Global knockout mice for NR4A1^−/−(Nur77), NR4A2^−/− (Nur1) and double knockout for NR4A1/2^−/− were generated by crossing NR4A1^flox/flox (obtained from the Jackson Laboratory) or NR4A2^flox/flox mice (kindly provided by Pierre Chambon (Sekiya et al., 2011), with CMV-GT-Rosa-CreERT2^TG/0. All knockout animals were generated on the C57BL/6 background. Global deletion of NR4A1 (A1KO) and/or NR4A2 (A2KO) was induced by administering tamoxifen (400 mg/kg in chow) for 10 days, followed by a 14-day tamoxifen-free “wash-out” period. In this study, control animals (WT) are defined as littermates, which did not receive tamoxifen treatment with the chow. All animals had free access to chow and water in a specified pathogen-free facility with a 12 h light/dark cycle and all animal experiments were performed in accordance with the German animal protection law and were carried out after approval by the local authorities (Regierungspräsidium Darmstadt, approval number FU1268). Every mouse received an identification number for each experiment and the experimenter was blinded for the genotype. Animal group sizes differed due to number of available littermates.

AEA is an oily substance, which limits application in vivo and controlled absorption. Therefore, the compound was applied as nanoformulation. The preparation of the AEA micellar nanoformulations was carried out as previously described (Martín Giménez et al., 2023). Briefly, 30 mg of the commercial co-polymer Pluronic^® F127 (PF127; BASF, CABA, Buenos Aires, Argentina) was accurately weighed and dissolved in 1 mL of Milli-Q water (Sigma-Aldrich, St. Louis, MO, USA) each. The mixture was continuously stirred at RT until homogeneous and a transparent dispersion was obtained. Subsequently, 750 µg of AEA, dissolved in absolute ethanol (using 15 µL of a commercial AEA ethanolic solution from Cayman Chemical, Ann Arbor, MI, USA), was incorporated (drop by drop) into the polymeric dispersion. Stirring was maintained until complete ethanol evaporation. AEA-free micellar nanoformulations of the PF127 polymer (Pluronic) served as control (CTL).

The AirPouch model (Paul-Clark et al., 2012; Pierron et al., 2023) was conducted on both tamoxifen-treated mice following intake and washout, as well as on untreated mice (CTL) that did not receive tamoxifen. Mice were anesthetized using isoflurane inhalation and placed on a heating pad to maintain body temperature. The dorsal skin region was shaved, and the skin was sterilized with 70% ethanol. An AirPouch was created by subcutaneous injection of 5 mL sterile air using a 23-gauge needle, followed by additional injects of 3 mL after 3 days to maintain the pouch. On day 6, 30 µg diclofenac and 10 µg AEA-Pluronic nanoparticles or Pluronic control nanoparticles without AEA (in 0.5 mL 0.9% NaCl) were injected into the pouch. Diclofenac was administered to avoid degradation of AEA by cyclooxygenases. After 1 h of pre-incubation, 1 mL of a 1% zymosan solution was injected into the pouch. After 6 h the pouch content was recovered by injection of 1 mL PBS into the pouch, massage and aspiration of the contained liquids. The concentration used for the final experiments was the result of a dose-escalation study on individual animals to determine an effective concentration.

Human peripheral blood mononuclear cells were isolated from commercially available buffy coats from anonymous donors (DRK-Blutspendedienst Baden-Württemberg-Hessen, Institut für Transfusionsmedizin und Immunhämatologie, Frankfurt, Germany) using Ficoll density centrifugation. Peripheral blood mononuclear cells were washed twice with PBS containing 2 mM EDTA and thereafter incubated for 2 h under growth conditions in RPMI 1640 media supplemented with penicillin (100 U/mL) and streptomycin (100 μg/mL) to enable adherence to culture dishes (Sarstedt, Nümbrecht, Germany). Non-adherent cells were removed. Monocytes were differentiated into naïve macrophages with RPMI 1640 media (Gibco) containing 3% AB-positive human serum (DRK-Blutspendedienst Baden-Württemberg-Hessen, Frankfurt, Germany) for at least 7 days. Differentiated macrophages were exposed to media with 1% FCS overnight. The next day, cells were incubated in RPMI 1640 media with 1% FCS alone, with EtOH as solvent for 2 h, with 10 µM diclofenac (Sigma Aldrich) for 1 h, and treated with 10 µM AEA or EtOH as solvent for 2 h. Diclofenac was used to arrest the cyclooxygenase-mediated breakdown of AEA. After treatment, macrophages were incubated with 10 μg/mL Zymosan (Sigma Aldrich) for 6 h.

Primary human peripheral blood cells were isolated from buffy coats of anonymous donors (DRK-Blutspendedienst Baden-Württemberg-Hessen, Institut für Transfusionsmedizin und Immunhämatologie, Frankfurt am Main). T cells were isolated using the EasySep™ Human T Cell Isolation Kit (Stemcell Technologies) through negative selection. The purity of T cells was greater than 95%, as confirmed by flow cytometry. The cells were cultured at concentration 1 × 10⁶ cells/mL in T-cell medium (RPMI 1640, penicillin (100 U/mL), streptomycin (100 μg/mL), FCS (10%), non-essential and essential amino acids (1%), sodium pyruvate (1%) and 1% 4-(2-hydroxyethyl)-1 piperazineethanesulfonic acid (HEPES)). Cells were supplemented with human recombinant IL-2 (10 ng/mL; PrepoTech) at days 0, 2, and 4 and β-mercaptoethanol (50 μM; Gibco). Cells were cultured for up to 6 days. T cells were treated with Diclofenac for 1 h, Diclofenac and ethanol for 2 h, and Diclofenac and AEA for 2 h. After treatment, T cells were left unstimulated and stimulated with an anti-CD3/CD28/CD2 T cell activator (Stemcell) for up to 6 days. At the endpoint, supernatants were collected for cytokine measurement, and cells were analyzed by flow cytometry.

The migration assay was performed using 6.5 mm diameter transwell cell culture inserts (5 µm pore size; Costar). Human PBMCs were isolated from the buffy coats by Ficoll density centrifugation, washed, counted, and incubated overnight in RPMI 1640 media with 10% FCS, human recombinant IL-2 (10 ng/mL), and β-mercaptoethanol (50 μM). The next day, 10⁵ cells were seeded in the insert in serum-free medium. Prior to the migration assay, PBMCs were treated with AEA and Diclofenac or left untreated. Treated cells were used to migrate toward the medium with FCS. PBMCs that did not receive treatment were used for migration toward macrophage-derived supernatants, medium with AEA (EtOH as control), Diclofenac, and Zymosan (10 μg/mL). Cells were allowed to migrate for 3 h in cell culture. Afterwards, migrated and non-migrated cells were analyzed by flow cytometry. The percentage of migrated immune cells was determined by the ratio of migrated/non-migrated cells.

The concentrations of CCL2, CCL20, CCL4, CCL17, CCL5, IL-8, CXCL1, CXCL10, and CXCL9 in macrophage supernatants were quantified using LegendPlex (BioLegend). Samples were analyzed via flow cytometry. Data were analyzed using FlowJo V.10 (Tree Star).

IFN-γ, IL-10, IL-13, and IL-17A concentrations in the T cell-derived supernatants were quantified using Cytometric Bead Array flex sets (BD Bioscience). Samples were analyzed via flow cytometry. Data were analyzed using FlowJo V.10 (Tree Star).

Single-cell suspensions from air pouches were blocked with FcR blocking reagent (Miltenyi Biotec) in 0.5% PBS-BSA for 10 min, stained with fluorochrome-conjugated antibodies (Table 1) and analyzed on a FACSSymphony A5 flow cytometer (BD Biosciences). Live single cells were identified by FSC/SSC characteristics. Data were analyzed using FlowJo V10 (TreeStar).

T cells and PBMCs were harvested, pelleted by centrifugation, blocked with an FcR blocking reagent (Miltenyi Biotec) in 0.5% PBS-BSA, stained with fluorochrome-conjugated antibodies (Tables 1, 2), and analyzed on a FACSymphony A5 flow cytometer (BD Biosciences). Data were analyzed using FlowJo V.10 (TreeStar).

All antibodies and secondary reagents were titrated to determine optimal concentrations. Comp-Beads (BD) were used for single-color compensation to create multicolor compensation matrices. For gating, fluorescence minus one controls were used. The instrument calibration was controlled daily using Cytometer Setup and Tracking beads (BD Biosciences). To determine the actual number of cells, counting beads were used (Bangs Laboratories).

All experiments were independently performed at least three times as indicated by the number (n) in the respective figure legend. Statistical analysis was performed using Prism 10.1.2. Shapiro-Wilk tests were used to test for normal Gaussian distribution. A paired two-tailed Student’s t-test was used to calculate statistically significant differences between two groups of human immune cells. A one-sample Student’s t-test was used to calculate statistically significant differences between normalized data. ANOVA followed by Tukey’s test or Student’s t-test was used to evaluate statistical significance in murine data. Values of p ≤ 0.05 were considered significant. All data are expressed as mean ± standard error of mean (SEM).

To determine a potential anti-inflammatory effect of AEA in vivo the AirPouch model of acute inflammation was used. AEA nanoformulations or nanoformulation control were injected into the pouch followed by Zymosan stimulation. Subsequently accumulation of inflammatory cells in the pouch was determined. As compared to control, the number of neutrophils and eosinophils was significantly lower in the AEA group. A similar trend existed for monocytes, but did not reach statistical significance. In contrast to this, the number of T cells as well as dendritic cells (DCs) recovered from the pouch was significantly higher in the AEA group as compared to the control group (Figure 1). Thus, AEA treatment reduced myeloid cell infiltration in the pouches, while the number of invaded T cells, which are typically involved in the later immune response, was increased by AEA.

To determine potential mechanisms explaining the findings in the AirPouch model, human primary immune cells were used. First, the impact of AEA on chemokine production by Zymosan-activated primary human macrophages was tested, since local macrophages are the first responders to Zymosan and affect recruitment of further immune cells, among other mechanisms by producing chemokines. Human macrophages were exposed to a medium containing 1% FCS overnight, followed by AEA treatment and activation with Zymosan. Initially, AEA was used at a concentration of 10 nM or 10 µM. However, 10 nM AEA did not have any impact on chemokine levels. When looking at overall chemokine levels, AEA at 10 µM did not significantly change the chemokine production in Zymosan-activated human macrophages (Figure 2; Supplementary Figure S1) due to high variability in chemokine levels between donors. However, when data were normalized, a decreased expression of CCL2 (Figure 2A), CCL4 (Figure 2B), and CCL20 (Figure 2C) in the presence of AEA compared to the Diclofenac and solvent controls (CCL2, CCL4, and CCL20), or only compared to solvent control (CCL20), was noted. Additionally, AEA decreased the production of CCL17 (Figure 2D) compared to Diclofenac and the solvent, even though the changes were minor. The expression of the remaining chemokines was consistent among control and treated groups (Supplementary Figure S1). These data suggest that high concentrations of AEA reduce the expression of chemokines, which participate in attracting cells to the site of inflammation. These findings, while being in line with reduced myeloid cell numbers, did not explain increased T cell numbers in the AirPouch model after AEA treatment. To investigate, whether changes in chemokine production directly alter the migratory capacity of immune cells, a boyden chamber assay was performed in which human PBMCs were allowed to migrate towards a supernatant derived from Zymosan-activated macrophages macrophages that were treated with either solvent controls or AEA. However, leukocyte migration towards supernatants derived from macrophages was not affect, whether they were treated with AEA or left untreated (data not shown). Unexpectedly, also CCR6+ CD4⁺ T cells and CCR6+ CD8⁺ T cells, as responders to CCL20, did not show altered migration towards supernatants from the AEA-treated as compared to control-treated macrophages. In addition, conditioned media of AEA-treated macrophages as compared to control conditions did not alter the migration capability of CCR2+ cells (data not shown). Taken together, AEA reduced the levels of the chemokines CCL20, CCL4, CCL2, and CCL17 in Zymosan-activated macrophages, but this did not directly impact immune cell migration.

To explore a direct effect of AEA on human PBMC migration, the Boyden chamber migration assay was employed to allow PBMCs to migrate towards media containing solvent (CTL), AEA, Zymosan, and AEA + Zymosan. Diclofenac was present in each group to prevent cyclooxygenase-mediated breakdown of AEA. Media with and without FCS were used as positive and negative controls respectively, and the proportion between migrated and non-migrated immune cells was determined by flow cytometry (Supplementary Figure S2). The presence of AEA alone decreased the migration rate of immune cells, while AEA in combination with Zymosan increased the percentage of migrated cells (Figures 3A, B). These data suggest that AEA can have a multifaceted role in immune cell migration. Similarly, the presence of AEA alone in the bottom chamber also decreased the migration rate of B cells (Figure 3C). AEA alone and in combination with Zymosan reduced the migration of NK cells (Figure 3D) and the cytotoxic NK cell subset (CD56dim CD16hi) (Figure 3E). However, there was no difference between groups in the migration rate of the CD56hi CD16dim NK cell subpopulation (Figure 3F). Importantly, the addition of AEA alone or together with Zymosan did not alter the migration rate of T cells per se (Figure 3G). However, there was a tendency for a reduced number of migrated CD4⁺ T cells (Figure 3H) towards AEA in combination with Zymosan, while AEA did not have impact on the migration rate of CD8⁺ T cells (Figure 3I) and CCR6+ CD4⁺ T cells (Figure 3J). When looking at myeloid cells, AEA did not affect the migration of myeloid cells per se (Figure 3K), as well as that of myeloid subsets such as classical monocytes (Figure 3L), intermediate monocytes (Figure 3M), mDCs (Figure 3N), and activated mDCs (Figure 3O). The migration of myeloid CD33low cells (Figure 3P), and CD195 (CCR5)+ myeloid CD33low cells (Figure 3Q), which may correspond to CD33⁺ NK cells, was decreased by AEA without the presence of Zymosan. Together, these data demonstrated that AEA directly affects migration of certain immune cell subsets, despite being applied at the site towards which the cells migrate.

Next, we tested whether or not pre-incubation of PBMCs with AEA would affect their ability to migrate towards FCS. Cells were pre-treated for 1 h with Diclofenac and for 2 h with AEA. Again, media with and without FCS were used as positive and negative controls respectively, and flow cytometry was employed to measure the migration rate of major immune cell populations. AEA dramatically reduced the migration rate of the cells studied including leukocytes (Figures 4A, B), B cells (Figure 4C), NK cells (Figure 4D) and cytotoxic NK cells (Figure 4E), while NK cell precursors treated with AEA demonstrated a tendency to reduced migration rate (Figure 4F). Importantly, pretreatment with AEA reduced migration of T cells (Figure 4G), where particularly the migration of CD8⁺ T cells (Figure 4I) treated with AEA was significantly reduced, while CD4⁺ T cells (Figure 4H) and CCR6+ CD4⁺ T cells (Figure 4J) did not respond to AEA treatment. Similarly, classical monocytes (Figure 4K), activated mDCs (Figure 4L), mDCs (Figure 4N), and myeloid CD33low cells (Figure 4O) treated with AEA showed a tendency towards lower migration rates. There was no difference in migrated cell number in myeloid cells (Figure 4M), CD195+ myeloid CD33low cells (Figure 4P), and intermediate monocytes (Figure 4Q) between groups. Collectively, these data indicate that AEA has a direct impact on immune cells and impairs their migration ability. Moreover, the data suggest that the increase in T cells in the animal model is not a consequence of an increased recruitment by AEA. Rather, AEA might block emigration of cells towards the lymphatics.

Following the findings that AEA altered T cell migration in mouse and human systems, the question whether or not AEA may affect the T cell phenotype at sites of inflammation was studied. Bead-isolated human T cells were treated with AEA or control and were afterward activated with CD3/CD28/CD2 activator cocktail for 5 days. Using flow cytometry, the expression levels of chemokine receptors (CCR6, CCR2, CXCR3, CXCR4), activation markers (CD25, CD38, CD69), and exhaustion markers (TIGIT, LAG3, PD1) were measured (Supplementary Figures S3, S4). Similar to the macrophage experiments, AEA was tested at concentrations 10 nM and 10 µM. Anandamide at lower concentrations did not have any impact on T cell activation, proliferation, and maturation status. However, a marked reduction in T cell cluster formation was observed (Figure 5A) when the cells were treated with 10 µM AEA. In contrast, the percentage of live T cells increased in the presence of AEA (Figure 5B). Both parameters indicate reduced T cell activation. There was no difference in clusters and live cell numbers between groups in non-activated T cells. Addition of AEA decreased percentage of CD4⁺ T cell among non and activated cells (Figure 5C). Surprisingly, a significant reduction in non-activated Th17 cells treated was observed in the AEA treated group (Figure 5D). Additionally, AEA decreased the percentage of activated TIGIT+CD4⁺ T cells (Figure 5E). AEA also reduced the expression of CD38 and CD25, both in activated and non-activated CD4⁺ T cells (Figures 5F, G). No alterations were identified in CCR2+ CD4⁺ T cells, LAG3+ CD4⁺ T cells, Th1 cells, and Th2 cells (Supplementary Figure S5). These data demonstrated that AEA decreases the activation and exhaustion status of CD4⁺ T cells and reduces the number of Th17 cells.

The impact of AEA on CD8⁺ T cells was measured alongside CD4⁺ T cells. By contrast, AEA was found not to alter the percentage of CD8⁺ T cells (Figure 6A). When looking closer at CD8⁺ T cell activation and exhaustion status, PD1+ CD8⁺ T cells (Figure 6B), TIGIT+ CD8⁺ T cells (Figure 6C), CD38⁺ CD8⁺ T cells (Figures 6F, H), and CD69⁺ CD8⁺ T cells (Figures 6G, H) populations were significantly decreased following AEA and activator addition. Similarly, to CD4⁺ T cells, a reduced Tc17 cells percentage in the AEA group in non-activated cells was observed (Figure 6D). Moreover, AEA presence decreased the Tc1 percentage in the non-activated group (Figure 6E). There was no observable effect on CCR2+ CD8⁺ T cells, LAG3+ CD8⁺ T cells, and Tc2 cells (Supplementary Figure S5). Taken together, these data demonstrated that AEA reduces the number of activated and exhausted CD8⁺ T cells and limits CD8⁺ T cell subsets such as Tc1 and Tc17 cells.

Based on Anandamide altering T cell activation surface markers, we further investigated if it would affect T cell cytokine production. To test this, we collected supernatants after T cells were treated with AEA and activated for 6 days. Despite reduced activation observed by flow cytometry, there was no significant difference in interferon-γ (IFN-γ) secretion (Figure 7A). However, we found a dramatic reduction in IL-10, IL-13, and IL-17A cytokine production following AEA and activator addition (Figures 7B–D). These data confirmed that AEA has an impact on T cell activation. Particularly, AEA decreases the production of Th2 and Th17 cytokines, but does not alter the Th1 cytokine IFN-γ. Previous studies showed that AEA at already low concentrations engages CB2 receptors in immune cells altering cell activation and differentiation (Cencioni et al., 2010). Therefore, we measured IL-10, IL-17A, IL-13, and IFN-γ cytokine production in activated T cells treated with AEA at concentrations of 10 nM, 100 nM, and 1 µM (data not shown), which would be sufficient for CB receptor engagement. However, we did not observe any effect in cytokine production when T cell were treated with lower concentrations of AEA. These data suggest a shift in T cell activation profiles rather than a global suppression of T cell activation by AEA. While AEA exhibits its central effects primarily through Gi-coupled cannabinoid receptors CB1 and CB2 receptors which are activated already at nanomolar concentrations (Munro et al., 1993), the lack of an anti-inflammatory effect at such a low concentration suggests alternative receptors with lower sensitivity like intracellular receptors. These require efficient cellular uptake and thus particularly high extracellular concentrations (Teichmann et al., 2024).

Based on our previous study suggesting that NR4A1 and NR4A2 may contribute to the signaling of AEA, and the importance of the NR4A family in T cell biology (Odagiu et al., 2020), the effect of genetic deletion of each receptor or both in combination was determined in the AirPouch model. Deletion of any of the two receptors blocked the effect of AEA. NR4A1 and NR4A2 deficient, as well as double knockout mice, did not exhibit a significant decrease in neutrophil or eosinophil accumulation. Although not statistically significant, it appeared that knockout of each receptor partially attenuated the response of AEA with the double knockout mouse exhibiting the strongest, indicating additive reduction (Figure 8).

Double knockout of NR4A1 and NR4A2 also prevented the AEA-induced increase in T cell and dendritic cells. A similar effect was observed in NR4A1 KO mice. In NR4A2 KO mice, AEA still induced an increase in T cells and dendritic cells, which for T cells did, however, not reach the significance level with the group size present (Figure 8). Collectively, these data suggest that AEA elicits an inhibitor effect on granulocyte accumulation in the AirPouch model by NR4A1 and NRA2, whereas it increases T cell and dendritic cell accumulation through NR4A.