacts LPS activation targets in NZBWF1 FLAMs as revealed by deconvolution analysis. These agents appeared to act on LPS-induced activation of both MyD88-dependent and TRIF-dependent pathways to drive M1 macrophage polarization (105). The MyD88 pathway triggers IκBα degradation via IKK, enabling NF-κB subunits (e.g., NFKB1, REL) to translocate to the nucleus (106), while parallel MAPK activation phosphorylates AP-1 components (e.g., JUN, FOS) (107). These TFs collaborate with coactivators like p300 to remodel chromatin, initiating robust transcription of proinflammatory cytokines such as TNF-α, IL-6, and IL-1β (106). Simultaneously, the TRIF pathway phosphorylates IRF7, which synergizes with NF-κB and AP-1 at the IFN-β enhanceosome. This multi-protein complex recruits p300/CBP to stabilize enhancer assembly and drive IFN-β production (105, 108). Resultant IFN-β activates the expression of ISGs, including IRF1, via the JAK/STAT/IRF9 pathway. Although baseline IRF1 levels are constitutively low in resting macrophages, it integrates into the enhanceosome complex when induced, binding regulatory elements to potentiate IFN-β and ISG transcription (109). These actions constitute an autocrine/paracrine loop where IFN-β reinforces its production, enhancing antiviral responses and solidifying M1 polarization through sustained enhancer activity. Altogether, DHA+DEX suppressed this cooperative interplay between NF-κB, AP-1, and IRFs at the enhanceosome.

In complementary fashion, DHA+DEX co-therapy inhibits LPS-induced suppression of multiple transcriptional and post-transcriptional regulators such as EGR1, YBX1, E2F, GLI2, MYC/MYCN, and NFIC that form a collaborative network for suppressing macrophage inflammatory signaling. For example, EGR1, in association with the NuRD complex, drives chromatin compaction and decreases accessibility at inflammatory enhancers (110), while YBX1 exerts post-transcriptional control by binding to and silencing inflammatory gene mRNAs (48, 49). This dual-layer repression system may be especially critical for maintaining stringent control over key cytokine loci, as disruption of either EGR1 or YBX1 only partially restores inflammatory responses in experimental models (54, 56). The inhibition of GLI2 and E2F family members adds additional redundancy, with GLI2 attenuating NF-κB through Hedgehog signaling and E2F proteins modulating NF-κB dynamics and competing with AP-1 at shared promoters (50, 57). Meanwhile, MYC’s role in driving glycolytic flux and stabilizing IRF4 introduces a metabolic checkpoint that intersects with NFIC’s transcriptional regulation of PTEN and SENP8, which together serve to reduce oxidative stress and further dampen immune activation.

In addition to these LPS-sensitive regulators, DHA+DEX also enriched suppressive TF activities that were not markedly affected by LPS, including BCL6, NFATC2, and HNF4A. BCL6 acts as a transcriptional repressor to suppress NF-κB-driven proinflammatory genes (e.g., IL-6, Ccl2) and restrain type I IFN signaling (52, 111). At the same time, NFATC2 integrates calcium signaling, TLR4 activation, and interferon pathways to regulate macrophage immunity (112). HNF4A promotes M2 macrophage polarization via the NCOA2/GR/STAB1 axis and attenuates acute-phase gene expression, with its activation linked to improved outcomes in models of sepsis (113). The shared enrichment of these factors by DHA and DEX suggests that these therapies reinforce multiple layers of anti-inflammatory control, potentially providing a broader and more robust defense against excessive immune activation.

DHA alone positively enriched regulatory TFs that promote anti-inflammatory and reparative functions, including NFE2L2, CEBPB, RXRA, TWIST1, STAT5A, FLI1, and FOXO1. These TFs play essential roles in resolving inflammation and maintaining macrophage homeostasis. NFE2L2 interferes with LPS-induced transcriptional upregulation of proinflammatory cytokines, including IL-6 and IL-1β, in macrophages by binding to the proximity of these genes and inhibiting RNA polymerase II recruitment (114). CEBPB is essential for M2 macrophage polarization, driving anti-inflammatory genes like Arg1 and IL-10 via CREB-mediated induction, which is critical for resolving tissue damage (65). RXRα (retinoid X receptor α) plays a significant role in modulating the host’s antiviral response by regulating the production of type I IFNs, particularly IFN-β (66, 115). RXRA also contributes to anti-inflammatory effects by modulating nuclear receptor-mediated gene expression networks. STAT5A activation promotes M2 macrophage polarization, favoring tissue repair and the production of anti-inflammatory cytokines (64). TWIST1 induces M2 macrophage polarization by upregulating profibrotic factors (ARG-1, CD206, IL-10, TGF-β) through direct activation of galectin-3 transcription, enhancing M2 phenotypic plasticity (63). FLI1 loss has been linked to increased IFN-regulated expression (116). FOXO1 activity is associated with both the M1 and M2 phenotypes, suggesting a more complex role in regulating macrophage polarization (68).

DHA monotherapy also uniquely negatively enriched the activity of other proinflammatory regulators, including SP3, PGR, and RFX5. SP3 plays a critical role in promoting proinflammatory macrophage activation (M1 phenotype) by driving NF-κB-mediated transcription activation. When SP3 activity is diminished, macrophages exhibit decreased expression of M1-associated proinflammatory genes, such as Nos2, Tnfa, Il1b, and Il6, and increased expression of M2-associated anti-inflammatory markers like Arg1 and Retnla. PGR, the progesterone receptor, modulates macrophage function through its activation. Stimulation of membrane-bound PGRs increases the transcription of proinflammatory genes such as Il1b, Tnfa, and Nos2 (117), suggesting that decreased PGR activity likely suppresses these proinflammatory responses, reducing the production of inflammatory cytokines. RFX5 regulates MHC class II expression and macrophage antigen presentation (118, 119). Reduced activity of RFX5 could impair antigen presentation capacity and potentially alter cytokine signaling pathways, indirectly influencing inflammatory responses. Collectively, the suppression of these TFs by DHA likely shifts macrophage activity away from a proinflammatory phenotype.

Interestingly, DHA alone also reduced enrichment scores for suppressive factors like TCF7L2 at the 4-hr time point and CREB1, RARα, SMAD3, and GATA3 at the 8-hr time point. TCF7L2 modulates inflammatory cytokine expression in macrophages by promoting polarization toward the anti-inflammatory M2 phenotype, which suppresses proinflammatory cytokines like TNF-α and IL-6 (120). CREB1 (cAMP response element-binding protein 1) primarily suppresses proinflammatory cytokines like TNF-α and inhibits NF-κB signaling in macrophages, maintaining anti-inflammatory responses. Reduced CREB1 activity diminishes IL-10 production and decreases NF-κB suppression, amplifying proinflammatory gene transcription (121). RARα is known for its anti-inflammatory effects through modulating macrophage polarization and proinflammatory gene expression (122). SMAD3 promotes an anti-inflammatory macrophage phenotype via TGF-β signaling; its suppression increases proinflammatory cytokines such as IL-1β and TNF-α while reducing anti-inflammatory mediators like IL-10 (123). GATA3 promotes M2 differentiation and suppresses proinflammatory cytokines; reduced GATA3 activity can increase these cytokines and ISGs (124). Given their anti-inflammatory roles, decreased enrichment of these pro-resolving regulatory factors by DHA might reflect complex, time-dependent homeostatic control and suggest the need for further investigation.

DEX monotherapy also negatively enriched TFAP2A and ATF4, which likely dampened inflammatory pathways, as these factors regulate stress-responsive and cytokine genes (125, 126). Accordingly, DEX orchestrates a dual mechanism in macrophages by activating anti-inflammatory TFs while suppressing key proinflammatory regulators. This comprehensive modulation attenuates LPS-induced cytokine and IFN-regulated gene expression, thereby reprogramming macrophages toward an anti-inflammatory state.

DEX alone further selectively enriched for factors that promote resolution, including NR3C1, PPARγ, ARNTL, CLOCK, LEF1, and TFDP1. NR3C1 (glucocorticoid receptor, GR) directly inhibits proinflammatory TFs AP-1 and NF-κB, a key mechanism for suppressing inflammation (127). PPARγ reinforces these anti-inflammatory effects by inhibiting AP-1 and NF-κB, another mechanism that promotes the M2 phenotype (128). ARNTL (BMAL1) and CLOCK are circadian regulators that suppress LPS-induced proinflammatory genes (e.g., TNF-α, IL-6) in macrophages by competitively displacing NF-κB/AP-1 at enhancers and reducing H3K27ac histone acetylation, thereby limiting transcriptional activation (59, 60). ARNTL further antagonizes STAT1-mediated IFN-β signaling and stabilizes metabolic rhythms, while CLOCK reinforces these anti-inflammatory effects by curbing excessive enhancer remodeling. LEF1 expression is positively correlated with the M2 phenotype (129). Although the role of TFDP1 in inflammation remains unclear, its association with E2F factors suggests regulatory influence over immune response genes (53).

Although we present herein compelling evidence for O3FA+GC synergy in inhibiting LPS-induced inflammation in SLE macrophages, we acknowledge that our study has limitations. First, our use of an in vitro LPS-activated NZBWF1 macrophage model, while mechanistically informative, does not capture the full complexity of human SLE, as it lacks multicellular interactions and the tissue-specific microenvironment present in an in vivo model. Also, the lack of non-lupus control in these studies limits the generalizability to other diseases. Second, our analysis was limited to early transcriptional responses (4–8 hr post-LPS treatment), creating uncertainty about whether the observed DHA and DEX synergy is sustained or subject to rebound effects during chronic inflammation. Third, although key transcriptional regulators were identified by functional enrichment, the precise molecular mechanisms underlying the observed synergy, such as direct receptor crosstalk, demonstration of altered TF activity, epigenetic changes, or metabolic reprogramming, remain uncharacterized. Finally, the efficacy and safety of low-dose DHA and DEX cotreatment have not yet been validated in preclinical SLE models or clinical settings, where factors such as bioavailability, off-target effects, and patient heterogeneity could significantly impact therapeutic outcomes. These limitations highlight the need for further mechanistic, longitudinal, and preclinical studies to advance this promising combinatorial approach toward clinical application.