The THP-1 cell line was purchased from American Type Culture Collection (ATCC). The cells were cultured in complete medium consisting of RPMI 1640 Medium containing sodium bicarbonate (Sigma-Aldrich) supplemented with 10% FBS (Sigma-Aldrich), 200 U/mL penicillin, 200 µg/mL streptomycin and GlutaMAX Supplement (Life Technologies) and incubated at 37 °C with 5% CO₂. Cells were tested for mycoplasma using the MycoAlert mycoplasma detection kit (Lonza). The THP-1 cells were differentiated into M0, M1 or M2 macrophages using a standard protocol with minor modifications (27). Briefly, 1x10⁶ THP-1 cells were seeded into 24 well plates and treated with 5 ng/mL phorbol-12-myristate-13-acetate (PMA, Sigma-Aldrich) for 24 hours. PMA was then replaced with fresh complete medium and the cells incubated for 72 hours, followed by stimulation with either 250 ng/mL lipopolysaccharide (LPS, Sigma-Aldrich) and 20 ng/mL interferon-γ (IFN-γ, Miltenyi Biotec), or 20 ng/mL interleukin-4 (IL-4, Miltenyi Biotec) for 48 hours, to differentiate them into M1 or M2 macrophages respectively. Perhexiline (PHX, Sigma-Aldrich) was added at different stages of the culture protocol i.e., during addition of PMA (priming), during resting stage for 72 hours (rest), or for 48 hours along with addition of cytokines and LPS (differentiation). For the IL-1β release experiments, cell-free supernatant was collected from M1 macrophages after differentiation. Then, the macrophages were washed twice with phosphate buffered saline (PBS). Finally, IL-1β release was further stimulated by treating with complete medium supplemented with 1 μg/mL LPS for 2 hours followed by the addition of 5 mM adenosine triphosphate (ATP, Sigma-Aldrich) for 1 hour, before the cell-free supernatant was collected. For unstimulated controls, the medium was complete medium without LPS and ATP.

Macrophages were stained with BD Horizon Fixable Viability Stain 780 (FVS780) (Biolegend). Cells were then treated with 50 μL of FC block (BD Biosciences), then stained using anti-human CD80 A647, anti-human CD209 BV421, anti-human CD14 FITC and anti-human B2M PE conjugated antibodies (Biolegend), prepared in FACS buffer, for 30 minutes at 4 °C. After washing, the cells were resuspended in FACS buffer and analysed using a FACS Canto II flow cytometry system (BD Biosciences). The data were analysed using FlowJo v10.4.1 software (BD Biosciences).

Cells were washed in PBS and RNA was extracted using PureLink RNA Mini Kit (Life Technologies). RNA concentrations were measured and purity was checked using a Nanodrop-1000 spectrophotometer (Thermo Fisher Scientific). Total RNA (100 ng) was reverse transcribed using the iScript cDNA Synthesis Kit (Bio-Rad). qPCR was performed using a ViiA 7 Real-Time PCR System (Life Technologies) using TaqMan Fast Advanced Master Mix (Life Technologies). Primers (Life Technologies) were purchased for the following genes: STAT1 (Hs01013996_m1), STAT6 (Hs00598625_m1), CCL2 (Hs00234140_m1), c-MYC (Hs00153408_m1), and RPML37 (Hs01102345_m1) as an internal control. All qPCR analyses were performed using the ΔCT method in Design and Analysis Quant studio Software (Thermo Fisher Scientific).

IL-1β from cell-free supernatants were quantified using ELISA MAX™ Deluxe Set Human IL-1β kit (Biolegend), following the manufacturer’s instruction. Absorbance was measured at 450 nm using a FLUOstar Optima microplate reader (BMG Labtech).

Vehicle and 5 µM PHX treated M1-like macrophages were lysed in 50 mM Tris (pH 8.5) buffer containing 1X cocktail of protease inhibitors (Thermo Fisher Scientific) using manual homogenisation. Cells were centrifuged at 42,000 × g for 5 min at 4 °C. The protein content of the supernatant was estimated using a NanoOrange Protein Quantitation Kit (Thermo Fisher Scientific). Protein and peptide single-pot solid-phase-enhanced sample preparation (SP3) with Sera-Mag SpeedBead Carboxylate-Modified Magnetic Particles (Cytiva) was performed on cell lysates adopting manufacturer’s instructions. Briefly, proteins were reduced with 10 mM tris (2-carboxyethyl) phosphine (TCEP) for 30 min at 56 °C, then alkylated with 20 mM chloroacetamide in the dark for 30 min at room temperature. Hydrophobic and hydrophilic beads were resuspended 1:1 in ultra-pure water at 10 μg/μL. A 10:1 bead /protein ratio was added to lysates. Ethanol (100%) was added at 1:1 volume to the bead/protein mix and incubated for 10 min at 1000 rpm using a ThermoMixer (Eppendorf). Samples were then placed on a magnetic rack for 2 min. Supernatants were aspirated and carefully washed 3x in 80% ethanol with each wash having a 1 min magnetic separation step. Beads were then air-dried for 30 s before being resuspended in 100 mM ammonium bicarbonate. Trypsin was added in a 1:20 enzyme-to substrate ratio to each sample and incubated at 37°C overnight to digest the protein. A follow-up peptide clean-up was then performed. Following a 2 min magnetic separation step, supernatants containing peptides were transferred into a clean tube before being bound to freshly prepared Sera-Mag SpeedBeads. Acetonitrile (100%) was then immediately added to reach a final concentration ≥ 95% to initiate peptide binding to beads. Samples were then incubated for 10 min at 1000 rpm using a Thermomixer and then washed three times in acetonitrile with magnetic separation in each step. Beads were then air-dried before peptides were eluted with 2% DMSO. The concentration of the peptides was brought to 1 μg/3 μL prior to mass spectrometry acquisition. For reproducibility and to build a chromatogram library, aliquots from five replicates were pooled to ensure that the pool contained all peptides.

Peptides were analysed with a Dionex Ultimate 3000 UPLC coupled with a Orbitrap Exploris 480 tandem mass spectrometer (Thermo Fisher Scientific). An in-house pulled column created from 75 µm inner diameter fused silica capillary packed with 1.9 µm ReproSil-Pur C18 beads (Dr. Maisch, Ammerbuch, Germany) to 25 cm, coupled with a PepMap™ 100 trap cartridge (0.3 x 5 mm, 5 µm C18, Thermo Fisher Scientific) were used. Solvent A was 0.1% formic acid in water and solvent B was 0.1% formic acid in 80% acetonitrile. For each injection, 1 µg of peptides was loaded and separated using a 120 min gradient from 3 to 31.2% solvent B, followed by a 30 min washing and equilibration gradient.

A pooled sample comprised of 1.5 µL of each protein digest was used to generate a sample project specific spectral library for data dependent analysis (DDA). Six gas phase fractionation (GPF) chromatogram library acquisitions, each spanning a narrow m/z range across the 350 – 1200 m/z total mass range (350-500 m/z method 1, 490-610 m/z method 2, 600-710 m/z method 3, 700-810 m/z method 4, 800-910 m/z method 5, 900-1200 m/z method 6). For each DDA-GPF analysis two µL of the pooled sample was used with a three second cycle time instrument method. Briefly, a narrow spectrum ms1 scan matching one of the six m/z mass ranges was performed using an orbitrap resolution of 60,000. A normalised AGC target of 3e6 with an auto maximum injection time mode was used. An intensity threshold of 2.5e5 and dynamic exclusion of time of 45 s was employed for all data dependent ms2 scans that were acquired at 15,000 resolution, AGC target 5e4, 33% normalised collision energy (NCE) in the HCD cell, with an auto maximum inject time mode used.

For the Data Independent Acquisition (DIA) runs, the Orbitrap Exploris 480 was configured to acquire 37 16 m/z precursor isolation windows (396.43 –1004.70 m/z), followed by 37 16 m/z windows (400.43 –1008.70 m/z) creating a staggered window pattern. An ms2 resolution of 15,000, AGC target 5e4, maximum inject time of 20 ms, and normalised HDC collision energy of 28 was employed for all DIA scans. Precursor spectra over a 390 - 1010 m/z mass range were acquired prior to DIA scans with a resolution of 60,000, AGC target 3e5, maximum inject time 100 ms were used for all full scan mass spectrometry (MS) spectra.

Spectronaut™ (version 15.0.210615.50606, Biognosis) was used for both spectral library generation and DIA data analysis. Factory default setting were used for all analysis steps.

Differential protein abundance analysis between the control and PHX treated groups was conducted in R using the DEqMS package (28). The number of precursors used for quantification counts and log2 abundance scores for each protein and sample were used in the analysis. A protein was determined to be present in a sample if it had at least two number of precursors, in order to guard against false positives (29). Proteins were included in the differential expression analysis if they were present in at least two samples in each group. All p-values were corrected using a false discovery rate (FDR) of 0.05.

Differentially expressed proteins were then submitted to the STRING database (30) to generate protein-protein interactions, and clusters from downregulated proteins were identified using inbuilt clustering tools with a k-means cluster of three. Gene set enrichment analysis (GSEA) of individual clusters was undertaken using the Reactome Knowledgebase (31).

Ranked GSEA was performed with the clusterProfiler package (32). Gene sets identified as significant (FDR < 0.2, p < 0.05) with GSEA were visualized using the enrichplot package (33). The following ontologies were included in the analysis of the Molecular Signatures Database (MSigDB) (34, 35): Hallmark gene set collection (36), Gene Ontology (GO) (37, 38): Biological Process, GO: Molecular Function, GO: Cellular Component, KEGG Pathway (39) and Reactome Knowledgebase (31).

Technical and biological replicates were included in all experiments. Statistical analysis is described in figure legends.

To generate macrophage subsets, we adopted a standardised protocol that reliably generates THP-1 derived M1 and M2 macrophages with molecular features that mimic human monocyte derived macrophages (27). A key feature of this protocol is the utilisation of three distinct culture stages (PMA priming, rest, and cytokine differentiation) to generate cultures enriched for M0, M1 and M2 macrophages ( Supplementary Figure 1A ). Importantly, this protocol enabled us to investigate at which stage/s in macrophage differentiation PHX might act. As reported, these macrophage subsets had distinct morphological features ( Supplementary Figure 1B ). Flow cytometric measurement of M1 (CD80) and M2 (CD209) cell surface markers confirmed that the polarizing conditions generated the expected macrophage subsets ( Supplementary Figures 1C, D ). Furthermore, viability was high following culture, although higher in M2 than M1 cultures ( Supplementary Figure 1E ). This may reflect the fact that LPS, a key inducing factor in M1 macrophage differentiation, is cytotoxic to macrophages. Reported LPS concentrations in M1 cultures ranges from 0.01 to 1 μg/mL (40). We titrated the LPS concentration used in our M1 cultures and showed that 250 ng/mL LPS was required to induce a homogenous M1 macrophage population ( Supplementary Figures 2A, B ). Furthermore, lowering the LPS concentration reduced the CD80 expression per cell (as measured by mean fluorescence intensity (MFI)) ( Supplementary Figure 2C ). Thus, we used 250 ng/mL LPS for all subsequent M1 cultures as in the previously published protocol (27).

Upon optimising our macrophage culture conditions, we investigated if PHX affected M1 or M2 macrophage polarization. As PHX is known to have cytotoxic activity, we first determined that THP-1 cells tolerated up to 5 µM PHX for 72 hours with no signs of toxicity, as shown by the alamarBlue viability assay (data not shown). Next, we determined the effect of 1.25, 2.5 and 5 µM PHX on the M0, M1 and M2 cultures ( Figure 1A ). Addition of PHX did not significantly reduce the viability in any of the cultures ( Figure 1B ). By itself PHX did not promote polarization of M0 cells into either M1 or M2 macrophages. We observed a PHX concentration dependent increase in the proportion of M1 macrophages in the M1 cultures, and a decrease in M2 macrophages in the M2 cultures ( Figures 1C–E ). Furthermore, the reduction in M2 macrophages was not the result of a shift to M1 macrophage polarization in the M2 culture. These data suggests that PHX treatment inhibited M2 macrophage differentiation, and enhanced M1 macrophage differentiation only in conjunction with M1 promoting factors such as LPS and IFN-γ.

To determine at which stage PHX had its effect, we selectively added it or vehicle at either the PMA priming, rest, or differentiation stage of the M0, M1, and M2 cultures ( Figure 2 ). Our results showed that the addition of PHX specifically at the differentiation stage of the M1 cultures was enough to enhance M1 macrophage polarization. We observed a significant increase in CD80⁺ M1 macrophages and a decrease in the already low count of CD209+ M2 macrophages present in our M1 cultures ( Figure 2C ). In support, PHX treatment at the differentiation stage in the M1 cultures significantly increased expression of key M1 macrophage genes STAT1 and CCL2 ( Figure 3 ). In contrast, we did not detect a PHX induced reduction in CD209+ M2 macrophages at any of the stages in the M2 cultures ( Figure 2 ). However, we did observe a significant decrease in M2 macrophage related transcription factors STAT6 and c-MYC when PHX was added during the cytokine differentiation stage of the M2 cultures ( Figure 3A ). This suggests that PHX may have partially inhibited M2 macrophage polarization, but higher concentrations or longer treatment times are required to induce a strong inhibitory effect as observed in Figure 1 . Taken together, our results suggest that PHX may enhance M1 and inhibit M2 macrophage polarization. Furthermore, PHX acts primarily when activated THP-1 macrophages are exposed to M1 differentiating factors.

Because the effect of PHX was most pronounced in M1 cultures, we next characterised the changes it induced in the proteome of M1 macrophages ( Supplementary Figure 3A ). Five experimental replicates of vehicle and PHX treated (at the cytokine differentiation stage) THP-1 M1 cultures were subjected to DIA MS, a technique that allows for reproducible quantitative deep proteome-wide profiling (41, 42). Each treatment showed high homogeneity between replicates ( Supplementary Figures 3B, C ). We identified a total of 2,999 unique proteins, with 62 and 15 proteins found exclusively in vehicle and PHX treated M1 cultures, respectively ( Supplementary Figure 3D ; Table S1 ). Interestingly, PHX treatment induced significant changes to the M1 macrophage proteome ( Supplementary Figure 3E ).

Using differential expression analysis of quantitative mass spectrometry data (DEqMS) (28), we identified 488 differentially expressed proteins (DEP, adj. p-value <0.05) with annotated gene names that were altered by PHX treatment ( Figure 4A ). Of these, 44 proteins were upregulated and 444 were downregulated ( Table S2 ), and 11 upregulated and 80 downregulated proteins had at least a 2-log fold change in expression ( Figure 4A ; Table 1 ). Using the lists of upregulated and downregulated DEPs, we used the STRING database to visualise the protein-protein interaction (PPI) networks that were affected by PHX treatment (30). We identified that the 433 downregulated DEPs generated a PPI network consisting of 3 main clusters ( Figure 4B ; Table S3 ). Gene set enrichment analysis (GSEA), based on the Reactome Knowledgebase (31), showed that Cluster 1 proteins were primarily involved in RNA metabolism and could be subdivided into RNA splicing and protein translation ( Supplementary Figure 4A ; Table S4 ). Clusters 2 and 3 had proteins primarily related to the immune system, with Cluster 3 consisting of interferon signalling proteins ( Supplementary Figure 4 ; Table S4 ). Cluster 2 contained diverse signalling pathways including Rho GTPase, mTOR, and Receptor Tyrosine Kinase (RTK). RTKs could be attributed to receptors for growth factors, cytokines, and hormones which in Cluster 2 included VEGFR2, MET and Eph receptors ( Supplementary Figure 4B ). Since PHX has been shown to downregulate mTOR signalling (43, 44), we confirmed that PHX downregulated mTOR, raptor, and other mTOR related proteins in M1 cultures ( Figure 4C ). In contrast, the 44 upregulated DEPs did not form any strong PPI networks (data not shown).

Recently a proteomic analysis of THP-1 derived M0, M1 and M2 macrophage cultures identified 68 M1-associated proteins and 20 M2-associated proteins that were found only in their respective cultures ( Table S5 ) (45). While our culture conditions do not exactly match that study, comparisons of the two data sets identified 33 M1- and three M2-associated proteins in common, supporting that our M1 culture conditions promote differentiation of M1 macrophage phenotypes ( Table S6 ). DEqMS for the identified 33 M1-associated proteins in common showed that PHX treatment significantly downregulated 11 M1-associated proteins and upregulated one, B2M ( Figure 5 ). Thus, PHX treatment while promoting M1 macrophage polarization may eventually downregulate the inflammatory properties of these cells.

To gain a deeper insight into the global changes within our proteomics data, pathway enrichment analysis was performed on all 2999 unique proteins identified by DIA MS. Using the MSigDB, we performed GSEA against Hallmark, Reactome Knowledgebase and Gene Ontology (GO) datasets (35, 36). Interestingly, PHX treatment significantly enriched Hallmark gene sets for oxidative phosphorylation, adipogenesis, cholesterol homeostasis, and fatty acid metabolism. Furthermore, in agreement with our differential enrichment analysis, PHX downregulated genes related to the inflammatory response ( Figure 6 ; Table S7 ). Reactome Knowledgebase pathway analysis supports these findings, with key pathways in the TCA cycle, mitochondrial fatty acid beta oxidation, and gluconeogenesis being upregulated ( Supplementary Figure 5 ; Table S8 ). Finally, our GO Cellular Component analysis shows that PHX upregulated proteins related to the cytoplasm, mitochondria and cell membrane, and downregulated proteins related to the nucleus and endosome ( Figure S6 ; Table S9 ). This once again matches our GO Biological Process ( Supplementary Figure 7 ; Table S10 ) and GO Molecular Function ( Supplementary Figure 8 ; Table S11 ), where there is an enrichment for metabolic process related to fatty acid metabolism and oxidative phosphorylation with a downregulation in RNA metabolism.

To validate our proteomics result, we assessed the cell surface expression of CD14 and B2M on PHX-treated M1 macrophages by flow cytometry. Consistent with our proteomics results ( Figure 5 ), PHX treatment of M1 macrophage cultures resulted in reduced expression of CD14 and enhanced expression of B2M ( Figures 7A, B ). Finally, as IL-1β was one of the most significantly downregulated proteins in our data ( Figures 4A , 6B ), we performed ELISA to investigate if PHX reduced IL-1β release in our M1 cultures ( Figure 7C ). Cell culture supernatants from PHX-treated M1 cultures had lower IL-1β levels compared to vehicle controls ( Figure 7D ). Furthermore, stimulation of these M1 cultures with LPS and ATP showed that while PHX treatment reduced the amount of IL-1β released, it did not prevent these macrophages from secreting IL-1β ( Figure 7E ). Taken together, for the first time, we define the PHX induced changes in the M1 macrophage proteome.