Resveratrol and ω-3 PUFAs Promote Human Macrophage Differentiation and Function

Monocytes differentiate into M1 and M2 macrophages, which are classically activated by microbial products such as LPS or IFN-γ and interleukins (e.g., the anti-inflammatory and Th2 promoting IL-4), respectively. The contribution of nutrients or nutrient-based substances such as ω-3 polyunsaturated fatty acids (ω-3 PUFAs) and resveratrol (Res) on the differentiation and function of M1 and M2 macrophages was evaluated. THP-1 cells and peripheral blood mononuclear cells (PBMCs) were differentiated into M1 and M2 cells and activated with LPS/IFN-γ or IL-4/IL-13. Macrophage lineage specific surface determinants (e.g., CD11b, CD11c, CD14, CD206, CD209, CD274, HLA-DR, CCR7, CCR2) were analysed by cytofluorometry. Res and ω-3 PUFAs altered CD14, CD206, CD274 and HL-DR surface expression patterns in M1 and M2 macrophages differentiated from PBMC. LPS/IFN-γ or IL-14/IL-13 activated macrophages subpopulations, which secreted cytokines and chemokines as measured by multiplex ELISA. Res and ω-3 PUFA reduced IL-1β, IL-6, TNF-α, CXCL10/IP-10, CCL13/MCP-4 and CCL20/MIP-3α in LPS/IFN-γ activated human leukaemia THP-1 cells, which is indicative of a dampening effect on M1 macrophages. However, Res increased M1 prototypic cytokines such as IL-1β or IL-6 in macrophages derived from PBMCs and also modified the expression of IL-12p70. Collectively, Res and ω-3 PUFAs distinctly promoted the differentiation and function of M1 and M2 macrophages. We conclude that these substances strengthen the macrophage-mediated effects on the innate and adaptive immune response.


Surface Determinant
Macrophage Subtype Colony stimulating factors mainly determine the fate of macrophage differentiation in conjunction with metabolic factors. In addition, nutrients and nutritional substances modulate cells of the adaptive and innate immune system including macrophages. Their anti-inflammatory and immunomodulatory effects could therefore contribute to skewing the balance between the M1/M2 phenotypes and functions. For instance, ω-3 polyunsaturated fatty acids (ω-3 PUFAs) have been shown to alter the functions of macrophages [28][29][30][31][32] and liver macrophages, i.e., Kupffer cells [33]. EPA and DHA reduced expression of pro-inflammatory cytokines and chemokines [30,31]. DHA favoured the differentiation of M2 phenotypes [34,35] and of RAW264.7 cells, a murine macrophage cell line [36][37][38][39]. Similarly, ω-3 PUFAs are associated with a shift in M1/M2 macrophage transition of Kupffer cells [33]. We and others have shown the anti-inflammatory effects of resveratrol (Res) on blood cells, macrophage/monocyte and endothelial cell lines after induction of the inflammatory response [40,41]. Res also modulated the M1/M2 balance in Kupffer cells [42].
The present study aimed evaluating the effects of nutrients on macrophage phenotypes and functions. For this purpose, peripheral blood mononuclear cells (PBMC) were differentiated into M1 and M2 cells and incubated with specific activators of M1 and M2. Similarly, the human leukaemia monocytic cell line THP-1, was treated with M1 and M2 activators. Phenotypes and functions of different cell populations were determined and related to M1 or M2 prototypic parameters. The results showed that Res and ω-3 PUFAs influenced the phenotypes and functions of M1 and M2 macrophages.

Isolation and Culture of Monocytes from Human Peripheral Blood Mononuclear Cells (PBMC)
Human peripheral blood leukocytes (PBL) were isolated from freshly drawn blood from healthy individuals as described previously [43]. Briefly, PBMC were isolated from total blood by Ficoll-Isopaque centrifugation, washed twice in an excess of PBS/EDTA (2 mM) to remove platelets and resuspended in culture medium. The purity of PBMC was analysed using the FSC/SSC cytofluorometric plot and was >80%. The PBMC population contained >25% of monocytes as determined by FSC/SSC analysis. PBMC were seeded in 12-well culture plates for 4 h at 37 • /5% CO 2 in humidified atmosphere.  [7,18,44]. Where appropriate, substances were added to culture medium at the beginning of the polarization period. Macrophages were harvested after 6 days and processed for cytofluorometric analysis. Alternatively, culture media were renewed, and the cells activated with LPS (100 ng/mL)/IFN-γ (20 ng/mL) or interleukins (i.e., IL-4 and IL-13, 20 ng/mL of each interleukin). After 24 h, culture supernatants were recovered for chemokine and cytokine analysis. Cultured cells were detached and analysed by cytofluorometry. DNA contents and therefore the relative cell numbers obtained at different culture conditions were determined using the DNA intercalating dye (CyQUANT™)) according to the manufacturer's instruction (Molecular Probes, www.thermofisher.com (accessed on 29 April 2022)).

THP-1 Cells
THP-1 cells, a human leukaemia monocytic cell line, were from Cell Lines Service (Eppelheim, Germany) and maintained at <2 × 10 5 cells/mL in RPMI 1640 medium supplemented with P/S, 10% FCS and 2 mM L-glutamine. Cell adherence was induced by treatment with 320 nM phorbol 12-myristate 13-acetate (PMA). Substances were added 4-6 h after beginning of PMA treatment. Cells were then activated with LPS (100 ng/mL)/IFNγ (20 ng/mL) and IL-4/IL-13 (20 ng/mL of each interleukin), respectively. Cell culture supernatants were recovered after 24 h and processed for chemokine and cytokine analysis. Where appropriate, THP-1 cells were detached, and the expression of surface determinant was measured by cytofluorometry. Cell numbers were determined with the CyQUANT™ DNA dye kit (Molecular Probes, www.thermofisher.com (accessed on 29 April 2022)).

Analysis of Cytokines, Chemokines and PGE 2
Secreted analytes (cytokines, interleukins, chemokines, PGE 2 ) were measured using multiparametric kits purchased from Bio-Rad Laboratories (Hercules, CA, USA) and EIA PGE 2 assay kits (Cayman Chemicals, Ann Arbor, MI, USA). Data were acquired using the LiquiChip Workstation IS 200 (Qiagen, Hilden, Germany) and Spectramax (Molecular Devices, San Jose, CA, USA), respectively. Data evaluation was made with the LiquiChip Analyser software provided by Qiagen [45]. All data were normalized to the number of cells determined by Cyquant at the end of the experimental series.

Statistical Analysis
Data were evaluated using statistical tools described previously [46]. A p value < 0.05 (calculated with SPSS software from IBM SPSS v22, Dynelytics AG, Zürich, Switzerland by using Student's t test or one-way ANOVA) was considered to reflect statistically significant differences.

Dosage Information
The final concentrations of substances in cell cultures were Res at 25 µM and omega-3 PUFAs at 20 µM. Unmodified Res had a plasma half-life of several hours [47], while omega-3 PUFAs persisted in plasma for days and weeks [48]. Dietary supplementation with Res and ω-3 PUFAs resulted in plasma concentrations of 2 µM [47] and 30 µM [48], respectively. Yet, transient cellular concentrations of substances might be considerably higher as shown for ascorbic acid (Schwager et al., 2015 [46] and references therein).
We observed subtle but significant changes in surface determinant expression when PMA-differentiated THP-1 cells were activated in the presence of Res or ω-3 PUFAs. These substances induced~2 to~100 fold variations in mean fluorescence levels. In LPS/IFN-γtreated (i.e., M1) THP-1 cells, the substances had minor effects on CD11b surface expression, which was increased by Res and ω-3 PUFAs (Figure 2). Res and ω-3 PUFAs reduced CD14 surface expression, whereas CD11c was enhanced ( Figure 2, panels A-E). Both Res and ω-3 PUFAs decreased CD14 and HLA-DR surface deposition. These investigations were extended to IL-4/IL-13-treated (i.e., M2) THP-1 cells. ω-3 PUFAs lessened levels of CD11b and CD11c ( Figure 2F,G), but it had no significant impact on CD14 (not shown). While ω-3 PUFAs had only a minor effect on CD274 surface expression, Res augmented it, whereas it blunted CD209 high expression ( Figure 2, panels F-I). Collectively, these substances modulated surface markers on THP-1 cells towards M2 phenotypes (revealed also by a reduced M1 surface marker expression).

CD11b
CD11c CD14 CD80 HLA-DR CD274 CD209   Cells were also cultured in M1 and M2 differentiation medium, in order to measure possible effects of GM-CSF and M-CSF on THP-1 cells. No significant phenotypic changes were observed within 6 days of culture (data not shown).

Res and Omega-3 PUFAs Modulate M1 and M2 Differentiation of Macrophages Derived from Peripheral Blood Mononuclear Cells (PBMCs)
Adherent cells, isolated from PBMCs, were cultured for 6 days with GM-CSF or M-CSF to induce M1 and M2 macrophages, respectively. They are also termed polarized M1 (M1p) and polarized M2 (M2p) macrophages [17,49,50]. At these culture conditions, M1 and M2 microphage differentiation is mirrored by the expression of distinct surface markers ( Figure 3, Table 1). Cell-specificity of antibodies was determined by staining freshly isolated PBMCs ( Figure S1). In vitro differentiated macrophages up-regulated CD14 and CD209, where the staining pattern indicated the appearance of intermediate-positive and brightpositive subpopulations (Figure 3). We compared surface expression on M0 macrophages (no CSF added to culture medium) to those of M1-or M2-differentiated macrophages. M1 macrophages displayed increased levels of CD11b and CD274, while the density of CD14 and CD209 was decreased ( Figure 3, Table 1). M0 and M2-differentiated macrophages had similar surface density of CD11b, CD209 and HLA-DR ( Figure 3). Collectively, the differences in expression of CD11b, CD14, and CCR7 (not shown) were diagnostic for M1 and M2 macrophage differentiation patterns (see also [7,18,44]. In addition, LPS/INF-γ activation on M1 and M2 macrophages caused down-regulation of CD14, CD11b, and CD11c (not shown).   Substances induced MFI intensity differences of ≈2-fold (e.g., CD274 (D)) to >100-fold (e.g., CD14 (B)). Staining with isotype control antibodies: light grey. MFI values are indicated for the major peak of the colour-coded cytofluorometric curves (see also Supplementary Table S1). Only surface determinants where marked differences between cultured conditions were observed are shown.
Next, cells were cultured in macrophage differentiation media containing Res or ω-3 PUFA. Cells, which differentiated in the presence of Res, formed aggregates (not shown).  In order to investigate whether the substances induced macrophage polarization, we cultured adherent PBMC during 7 days with Res or ω-3 PUFAs in M0 differentiation medium. The substances per se did not induce a distinct macrophage phenotype (not shown).
Subsequently, we investigated the effects of substances on the functional properties of activate THP-1 cells. Res and ω-3 PUFAs mitigated the production of TNF-α, IL-1β, IL-6, CXCL10/IP-10, but also CCL13/MCP-4 and CCL20/MIP-3α. Res reduced cytokine secretion more potently than ω-3 PUFAs (Table 3, Figure 5). THP-1 cells activated with IL-4/IL-13 secreted similar amounts of cytokines and chemokines as unstimulated cells (except for CCL18/PARC), which indicates a low responsiveness of THP-1 cells to M2 activation signals. CCL18/PARC production was low at the cultured conditions used and only marginally altered by Res. Thus, Res and ω-3 PUFAs reduced the prototype M1 cytokines in THP-1 cells. Due to low basal expression after IL-4/IL-13 stimulation their effects on M2 polarized THP-1 cells could not be determined ( Figure 5).

Macrophages
The secretion of prototype chemokines and cytokines was determined in different macrophage populations derived from PBMCs (Table 1). During a 7 days differentiation period, secreted chemokines and cytokines were similar at various culture conditions (Table 2, Figure 6). For instance, TNF-α and IL-6 levels were ≈900 pg/mL and ≈25 ng/mL at all conditions (Table 2). In addition, the M2 prototype chemokines CCL18/PARC and CXCL1/GRO-α were similar. This implies that during the in vitro polarization, PBMCsderived M1 and M2 macrophages released similar amounts of M1 and M2 prototype chemokines and cytokines. This contrasts with different phenotypes observed at various differentiation procedures (Figure 3). Consequently, the pattern of secreted cytokines and chemokines did not fully match the phenotypes of in vitro polarized un-activated macrophages.

Res and ω-3 PUFAs Modulated the Secretion Pattern of PBMC-Derived M1 and M2 Macrophages
The secretion of prototype chemokines and cytokines was determined in different macrophage populations derived from PBMCs (Table 1). During a 7 days differentiation period, secreted chemokines and cytokines were similar at various culture conditions (Table 2, Figure 6). For instance, TNF-α and IL-6 levels were ≈900 pg/mL and ≈25 ng/mL at all conditions (Table 2). In addition, the M2 prototype chemokines CCL18/PARC and CXCL1/GRO-α were similar. This implies that during the in vitro polarization, PBMCs-derived M1 and M2 macrophages released similar amounts of M1 and M2 prototype chemokines and cytokines. This contrasts with different phenotypes observed at various differentiation procedures (Figure 3). Consequently, the pattern of secreted cytokines and chemokines did not fully match the phenotypes of in vitro polarized un-activated macrophages.
M1p macrophages responded to LPS/IFN-γ stimulation since levels of TNF-α were higher than in the un-stimulated counterparts ( Figure 6). Similar observations were made for CXCL10/IP-10. M1p macrophages responded to IL-4/IL-13 activation with an excess of CCL13/MCP-4 and CCL18/PARC, which were significantly higher than in M1p medium-treated cells. M2p macrophages responded to IL-4/IL-13 activation with an increased secretion of CCL13/MCP4 and CCL18/PARC, whereas TNF-α, IL-6, CXCL10/IP10 and CCL20/MIP3α were secreted at lower amounts or barely matched the levels of unstimulated cells. Yet, M2p macrophages still retained full responsiveness to LPS/IFN-γ, since they produced most of the metabolites at similar levels as LPS/INF-γ activated M1p macrophages Consequently, we observed only a quantitative functional difference between M0p, M1p and M2p macrophages. The LPS/IFN-γ to IL-4/IL-13 ratio, however, indicated that TNF-α or CXCL10/IP10 and CCL3/MCP-4 were functional markers for M1p and M2p macrophages respectively (Table 4). In contrast, PGE 2 production was a specific read-out for M1p activated macrophages: LPS/IFN-γ induced significant levels of PGE 2 only in M1p rather than in M0p and M2p cells ( Figure S2). M1p macrophages responded to LPS/IFN-γ stimulation since levels of TNF-α were higher than in the un-stimulated counterparts ( Figure 6). Similar observations were made for CXCL10/IP-10. M1p macrophages responded to IL-4/IL-13 activation with an excess of CCL13/MCP-4 and CCL18/PARC, which were significantly higher than in M1p mediumtreated cells. M2p macrophages responded to IL-4/IL-13 activation with an increased secretion of CCL13/MCP4 and CCL18/PARC, whereas TNF-α, IL-6, CXCL10/IP10 and CCL20/MIP3α were secreted at lower amounts or barely matched the levels of un-   Next, we investigated the impact of substances on the function of PBMCs-derived M1 and M2 macrophages. Both Res and ω-3 PUFAs modulated the expression of metabolites ( Figure 6). ω-3 PUFAs only marginally influenced the cytokines and chemokines produced by PMBC derived M1 or M2 macrophages. Res enhanced the secretion of IL-1β, TNF-α and CCL20/MIP-3α, but it blunted CCL18/PARC production. We studied the impact of substances on the metabolites, that were secreted by activated M1p macrophages ( Figure 6). The production of TNF-α and CXCL10/IP-10 was impeded by Res in M1p macrophages regardless of the activation signals (i.e., LPS/IFN-γ or IL-4/IL-13), whereas IL-1β production was increased in M1p macrophages. IL-6 was moderately blunted by ω-3 PUFAs in M1p and M2p populations. IL-12p70 was enhanced by ω-3 PUFAs and impeded by Res, whereas both substances had no significant impact on the expression of IL-10. We also observed significant effects of Res on IL-4/IL-13 activated M1 and M2 macrophages: the production of CXCL10/IP-10, CCL20/MIP-3α and CXCL/GRO-α was increased, while the M2 prototype chemokines CCL13/MCP-4 and CCL18/PARC were secreted at a lower amount by Res-treated cells.

Discussion
In the current study we evaluated whether nutritional substances contributed to the development and differentiation of macrophage populations and functions. In vitro polarization and activation procedures to induce M1 and M2 macrophages were applied to the human leukaemia monocytic cell line, THP1, and to cells, that differentiated from adherent human PBMCs. THP-1 cells might be considered to represent an M1 predifferentiated macrophage line and therefore less inducible to differentiate into both M1 and M2 macrophages [49]. Accordingly, THP-1 cells did not further differentiate in the presence of colony stimulating factors including GM-CSF or M-CSF. However, they responded to M1 activation signals (i.e., LPS and INF-γ) and less to M2 activation signals (i.e., IL-4/IL-13). This suggests that THP-1 cells were biased to the M1 phenotype. Alternatively, the differentiation into M2 macrophages requires more complex interactions between chemokines and cytokines, which were not met in the THP-1 culture conditions procedure ( [49]; this study). It should be noted that cell lines such as THP-1 cells were valid surrogates but not perfect counterparts of the PBMC derived macrophage populations [49].
Macrophages derived in vitro from adherent PBMCs differentiated into various macrophage populations. In two seminal studies, the features of M1 and M2 macrophage populations were identified at the level of gene transcription, phenotypes and functions [17,18]. Here, we confirmed the pertinence of the respective parameters. Conditions for in vitro differentiation of all macrophage populations may be difficult to establish [9,10,25]. Yet, this and other studies have shown that distinct experimental conditions permit to induce the expression of surface determinants and cytokines/chemokines that were diagnostic for M1 and M2 macrophages. In fact, the induction of M1 or M2 macrophages was reflected by changes in macrophage-specific surface determinants (Table 1). With regard to functional parameters, differences in the secretion of, e.g., TNF-α and IL-1β by M1 and M2 macrophages were not spectacular, but marked for CXCl10, CCL5, and the (surface determinant) chemokine receptor CCR7. The transcription of cytokines and interleukins such as TNF-α, IL-1β, and IL-6 is regulated by the NF-κB pathway. Res and other anti-oxidants effected directly the expression of NF-κB factors ( [40,51,52]; reviewed in [53]). Consequently, Res could regulate numerous prototype M1 macrophage genes (compared to M2 prototype genes). Yet, the activity of Res seems to be rather complex and cell-specific: Res enhanced the IL-6 expression in PBMCs, whereas it strongly impaired IL-6 expression in murine RAW264.7 macrophages [40,43,54]. This corroborates the overall anti-inflammatory effect of Res. We hypothesize that it might transiently increase secretion of pro-inflammatory cytokines in order to promote rapid resolution of the inflammatory response. It should be mentioned that Res increased CD206 surface deposition in M1-differentiated cells and therefore the shift to the M2 phenotype ( Figure 4). This is in line with other studies where Res modulated CD206 expression [55] and acted on the M1/M2 macrophage balance at inflammatory conditions [56].
ω-3 PUFAs are membrane constituents and contribute to increased generation of EPA and DHA-derived resolvins but also prostaglandins (reviewed in [57]). Similar to Res, they further exemplify the dual role of nutrients as anti-inflammatory and pro-inflammatory agents [46]. DHA favoured M2 polarization and promoted the M1 to M2 macrophage switch [58,59].
Collectively, both Res and ω-3 PUFAs modulate the phenotype and the functions of M1 and M2 macrophages, which each substance having its particular and idiosyncratic activity profile. Importantly, when combined they two nutritional substances might have synergistic or antagonistic effects on macrophage phenotypes and functions. The in vivo significance of these results remains to be demonstrated by further investigations. It should be noticed that after absorption Res is rapidly metabolized and has only a short plasma half-life [47], yet the kinetics of ω-3 PUFAs indicate that they circulate in the plasma for weeks [48]. Res is supposed to have potent but transient and short-range effects on macrophage function, while ω-3 PUFAs have long-lasting effects. Both substances might have context-dependent but also macrophage population-specific effects [40]. We hypothesize that these distinct physiological profiles offer an advantage in targeting the effects of the two nutritional substances in transient (for Res) and prolonged or chronical (for omega-3 PUFAs) dietary supplementation.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/biomedicines10071524/s1, Figure S1: Cytofluoremetric analysis of freshly isolated adherent PBMC for surface determinants.; Figure S2: Production of PGE 2 by in vitro differentiated macrophages from PBMC; Table S1: Peak fluorescence intensity (PFI) data related to Figure 3; Table S2: Peak fluorescence intensity (PFI) data related to