Jurkat T cells were purchased from ATCC and maintained in RPMI 1640 (Gibco) supplemented with 10% heat-inactivated FCS, 100 U/mL penicillin and 100 μg/mL streptomycin. Human monocytes were isolated from commercially available buffy coats from anonymous donors (DRK-Blutspendedienst Baden-Württemberg - Hessen, Institut für Transfusionsmedizin und Immunhämatologie, Frankfurt, Germany) using Ficoll (Biochrom) density centrifugation. Monocytes were isolated from peripheral blood mononuclear cells (PBMCs) using positive selection with CD14 antibody-coupled magnetic beads (MACS Miltenyi Biotec) and LS columns (MACS Miltenyi Biotec) following the manufacturer’s protocol.

CD14⁺ monocytes were differentiated in macrophage serum-free medium (ThermoFisher Scientific) supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin and 50 ng/mL of macrophage colony-stimulating factor (M-CSF) (Immunotools) for 7 days followed by culture in RPMI 1640 medium supplemented with 10% heat-inactivated FCS, 100 U/mL penicillin and 100 μg/mL streptomycin. When indicated, cells were treated with 1 µM of T0070907 (Cayman Chemical), 100 nM GW501516 (Cayman Chemical), 1 µM of Rosiglitazone (Sigma-Aldrich), 10 nM of T0901317 (Tocris Bioscience), 1 µM of GSK3787 (Cayman Chemical), 5 µM of GSK2033 (Sigma-Aldrich), 100 nM of concanamycin (Santa Cruz Biotechnology), 10 µM of isopropyl dodecylphosphonofluoridate (IDFP) (Cayman Chemical), 10 µM of probucol (Cayman Chemical) and 10 ng/mL of LPS (Sigma-Aldrich). Treatments did not compromise Mφ viability.

To induce apoptosis, Jurkat cells were seeded in 10 cm dishes in serum-free RPMI 1640 medium and exposed to 100 mJ/cm² UV-C (254nm) irradiation (UVP Crosslinker CL-1000, Jena Analytik) followed by incubation for 3 hours at 37°C with 5% CO₂. Human Mφ were stimulated with apoptotic Jurkat cells at a 1:3 ratio in 6-well plates for 3 or 6 hours. Upon efferocytosis, non-phagocytosed cells were removed and Mφ were washed 3x with PBS, followed by incubation with RPMI 1640 supplemented with 10% heat-inactivated FCS, 100 U/mL penicillin and 100 μg/mL streptomycin for 3, 6, 9 and 21 hours.

Apoptosis of Jurkat cells was assessed with Annexin V-FITC/propidium Iodide (PI) double staining. Upon UV treatment, Jurkat cells were washed with PBS and resuspended in Annexin V Binding Buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 2.5 mM CaCl₂ in PBS) with 1 µg/mL of Annexin V-FITC (Immunotools) and 1 µg/mL of PI (Thermofisher Scientific). Samples were incubated at room temperature in the dark for 15 minutes followed by flow cytometry analysis with a LSRII/Fortessa flow cytometer (BD Biosciences).

Jurkat cells were labelled using CellTracker™ Orange CMRA (ThermoFisher Scientific) according to the manufacturer’s recommendations. Mφ were seeded onto 8-well chambered coverslips (µ-slide, ibidi GmbH), stained with CFSE Cell Division Tracker Kit (Biolegend) and incubated with labelled apoptotic Jurkat cells at a 1:3 ratio. Upon removal of the non-phagocytosed cells, Mφ were analyzed by fluorescence imaging using a Plan-Apochromat 20x long range objective on a Zeiss LSM800 confocal microscope driven by the Zen 2009 software (Carl Zeiss) or by flow cytometry with a LSRII/Fortessa flow cytometer (BD Biosciences).

Total RNA of non-treated and efferocytotic Mφ (6 biological replicates each) was isolated using RNeasy Micro Kit (Qiagen) according to the manufacturer’s protocol. cDNA library preparation was carried out using QuantSeq 3’ mRNA-Seq Library Prep Kit FWD from Illumina (Lexogen) according to the manufacturer’s procedure. RNA and DNA quantification was done using Qubit cDNA HS Assay Kits (ThermoFisher Scientific) and quality control was performed using an Agilent 2100 Bioanalyzer with RNA Nano Chip (Agilent) as well as High Sensivity DNA chips (Agilent). Libraries were diluted and denatured according to the Illumina Denature and Dilute Libraries Guide, followed by mixing with 1% Phix Control (Illumina). 12 Libraries were loaded on one sequencing cartridge of the TG NextSeq 500/550 High Output Kit v2 (75 cycles) (Illumina) and RNA sequencing was performed on a NextSeq500 system (Illumina).

Statistics of the individual RNA sequence data sets were monitored by FastQC analysis. RNA-seq data processing and differential expression analysis was performed using the QuantSeq data analysis pipeline from Bluebee Genomics analysis platform following manufacturer’s instructions. Genes significantly regulated by efferocytosis were extracted by setting a threshold based on the average number of reads to avoid Jurkat T cell-specific genes (set at the level of CD3). 39.555 genes with lower number of reads were removed. After this selection, a 20.646 gene set was further analyzed using Gene Set Enrichment Analysis 3.0. Gene list of up- and down-regulated genes of efferocytotic versus non-treated MΦ was generated using the following inclusion criteria: |log₂FC|>1 and Padjusted≤0.05, and normalized base mean above 50. The lists were ranked based on adjusted P value. The RNA-seq data are available at the Gene Expression Omnibus database under accession number GSE169160.

Total RNA was isolated with peqGOLD RNAPure reagent (PeqLab Biotechnology) according to manufacturer’s recommendations followed by reverse transcription using Maxima first-strand cDNA synthesis kit (ThermoFisher Scientific). Quantitative real-time PCR (Q-PCR) assays were performed with PowerUp SYBR Green Master Mix (Applied Biosystems) using Quant Studio Real Time PCR System (Applied Biosystems). Relative transcript amounts were quantified using the Δct method with β-microglobulin (βMG) as a housekeeping gene and normalized to the untreated or apoptotic cell-treated controls.

Knockdowns of PPARδ, PPARγ, LIPA and PLA2G15 were performed using transfections with 50 nM siRNA (siGENOME human SMARTpool, Thermo Scientific) and Hyperfect transfection reagent (Qiagen) according to manufacturer’s recommendations. Cells were treated 96 hours post-transfection.

Cell pellets were harvested in lysis buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 5 mM EDTA, 10 mM NaF, 1 mM Na₃VO₄, 0.5% NP-40, 1 mM PMSF, protease inhibitor cocktail). Protein lysates were sonicated and centrifuged at 12.000g for 10 minutes at 4°C. Supernatants were heat-denatured at 95°C, separated on SDS-PAGE gels, and transferred onto nitrocellulose membranes. For ABCA1 and ABCG1 Western blots, the heating step was omitted. Primary antibodies directed against ABCA1 (#NB400105SS, Novus Biologicals), ABCG1 (#13578-1-AP, Proteintech), CPT1A (#15184-1-AP, Proteintech) and nucleolin (#sc-55486, Santa Cruz Biotechnology) were used followed by IRDye 680 or IRDye 800-coupled secondary antibodies (LICOR Biosciences). Blots were visualized using LICOR ODYSSEY scanner, and analyzed by Image Studio™ Lite software (LICOR Biosciences).

Cellular oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) were analyzed using a Seahorse 96 extracellular flux analyzer (Agilent). Human Mφ were plated in Seahorse 96-well cell culture plates one day before measurements and equilibrated in Krebs-Henseleit buffer (111 mM NaCl, 4.7 mM KCl, 1.25 mM CaCl₂, 2 mM MgSO₄, 1.2 mM Na₂HPO₄) supplemented with 11 mM L-glucose and 0.2 mM L-glutamine for 1 hour prior to the assay. Cells were treated with 2.5 µM oligomycin (Sigma-Aldrich), 1 µM carbonyl cyanide m-chlorophenyl hydrazine (CCCP, Sigma-Aldrich), 1 µM rotenone (Sigma-Aldrich), and 1 µg/mL antimycin (Sigma-Aldrich).

Human Mφ were stimulated with apoptotic Jurkat cells at a 1:3 ratio in 6-well plates for 3 hours. Upon efferocytosis, non-phagocytosed cells were removed and Mφ were washed 3 times with PBS, followed by incubation with phenol red-free RPMI 1640 supplemented with 10 µg/mL of human recombinant ApoAI (Calbiochem), 100 U/mL penicillin and 100 μg/mL streptomycin for 21 hours. Cholesterol levels in the supernatant of efferocytotic Mφ were measured using the Amplex Red Cholesterol Assay Kit (Invitrogen) following the manufacturer’s protocol.

Statistical analysis was performed using GraphPad Prism 8.2. Data were analyzed using Student unpaired, two-tailed t test or by one-way ANOVA with Bonferoni multiple comparisons. Comparisons to normalized controls were analyzed using one-sample t test. Graphical data are presented as means ± SEM for at least three independent experiments. Asterisks indicate significant differences between experimental groups (*p<0.05, **p<0.01, ***p<0.005).

PPARs and LXRs are well-established master transcriptional regulators of efferocytotic Mφ, although their functional input to shape the phenotype of human Mφ is only poorly understood. To follow LXR/PPAR activation, we set up an in vitro human efferocytosis model, consisting of monocyte-derived human Mφ and apoptotic Jurkat T cells ( Figure 1A ). Apoptosis of Jurkat cells was induced by UV-C exposure, which caused 75-85% apoptotic cells after 3 hours. To confirm time-dependent efferocytosis, CMRA Orange-labelled AC were added to Mφ for 3 hours, followed by the removal of the non-phagocytosed cells, and subsequent incubations in fresh medium for up to 24 hours. Confocal microscopy showed efficient efferocytosis after 3 and 6 hours when Mφ engulfed substantial amounts of apoptotic material ( Figure 1B ). Quantification of these results by flow cytometry revealed roughly 60% CFSE⁺/CMRA Orange⁺ Mφ after 3 hours ( Figure 1C ). Using microscopy or flow cytometry we noticed a decreasing fluorescence signal intensity originating from AC, suggesting that Mφ digest the apoptotic material over time. We then analyzed the mRNA expression of classical LXR target genes, i.e., ABCA1 and ABCG1 as well as PPARδ/γ targets pyruvate dehydrogenase kinase 4 (PDK4) and carnitine palmitoyltransferase 1A (CPT1A) ( Figure 1D ). Expression of LXR and PPAR mRNA targets peaked at 6-9 hours post-efferocytosis and returned to baseline after 24 hours. Therefore, we choose 6 hours (3 hours exposure to AC, followed by 3 hours after their removal) for subsequent gene expression analyses. Q-PCR analyses of genes specifically expressed in Jurkat cells (CD3E, CD3D, LCK) revealed that the amount of Jurkat mRNA did not exceed 1% of total mRNA in the samples. Furthermore, Jurkat cells showed negligible ABCA1 and PDK4 mRNA expression as compared to macrophages. Therefore, we excluded significant contribution of RNA from apoptotic cells in Mφ. Data so far confirm rapid activation of PPAR and LXR target genes in efferocytotic human Mφ.

To explore global transcriptional responses of efferocytotic human Mφ we performed RNA-seq analysis using 6 biological replicates. Again, Mφ were exposed to AC for 3 hours, followed by AC removal and subsequent incubations for 3 hours. To group differentially expressed, functional subsets of genes, we performed gene set enrichment analysis (GSEA). Efferocytotic Mφ exhibited transcriptional changes related to several metabolic and stress-signaling pathways. We noticed upregulation of oxidative stress responses and hypoxia pathways, as previously reported by us and others (15–17) ( Figure 2A ). Moreover, in addition to upregulating genes referring to PPAR and LXR pathways, GSEA data revealed strong downregulation of SREBP-2 targets associated with cholesterol biosynthesis and sterol metabolism ( Figure 2B ). Conclusively, changes in the activity of PPARs, LXRs and SREBP-2 point to the importance of lipid-related changes in efferocytotic Mφ.

LXR and PPAR transcription factor families consist of two (LXRα and LXRβ), respectively three (PPARα, PPARγ, and PPARδ) members. In human Mφ LXRα is predominantly expressed (18) and linked to efferocytotic responses, while both PPARγ and PPARδ are equally expressed and are known to regulate efferocytosis-induced gene expression in various experimental systems. To understand the relative contribution of PPARδ, PPARγ and LXRs to individual target gene expression during efferocytosis, we inhibited PPARδ and LXRs during the uptake of apoptotic Jurkat cells with the PPARδ antagonist GSK3787 (19) or the LXR antagonist GSK2033 (20) followed by mRNA expression analysis of ABCA1, ABCG1, PDK4, and CPT1A. GSK3787 prevented the induction of PDK4 and CPT1A in efferocytotic Mφ, while expression of ABCA1 and ABCG1 was not affected ( Figure 3A ). In contrast, GSK2033 suppressed ABCA1 and ABCG1 expression but not that of PDK4 or CPT1A ( Figure 3B ). Strikingly, GSK2033 increased PDK4 expression in the absence of AC, suggesting that this inhibitor might activate PPARδ. This appears rational as GSK2033 had been described as a ligand for multiple nuclear receptors (21) and considering that GSK3787 attenuated effects of GSK2033 ( Supplementary Figure 1A ). PDK4 is still induced by AC in the presence of GSK2033 ( Figure 3B ). PDK4 mRNA is also induced by the synthetic PPARγ ligand rosiglitazone ( Supplementary Figure 1B ), but this induction is insensitive to GSK3787, confirming the specificity of the latter one as a PPARδ antagonist. To further explore the contribution of PPARγ and PPARδ to the induction of PDK4, we silenced these transcription factors in efferocytotic Mφ. Consistent with the repressive function of PPARδ in the absence of a ligand (22), knocking down PPARδ substantially upregulated PDK4 expression ( Figure 3C ). AC failed to induce PDK4 RNA in PPARδ-silenced cells, confirming a PPARδ-dependence of PDK4 induction in efferocytotic Mφ. In contrast, silencing PPARγ preserved a PDK4 mRNA increase in response to AC, indicating that PPARγ does not substantially contribute to PDK4 induction upon efferocytosis. Unfortunately, we could not use the pharmacological PPARγ antagonists GW9662 or T0070907, since both of them elicited strong PDK4 induction, probably through PPARδ activation (23) (data not shown). Collectively, our data suggest that AC increase PDK4 and CPT1A mRNA expression through PPARδ, whereas ABCA1 and ABCG1 are induced through LXRs.

Currently, activation of PPARs and LXRs upon efferocytosis is only incompletely understood. We hypothesized that fatty acids and sterols, liberated upon lysosomal digestion of ingested material, might cause their transcriptional activation. To address this possibility, we inhibited lysosomal acidification using the lysosomal v-ATPase inhibitor concanamycin A. Blocking lysosomal acidification interfered not only with induction of LXR targets ABCA1 and ABCG1, but also the PPARδ targets PDK4 and CPT1A ( Figure 4A ). Conclusively, lysosomal processing of apoptotic cells is a pre-requisite for LXR and PPARδ activation.

Recently, Viaud and colleagues (14) suggested lysosomal acid lipase (LIPA) as an essential component to generate LXR ligands in efferocytotic THP-1 cells. To explore the relevance of this pathway for primary human Mφ we silenced LIPA with over 90% efficiency both at the mRNA and protein level ( Figures 4B, C ). In agreement with published observations (14), LIPA silencing prevented induction of the LXR target ABCA1 mRNA by apoptotic cells. Strikingly, a LIPA knockdown also interfered with PPARδ target gene expression in efferocytotic Mφ ( Figure 4B ). Thus, LIPA is critical for processing ingested material to generate both LXR and PPARδ ligands. Also, the hydrolysis of ingested phospholipids by lysosomal phospholipase A₂ (PLA2G15) may generate PPAR activating ligands. To inhibit PLA2G15, we used isopropyl dodecylfluorophosphonate (IDFP), a covalent modifier of the catalytic serine-165 residue (24). As seen in Figure 4D , IDFP lowered PDK4 and CPT1A expression in efferocytotic MΦ, with no regulation of ABCA1 or ABCG1. To verify the role of PLA2G15, we used a knockdown strategy and analyzed mRNA expression of PPARδ targets in efferocytotic Mφ ( Figure 4E ). Silencing PLA2G15 in Mφ indeed attenuated PDK4 and CPT1A expression.

Previous studies suggested that LXRs and PPARs play important roles in controlling the expression of efferocytotic receptors, cholesterol efflux, and suppressing inflammatory responses in efferocytotic Mφ (8, 25). However, most of these studies employed Mφ with a constitutive knockout of these transcription factors, with an altered expression of efferocytotic receptors or cholesterol transporters. We took advantage of acute pharmacological inhibition of LXRs and PPARδ during efferocytosis to examine how LXR and PPARδ activation in efferocytotic macrophages affects cholesterol efflux, mitochondrial metabolism, and inflammatory response.

Surprisingly, and contrary to mRNA expression data, we noticed that protein expression of ABCA1 ( Figure 5A ) was not significantly altered during efferocytosis. Analysis of Jurkat-specific protein Zap70 showed no contamination of the samples with proteins from engulfed cells. Nevertheless, when we analyzed cholesterol efflux to apoAI in the supernatant of Mφ 24 hours post-efferocytosis ( Figure 5B ), we observed an increased cholesterol accumulation in apoAI-containing medium. This increase is sensitive to the co-treatment with GSK2033 or the ABCA1 inhibitor probucol. Apparently, pharmacological LXR inhibition acutely suppresses ABCA1-mediated cholesterol efflux from efferocytes.

PDK4 and CPT1A support mitochondrial fatty acid oxidation through inhibition of the mitochondrial pyruvate dehydrogenase complex (PDH) and maintaining fatty acid transport into mitochondria, respectively. While we were unable to find commercial antibodies specifically detecting PDK4 in human Mφ, we did not observe changes in PDH phosphorylation upon efferocytosis ( Supplementary Figure 2A ). Similarly, CPT1A protein remained unaltered in efferocytotic Mφ ( Supplementary Figure 2A ). Finally, analysis of mitochondrial oxygen consumption and extracellular acidification rates using Seahorse extracellular flux analysis showed no alterations in efferocytotic Mφ, suggesting that efferocytosis doesn’t significantly affect mitochondrial metabolism in human Mφ ( Supplementary Figure 2B ).

To assess the importance of LXR and PPARδ in modulating the inflammatory profile of efferocytotic Mφ, we acutely inhibited these transcription factors using GSK2033 and GSK3787, respectively, and simultaneously treated Mφ with lipopolysaccharide (LPS) and apoptotic cells ( Figure 5C ). LPS stimulated the mRNA expression of inflammatory chemokines, i.e., chemokine (C-X-C motif) ligand 9 and 10 (CXCL9 and CXCL10) as well as the pro-inflammatory interleukin 12A (IL12A). In efferocytotic Mφ expression of CXCL9/10 and IL12A mRNA was attenuated. Whereas PPARδ inhibition by GSK3787 attenuated the expression of these targets in LPS-stimulated Mφ, apoptotic cells still exhibited anti-inflammatory effects in the presence of GSK2033 or GSK3787. We did not observe significant alterations of anti-inflammatory IL10 in efferocytotic Mφ under these conditions ( Supplementary Figure 2C ). In summary, these observations indicate that apoptotic cells attenuate LPS-induced inflammatory gene expression probably independently of LXR or PPARδ agonsim.