Human monocytic THP-1 and murine macrophage RAW 264.7 cells obtained from American type culture collection (ATCC^® TIB-202™ and TIB-71™, respectively) were cultured in RPMI-1640—GlutaMax™ media (Life Technologies™) and DMEM 6546 (Sigma-Aldrich^®), respectively. MEF cells lacking NF-κB subunit p65 (p65^−/−), p50 (p50^−/−), p52 (p52^−/−), Relb (Relb^-/−), and wild-type (WT) controls were a generous gift from Alexander Hoffman (UCLA) and cultured in DMEM 6546 (Sigma-Aldrich^®). All cell media were supplemented with 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin. Cells were cultured at 21% O₂, 5% CO₂, and maintained in a humidified tissue culture incubator at 37°C balanced using nitrogen. Stable knockdown cells were maintained in an additional 5 µg/ml Puromycin dihydrochloride (Sigma-Aldrich^®) for stable knockdown selectivity. Reagents such as LPS, BAY-11-7082, and Cytosporone-b (Csn-B) were supplied by Sigma-Aldrich, and TNFα was supplied by RnD systems.

Peripheral venous blood was collected from healthy volunteers at the Conway Institute, University College Dublin (UCD). Institutional review board approval was obtained from the Ethics Committee at St. Vincent’s University Hospital (UCD affiliated teaching hospital), and written informed consent was obtained from all volunteers. A total of 30 ml of blood was collected into a syringe, layered slowly onto polymorphoprep solution (1:1), and centrifuged at 500 × g for 35 min at 20°C with the brake off. The mononuclear layer was removed and mixed with equal volumes of 0.45% NaCl by gentle inversion followed by centrifugation at 400 × g for 10 min at 20°C with the brake on. Supernatant was discarded and pellet was re-suspended in 12 ml ice-cold water and mixed by inverting gently for 1 min followed by the addition of 12 ml 1.8% NaCl and subsequent centrifugation at 300 × g for 5 min at 20°C with the brake on. Cells were then re-suspended in media, counted, and seeded at a density of 2.5 × 10⁵ cells/ml for RNA experiments.

The study was approved by the Ethics Committee of St. Vincent’s University Hospital, Dublin, and in accordance with International guidelines and Helsinki declaration principles. All patients (n = 6) gave written informed consent. This patient cohort has been recently described and analyzed (16). Patients, symptomatic and asymptomatic, with clinical and angiographic evidence of atherosclerosis undergoing revascularization surgery were recruited to the study, and clinical details obtained for patients are supplied as Table S1 in Supplementary Material. Three “symptomatic” patients [defined as patients who had had a first-ever transient ischemic attack or non-disabling ischemic stroke from atherosclerotic stenosis and without any known history or diagnosis of cardiovascular disease (CVD)]; and three “asymptomatic” patients (defined as those without any previous ischemic events or known CVD) were studied.

Immediately following carotid endarterectomy, human carotid atherosclerotic plaque specimens were placed in RNAlater^® solution (Thermo Fisher), and stored at −80°C. Each plaque specimen was then dissected into four distinct sections: common carotid (CC), Internal Carotid (IC), external carotid (EC), and relatively disease free (RDF), as shown in the schematic (Figure S3A in Supplementary Material). Subsequent analysis was performed on IC and RDF samples only as described below in qRT-PCR methods.

Stable knockdown of human NR4A2, NR4A3, and murine NR4A2 was achieved with the transduction of shRNA using a Mission™ Lentiviral packaging mix as per instructions (Sigma-Aldrich) using THP-1. For controls, cells were transduced with scrambled shRNA (Sigma-Aldrich). Briefly, for THP-1 cells, 150 µl of cells from a stock concentration of 2.5 × 10⁵ cells/ml were pipetted into a 96-well plate; for Raw mac 264.7 cells, 250 µl of cells from a stock concentration of 1.5 × 10⁵ cells/ml were pipetted into a 24-well plate and both were incubated overnight. Hexadimethrine bromide was added to a final concentration of 8 µg/ml followed by the addition of 10 µl of lentiviral particles and subsequent incubation for 48 h in a tissue culture incubator at 37°C. Subsequent to this, puromycin was added to the cells (5 µg/ml), and medium was changed every 3–4 days until resistant colonies were identified and subsequently expanded. Western blotting was used to confirm knockdown levels for human THP-1 cells and qRT-PCR was used for murine Raw mac 264.7.

Mouse embryonic fibroblast cells were transfected with previously described/published plasmids p-CMX-NR4A2 (WT) and p-CMX-NR4A2 mutant (C283G) (10, 17). MEF cells were trypsinized, spun down to obtain pelleted cells, re-suspended in fresh media, and brought to a final cell concentration of 1.2 × 10 cells/ml. Optimem, plasmid, and transfection reagent (TransIT2020) (Mirus) were mixed and left for 30 min to allow complexes to form as per manufacturer’s instructions in the ratio 100 μl:500 ng:3 μl (Optimem:Plasmid:TransIT2020). Subsequently 100 µl of mixture/complexes was added to 2 ml of cell suspension in 6-well plates and incubated for 48 h. At 48 h, cells were treated as described in figure legends.

Whole-cell lysates: suspension cells (THP-1) were pelleted by centrifugation @ 1,200 RPM followed by re-suspension into 100 µl radio-immuno-precipitation assay (RIPA) buffer. The re-suspended pellets were then pipetted back onto their respective treatment dishes and were scraped into a new Eppendorf tube to ensure collection of floating cells and cells that adhered during the course of stimulations. Adherent cells (Raw mac 264.7) and MEFs were scraped directly into RIPA buffer. Cell suspensions were then agitated vigorously on a rocker for 30 min at 4°C followed by centrifugation at 15,000 × g for 15 min. Supernatants were removed as whole-cell extracts and stored at −20°C.

Protein content was assessed by the Bradford method (Bio-Rad Laboratories) and electrophoresed on 10% SDS PAGE gel followed by transfer to nitrocellulose membranes (Bio-Rad). Protein expression levels were measured by Western blot analysis using specific antibodies for NR4A2 and NR4A3 (RnD systems); p65 (Cell Signaling); Relb (Cell Signaling); CTGF (Santa Cruz); and β-tubulin (Sigma). Briefly, nitrocellulose membranes were incubated with primary antibodies overnight at 4°C followed by washing three times for 10 min with 1× TBST [20 mM Tris, 137 mM NaCl, and 0.05% (v/v) tween 20] and subsequent incubation for 1 h with species-specific HRP-conjugated secondary antibodies at room temperature. Membranes were washed again as described followed by signal detection using enhanced chemiluminescence reagent (Pierce). Densitometric analysis included for Western blot data was determined using LI-COR^® Image Studio Lite version 3.1. Briefly, the intensity of proteins of interest (NR4A2, NR4A3, p65, Relb, and CTGF) was quantified using LI-COR program relative to loading control protein (β-tubulin).

Medium from THP-1-treated cells was centrifuged briefly to remove cell debris, and supernatant was stored at −20°C. MCP-1 and MIP-3α (RayBiotech) protein levels were measured using an enzyme-linked immunosorbent assay. IFNγ, IL-12p70, IL-10, IL-1β, IL-6, IL-8, and TNFα were measured using the Human ProInflammatory 7-Plex Tissue Culture Kit (MSC). ELISA was performed as per manufacturer’s instructions. Changes incorporated into the protocols specific for our data are as follows. For each ELISA, the dilution value was calculated to yield optimum results within the linear range of the assay. From this, we determined that samples were used neat on the 7-Plex ELISA, diluted 1:10 for MCP-1 and 1:6 for MIP-3α.

RNA was extracted from in vitro cells using the column-based EZNA^® total RNA extraction kit (Omega Bio-Tek, GA, USA). Trizol (Invitrogen) extraction was used to isolate total RNA from patient samples stored in RNAlater^®. cDNA synthesis was subsequently performed as previously reported (14). qRT-PCR was performed using SYBR Green Master Mix (Applied Biosystems) and on an ABI 7300 thermocycler (Applied Biosystems). Primer pair sequences used are detailed in Table S2 in Supplementary Material. Relative expression/abundance levels of target gene transcripts were determined using qBase plus software (Biogazelle, Ghent University, Belgium) with GAPDH or β-actin (in vitro samples) or ribosomal 18s (patient samples) as a reference target. Results are expressed as fold over untreated control (FOC).

An in-house PCR array, consisting of 92 inflammatory genes and 4 reference genes, was performed. For this array, cDNA synthesis was performed using 500 ng of total RNA using the RT² First Strand Kit (Qiagen Ltd.) according to the manufacturer’s instructions. Real-time reverse transcriptase PCR was performed on a 20 µl reaction mixture per well containing 1 µl cDNA (after 1:5 dilution), 9 µl water, and 10 ml RT² SYBR Green/ROX qPCR master mix (SABiosciences) in a 7300 Real-time PCR system (Applied Biosystems). Using a web-based PCR array data analysis tool (SABiosciences), the C_T values of the target genes were normalized against the four reference genes (GAPDH, HPRT1, TBP, and ACTB) and the relative abundance of gene transcripts were calculated as log 2△CT. Subsequently, expression fold changes were calculated as fold over untreated control sample (FOC) and are displayed in Table S3 in Supplementary Material, alongside raw C_T values. Using the PCR array data tool, a clustergram was made and displayed in Figure 6A. A clustergram performs non-supervised hierarchical clustering of the entire data set and is represented as a heat map with a dendrogram showing individual genes and overall treatment groups with a similar regulation pattern.

The Genomatix software suite program MatInspector (© Genomatix Software GmbH) was used to analyze the human and murine MIP-3α promoter 500 bp upstream of the transcription start site (TSS). Two stringency tests were applied as per Genomatix software instructions in order to select binding sites that displayed the best probability as “real” binding motifs for transcription factors (Core similarity of 1.0 and matrix similarity >0.90). The first test is the core similarity test and the maximum core similarity of 1.0 is only reached when the highest conserved bases of a matrix match exactly in the given sequence. The second test is the matrix similarity test and the score must be greater than 0.90.

All data is presented as mean ± SEM and representative Western blots for a minimum of three individual n numbers (unless otherwise stated). Statistical significance was performed using a two-tailed Student’s t-test for patient tissue analysis and one-way analysis of variance followed by Tukey’s post hoc test was utilized throughout all other results (GraphPad Prism^®).

TLR4 ligands, including LPS, have been shown to drive NR4A gene expression in immune cells (14, 18). Several studies have shown in both human and mouse macrophages, using NR4A1–3 overexpression and/or knockout cells, that NR4A receptors regulate expression of several downstream NF-κB target genes including IKKi, TNFα, MCP-1, iNos, IP-10, and IL-12 (9, 11, 13–15). NR4A family members can modulate NF-κB activity downstream of TLR4 signaling in a dynamic fashion, either repressing or enhancing NF-κB target gene expression (9, 11, 13–15). The majority of the work to date suggests that NR4A1 and 2 depletion promotes a pro-inflammatory macrophage phenotype, displaying enhanced NF-κB activity (9, 11, 13–15). However, there are minimal studies elucidating the role monocyte/macrophage NR4A3 plays in modulating such regulatory pathways.

To examine TLR4-driven NR4A2 and NR4A3 gene and protein expression, we utilized a range of human myeloid cells including primary human mononuclear cells (PBMCs), THP-1s, Phororbol-12-myristate-13-acetate (PMA) differentiated THP-1s, and murine RAW mac264.7 cells. In human PBMCs, we reveal temporal changes in NR4A1–3 gene expression in response to LPS (1.0 µg/ml) (Figure 1A). We observe NR4A1 gene expression rises sharply and significantly at 1 h and is reduced by 2 and 3 h. NR4A2 gene expression is significantly increased at 1 h albeit to a very low level, while NR4A3 gene expression is significantly increased at 1, 2, and 3 h post stimulation (Figure 1A). Examining PMA (20 nM) differentiated THP-1 cells reveals a significant induction of all NR4A family members at 2 h exposed to 1.0 µg/ml LPS (Figure 1B). Using undifferentiated THP-1 cells, we observe a significant and selective induction of NR4A2 and 3 gene expression within 2 h LPS (100 ng/ml) treatment (Figure 1C) with minimal changes in NR4A1 expression.

Thus, we examined temporal levels of NR4A2 and 3 protein expression in undifferentiated THP-1 cells exposed to LPS (1.0 μg/ml) (Figure 1D). Undifferentiated THP-1 cells were chosen for extended analysis based on their similarity to the expression profile observed in primary PBMCs for NR4A receptors in response to LPS. Additionally, we have previously shown undifferentiated THP-1 cells closely reflect primary cell responses to various stimuli such as LPS and adenosine and support that they are a relevant immortalized cell line to use in such myeloid cell studies (14). NR4A3 protein is dramatically increased within 4 h and maintained at 6 h post LPS, while NR4A2 is enhanced as early as 2 h with protein levels reducing over 4–6 h (Figure 1D) revealing distinct temporal changes in protein expression under these cellular conditions. We and others have shown previously that LPS-driven NR4A transcription is NF-κB dependent (14, 18). Here, we support the studies confirming that pretreatment with the NF-κB inhibitor BAY 11-7082 (10 µM) abolishes LPS-driven NR4A2 and 3 protein expression (Figure 1D). We next examined the expression of NR4A genes in response to LPS in the murine macrophage cell line Raw Mac 264.7 (Figure 1E). LPS (1 µg/ml) significantly drives NR4A1 and NR4A2 gene expression within 1 h; in contrast, NR4A3 mRNA levels remain unchanged (Figure 1E). Using the NF-κB inhibitor BAY 11-7082 (NF-κBi) (10 µM), we further establish that LPS-induced NR4A1/2 is NF-κB dependent in these cells (Figure 1E).

NR4A2 and 3 virally delivered shRNA depleted cells were utilized (14) to further identify the transcriptional role of NR4A2 and 3 in TLR4 driven gene expression. As shown in Figure 1D, which incorporates NR4A2/3 depleted cells alongside scrambled control cells, we demonstrate a significant depletion of NR4A2 (≈96%) and NR4A3 (≈79%) (% calculated from densitometry band intensity values) (Figure 1D). Taken together, we establish that LPS drives temporal expression of NR4A1–3 genes in human and murine myeloid cells and further demonstrate that LPS-driven NR4A2 and 3 expression is NF-κB dependent.

NR4A receptors have the ability to promote or repress NF-κB target gene expression, albeit to date such divergent mechanisms have been established distinctly, and not simultaneously. We therefore investigated the effect of NR4A depletion on a number of known NF-κB target genes in myeloid cells. A multiple ELISA was performed to quantify IFNγ, IL-12p70, IL-10, IL-1β, IL-6, IL-8, TNFα, and additional single ELISA for MCP-1 and MIP-3α expression levels. These targets were selected having identified them as NF-κB-regulated genes using the NF-κB inhibitor BAY-11-7082 in THP-1 cells [data not shown, Figure 2 and Ref. (9, 14)]. Second, NR4A receptors have been shown to be critical in regulating their appropriate expression during inflammatory responses by our group and others (3, 9, 10, 14, 19).

ELISA analysis reveals NR4A2 and 3 depletion results in a more pronounced TLR4 response, displaying enhanced LPS-driven IL-12p70, Il-1β, IL-10, TNF-α, and IL-6 protein secretion (Figure S1 in Supplementary Material). We also observed enhanced protein secretion and mRNA expression of MCP-1 within NR4A2 and 3 depleted cells, which was more profound than any of the other targets described above (Figures 2A,B). Furthermore, we show that this LPS-induced MCP-1 is inhibited by the use of the NF-κB inhibitor BAY-11-7082 (NF-κBi), confirming MCP-1 to be an NF-κB-dependent gene (Figure 2C). As NR4A2 and NR4A3 display distinct expression changes in murine Raw mac 264.7 cells (Figure 1E), stably depleted NR4A2 cell line was selected using viral delivery of shRNA directed against NR4A2 and scrambled control. Gene expression analysis reveals that 95% depletion of LPS (1 µg/ml) induced NR4A2 gene expression in NR4A2 shRNA-transduced cells compared to scrambled control (Figure 2D). We further reveal scrambled Raw mac 264.7 cells exposed to LPS for 1 h results in significant MCP-1 gene expression, and this regulation is further potentiated with NR4A2 depletion (Figure 2E). Additionally, using the NF-κB inhibitor BAY 11-7082 (NF-κBi), we show that LPS-driven MCP-1 in Raw mac 264.7 cells is also NF-κB dependent (Figure 2F).

Intriguingly however, having demonstrated that NR4As negatively regulate the above NF-κB target genes, we observe LPS-dependent secretion of MIP-3α protein is significantly attenuated in NR4A2 and NR4A3 depleted cells (p ≤ 0.05 and p ≤ 0.01, respectively) (Figure 3A). Further analysis shows LPS drives MIP-3α gene expression within 2 h and expression returns to basal levels by 8 h. However, this induction is significantly attenuated (≈50%) in cells depleted of NR4A2 or NR4A3 expression (Figure 3B). Additionally, we observe a significant induction in MIP-3α gene expression in human PBMCs treated with 1 µg/ml LPS for 2 h (Figure 3C). Furthermore, LPS significantly drives MIP-3α mRNA in undifferentiated THP-1 cells and use of the NF-κB inhibitor, BAY 11-7082 (NF-κBi) (10 µM), confirms this induction is NF-κB dependent (Figure 3D). To further support NF-κB-dependent regulation of MIP-3α, we utilized mouse embryonic fibroblast (MEF) cells lacking the NF-κB subunit p65 (p65^−/−) and WT control (Figure 3E). We demonstrate that LPS drives MIP-3α gene expression in WT MEF cells and that this induction is abolished in the p65^−/− MEF cells (Figure 3F). Raw mac 264.7 cells exposed to 1 µg/ml LPS for 1 h results in a significant induction of MIP-3α gene expression and that this induction is significantly attenuated in NR4A2-depleted cells (Figure 3G). Using the NF-κB inhibitor BAY 11-7082 (NF-κBi) (10 µM), LPS-driven MIP-3α is shown to be NF-κB dependent in Raw mac 264.7 cells (Figure 3G). Importantly, using the MEF cells lacking the NF-κB subunit p65 (p65^−/−) analysis confirms a significant reduction in LPS-driven MIP-3α gene expression compared to WT control (Figure 3F), and here we reveal that NR4A2 and NR4A3 expression is also reduced in the MEF p65^−/− cells during TLR4 stimulation using LPS compared to WT cells (Figure 3H).

In order to address the role of NR4A1 in mediating MIP-3α expression, we utilized the NR4A1 agonist Cytosporone-B (Csn-B). We confirm Csn-B activity within our cell system by examining the expression of TNFα, previously shown to be repressed by NR4A1 activity in myeloid cells (20). Here, we confirm that Csn-B co-treatment significantly attenuates LPS-driven TNFα (Figure 3I). Further analysis reveals that stimulation of THP-1 cells with LPS (100 ng/ml), in the presence of Csn-B, does not significantly alter MIP-3α induction (Figure 3J).

We next investigated whether this differential regulation of NF-κB targets by NR4A receptors was specific to LPS-TLR4-driven inflammatory responses by using TNFα in the above systems. Here, we reveal that TNFα drives MIP-3α expression and protein secretion in primary and THP-1 cells (Figures S2A,B in Supplementary Material). Additionally, we show that TNFα-driven MIP-3α is NF-κB dependent using the BAY 11-7082 NF-κB inhibitor (Figure S2C in Supplementary Material). We observe analogous regulation of MIP-3α expression in both scrambled control and NR4A2/3 depleted cells, demonstrating that this regulatory mechanism is not limited to TLR4 signaling (Figure S2D in Supplementary Material). Furthermore, we observe a similar pattern of regulation of MCP-1 in scrambled and NR4A2/3 depleted cells in response to TNFα (Figures S2E,F in Supplementary Material).

We have established using myeloid cells the involvement of NR4A2/3 receptors in the regulation of MIP-3α downstream of TLR4 or TNFα signaling. To expand these observations, we sought to examine the expression profile of NR4A2/3 and MIP-3α expression in vivo in human carotid plaque tissue, given the fact that both genes have been identified as important factors in the pathogenesis of atherosclerosis (21–25). Figure S3A in Supplementary Material identifies the specific regions of the carotid where portions examined are taken from, IC or RDF from asymptomatic and symptomatic patients, as described within the Section “Materials and Methods.” Our analysis reveals that, in symptomatic plaques, NR4A3 and MIP-3α expression is significantly elevated in the IC-diseased portion compared to their respective RDF or IC sections from asymptomatic patients. Interestingly, NR4A2 expression is not significantly increased in symptomatic patients compared to asymptomatic tissue (Figure S3B in Supplementary Material). Therefore, NR4A2 and NR4A3 in vitro have been shown to regulate MIP-3α gene expression, while in vivo, specific NR4A3 expression levels parallel MIP-3α expression suggesting NR4A3 may be involved in an in vivo setting, albeit patient numbers are limited and the role of NR4A3 regulating MIP-3α in such disease warrants further investigation.

Thus, data herein strengthen the hypothesis that NR4As are not merely repressors or enhancers of NF-κB target genes in isolation but can lead to distinct cellular outcomes by both positive and negative modulation of target gene expression concurrently. Thus, NR4A2 and 3 are key regulators of MIP-3α gene expression in human and murine myeloid cells. We demonstrate that NR4A2 and 3 receptors concurrently enhance and repress NF-κB-dependent target gene expression, positively regulating MIP-3α while negatively regulating targets such as MCP-1.

In order to investigate how NR4As were positively regulating MIP-3α expression in both human and murine, we performed promoter analysis using the Genomatix MatInspector software suite. Figure 4A shows a graphical representation obtained from Genomatix showing two NBRE (NR4A-binding motif) and one κB (NF-κB-binding motif) site on the human MIP-3α proximal promoter, and on NBRE and κB site on the murine promoter. Intriguingly, no NBRE sites were identified in the promoters of the negatively regulated NR4A target genes IL-12P70 (analyzed IL-12p40 and p35 subunits), IL-10, TNF-α, IL-6, and MCP-1, while one NBRE site was identified in IL-1β (data not shown). Expanded from this graphic is the sequence of both human and murine showing the NBRE and κB sites identified nearest the TSS of both promoters (Figure 4A). We and others have previously shown that p65 is capable of binding to this κB site nearest the TSS (14, 26).

Following this, we investigated whether NR4A receptors needed to bind their NBRE motif in order to enhance MIP-3α expression. Using MEF cells, we overexpressed p-CMX-NR4A2 WT plasmid and a mutant p-CMX-NR4A2 (C283G) plasmid, previously published by our group and others, which lacks the ability to bind to the NBRE sequence on DNA (10, 17). MEF cells displayed successful transfection with significant increases in NR4A2 mRNA in both WT and mutant over backbone control (278 ± 105- and 202 ± 49-fold over control, respectively). We reveal that overexpression of WT NR4A2 can enhance MIP-3α gene expression during LPS stimulation and the DNA-binding mutant NR4A2 does not effect this enhancement (Figure 4B). Furthermore, we show that IL-6 expression due to LPS stimulation is suppressed in both WT and mutant NR4A2 (Figure 4B). Importantly, we confirm using the MEF cells lacking p65^−/− that LPS-driven IL-6 in these cells is NF-κB dependent (Figure 4C). Thus, NR4A2 does not require the ability to bind to DNA in order to enhance or repress NF-κB target gene expression.

Following the knowledge that NR4A2 does not require DNA binding to enhance and repress NF-κB target gene expression, we decided to examine more extensively the regulatory elements involved in MIP-3α production. In order to do this, we used MEF cells lacking the NF-κB family members p65^−/−, p50^−/−, p52^−/−, and Relb^−/−. We have shown that MIP-3α is NF-κB subunit p65 dependent in MEF cells (Figure 3F); additionally, we show that LPS-driven MIP-3α is also NF-κB subunit p50 dependent (data not shown), clearly identifying that it is a canonical NF-κB target gene. Furthermore, we identify that LPS is still capable of driving MIP-3α expression in MEF cells lacking p52^−/− (data not shown), however intriguingly we show cells lacking Relb^−/− display significantly attenuated MIP-3α expression (Figure 5A). Given that MIP-3α in these cells is confirmed as a p65-dependent gene, we examined whether Relb was also regulated in this manner. Here, we reveal that LPS stimulation significantly drives Relb protein and gene expression in a NF-κB p65-dependent manner (Figures 5B,C). Additionally, we observe that LPS-driven Relb is NF-κB p50-dependent (data not shown). Figure 5B also includes the Relb^−/− MEF cells to confirm that the cells lack Relb expression. Furthermore, we show that LPS can drive Relb expression in THP-1 cells and that this is NF-κB dependent, shown using the NF-κB inhibitor BAY 11-7082 (Figure 5D). We also show that Relb gene expression is induced following LPS stimulation using primary PBMC cells (Figure 5E), confirming that this induction is an observation not limited only to immortalized THP-1 and MEF cells.

Taking into account that NR4As are positive regulators of MIP-3α in THP-1 (NR4A2 and 3), Raw mac 264.7, and MEF cells (NR4A2 only tested), we asked whether they were involved in LPS-driven Relb expression. Here, we reveal for the first time that Relb is a novel NR4A target gene, displaying a significant reduction in expression during LPS stimulation in the both NR4A2 and 3 depleted cells (Figure 5F). Thus, we have identified that MIP-3α is both a NR4A and Relb target gene during LPS stimulation, and further reveal that NR4As are important in the LPS regulation of Relb itself. Importantly, we show that NF-κB subunits p65 and p50 (canonical NF-κB) are upstream of all of these factors.

In order to expand our knowledge of NR4A2/3 regulated target genes during inflammatory pathogenesis, we performed an in-house PCR array. Scrambled control THP-1 cells, NR4A2 and 3 depleted cells were stimulated with LPS (100 ng/ml) for 2 h. A heat map represents changes in gene expression from the array and similarities in the pattern of gene regulation and overall treatment similarities are represented by the inclusion of a dendrogram (Figure 6A). All data obtained from the PCR array analysis are shown in Table S3 in Supplementary Material. Genes of interest were further investigated within the same experimental model and analyzed using qRT-PCR (Figure 6B). LPS significantly drives CTGF in scrambled control THP-1 cells, and this induction is significantly attenuated in NR4A2 and 3 depleted cells (Figure 6B). Intriguingly while LPS treatment alone does not alter PAI-1 expression, an established NR4A1 regulated target gene in human umbilical vein endothelial cells, MCF-7, and JURKAT cells during TNFα stimulation (27), NR4A2 depletion significantly potentiates LPS effects on PAI-1gene expression and NR4A3 depleted cells display significantly reduced PAI-1 levels. Using the NR4A1 agonist Csn-B (100 nM), we reveal that enhanced NR4A1 activity reduces CTGF expression (Figure 6C).

Here, we establish that TLR4 stimulation drives CTGF expression, an important atherogenic and fibrotic factor, and for the first time show NR4A receptor activity is crucial to this induction. To further characterize the mechanisms involved in this regulation, we establish that LPS (1 µg/ml) significantly drives CTGF protein expression as early as 2 h and this induction declines over 4 and 6 h (Figure 6D). Further analysis supports the changes in gene expression using the PCR array, whereby LPS (1 µg/ml) induces CTGF protein expression at 4 h, which is significantly attenuated in NR4A2 and 3 depleted cells (Figure 6E). Using the NF-κB inhibitor BAY 11-7082 (10 µM), we confirm that LPS (1 µg/ml) and TNFα-induced CTGF gene and protein expression is NF-κB dependent (Figures 6F,G; Figure S4 in Supplementary Material), consistent with other LPS-induced targets controlled by NR4A activity including MIP-3α and MCP-1. We next examined whether CTGF was a Relb target gene, similar to that of MIP-3α, the other positively regulated NR4A gene identified in this study. Unfortunately however, examination of the MEF cells reveals no enhancement of CTGF during LPS stimulation, a mechanism that may be cell type specific and therefore we could not investigate the Relb involvement using these MEF Relb^−/− cells. Moreover, promoter analysis shows no NR4A DNA-binding motifs (NBRE sites) on the promoter of CTGF (data not shown). Thus, we have identified NR4A receptors regulate novel target gene CTGF, an important fibrotic and atherogenic target gene.