Human monocyte isolation followed protocol described by Schenk et al.13. Peripheral blood mononuclear cells (PBMCs) obtained from donors after consent approved by the UCLA Institutional Review Board in accordance with Declaration of Helsinki Principles. PBMCs isolated using Ficoll (GE Healthcare, Uppsala, Sweden) gradient centrifugation. Monocytes enriched with Percoll density gradient (GE Healthcare) and adherence in RPMI 1640 medium (Thermo Fisher Scientific) with 1% fetal calf serum (FCS; Omega Scientific, Tarzana, CA, USA) for 2 h. Adherent monocytes cultured in 10% FCS. Monocyte purity >80% as measured by CD14 expression.

PA strain 6919 obtained from American Type Culture Collections (ATCC; Manassas, VA, USA) and grown on brucella agar with 5% sheep blood, hemin, and vitamin K (Thermo Fisher Scientific Remel Productions, Lenexa, KS, USA) at 37℃ for 3~5 days in sealed chambers with Anaero Packs (Mitsubishi Gas Chemical Co., Inc., Tokyo, Japan). Cultures grown under same conditions in reinforced clostridial medium (Oxoid, Basingstoke, England). Spectrophotometer OD₆₀₀ used to determine bacterial log phase. Cultures diluted to final concentration of 0.5 multiplicity of infection (MOI) to stimulate cells.

Supernatants from monocytes stimulated for 24 hours with 0.5 MOI PA were processed to measure concentrations of HODEs and hydroxyeicosatetraenoic acids (HETEs) using method adopted from Imaizumi et al.14 LC/MS/MS performed using 4000 QTRAP (Applied Biosystems, Foster City, CA, USA) with electrospray ionization (ESI) source. High-performance liquid chromatography (HPLC) system used Agilent 1200 series LC pump with thermostatted autosampler (Agilent Technologies, Santa Clara, CA, USA). Chromatography performed using Luna C-18(2) column (3 µm particle, 150×3.0 mm). Data acquisitions and instrument control done with Analyst 1.4.2 software (Applied Biosystems). Detection accomplished with multiple reaction monitoring modes with negative ion detection. Collision energy, declustering potential, and collision cell exit potential optimized for each compound to obtain optimum sensitivity. Transitions monitored were mass-to-charge ratio (m/z): m/z 295.1→194.8 for 13-HODE; 295.0→171.0 for 9-HODE; 319.1→219.0 for 15-HETE; 319.1→115.0 for 5-HETE; 319→179 for 12-HETE.

IL-6, IL-8, and IL-1β levels in culture supernatants quantified by enzyme-linked immunosorbent assay (ELISA) using cytokine-specific capture antibodies and secondary biotin-conjugated antibodies (Thermo Fisher Scientific) in reference to recombinant proteins (eBioscience, San Diego, CA, USA) in 96-well plates (Corning, NY, USA). Plates blocked using phosphate-buffered saline (PBS) solution with 10% bovine serum albumin (BSA; Thermo Fisher Scientific, Waltham, MA, USA). Plates washed with PBS containing 0.005% Tween 20 (Sigma-Aldrich, St. Louis, MO, USA). Samples were assayed in triplicates.

Total RNA isolated using Trizol reagent (Thermo Fisher Scientific) following manufacturer's recommendations and treated with RNase-free DNase. RNA isolated with Quick-RNA MiniPrep kit (Zymo Research, Irvine, CA, USA). RNA concentration and quality determined by spectrophotometry at OD₄₆₀ and OD₄₈₀. Samples reverse transcribed with iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) at 25℃ 5 minutes, 42℃ 30 minutes, and 85℃ 5 minutes. Quantitative polymerase chain reaction (qPCR) performed using iQ SYBR Green Supermix (Bio-Rad) (40 cycles of 95℃ 5 minutes, 95℃ 10 seconds, 55℃ 20 seconds, and 72℃ 20 seconds). Glyceraldehyde-3-phosphate dehydrogenase amplification used as internal standard. Primers designed using Primer Express (Applied Biosystems). Gene expression level quantified by comparative method 2^−ΔΔCT.

Cells for detection of G2A treated with 0.5 MOI PA, isotretinoin, and/or all-trans-retinoic acid (ATRA) incubated with anti-G2A polyclonal antibody (MBL International; Woburn, MA, USA), Immunoglobulin (Ig)G isotype control (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and secondary Alexa Fluor 488-conjugated antibody (Thermo Fisher Scientific). Between incubation steps, cells washed with PBS containing 1% bovine serum albumin. Cells fixed with 2% paraformaldehyde (Thermo Fisher Scientific), acquired on BD FACScan System, and analyzed using CellQuest-Pro software (BD, Franklin Lakes, NJ, USA).

Small interfering RNA (siRNA) transfection performed with Lonza Nucleofector system with either the Lonza Human Monocyte Nucleofector kit or Lonza Cell Line Nucleofector kit V for human promyelocytic cell line THP-1 (Lonza, Basel, Switzerland). Scramble control and anti-G2A siRNA constructs were used at 1 µg per transfection (Thermo Fisher Scientific). Transfection efficiency of expression constructs assessed using the pmax-GFP construct (Lonza), and yielded average of 50% transfection rate.

Detections of phospho-p44/42 mitogen-activated protein kinase (MAPK), phospho-p38 MAPK, and NF-κB p65 were performed using western blot. After PA treatment, cells lysed in buffer containing 1% Triton X-100 (Sigma-Aldrich) and Protease/Phosphotase Inhibitor Cocktail (Cell Signaling, Danvers, MA, USA). Protein concentrations assessed with Quick Start Bradford Protein Assay kit following manufacturer's protocol. Lysates loaded with Laemmli Sample Buffer onto Any kD Mini-PROTEAN-TGX pre-cast polyacrylamide gel (Bio-Rad) and separated by electrophoresis. Gels transferred onto polyvinylidene difluoride membranes (EMD Millipore, Billerica, MA, USA) and detected with primary antibodies for MAPK (Cell Signaling) or NF-κB (Sigma-Aldrich) pathways, secondary horse radish peroxidase-conjugated antibody (Abcam, Cambridge, England), and chemiluminescent substrate (Thermo Fisher Scientific) before visualization with ChemiDoc MP System (Bio-Rad).

Total RNA from five acne lesions and five matched non-lesional controls was obtained from Diane Thiboutot (Pennsylvania State University, Hershey, PA, USA)15. Biopsies taken from backs of men and women without other skin diseases in biopsy areas. Gene expression profiles were re-measured using the Affymetrix GeneChip, Human Genome (HG)-U133A Plus 2.0 array platform to include GPR132, which was not included in the Trivedi et al. report15. Data analyzed as previously described16. Analysis identified 890 genes that were significantly elevated in acne lesions (fold-change≥1.5 and p<0.05), compared to controls.

Neutralizing antibodies against Toll-like receptor 2 (TLR2) and TLR4, and control antibodies (BioLegend, San Diego, CA, USA); 9(S)-HODE (Cayman Chemical Company, Ann Arbor, MI, USA); human acute monocytic leukemia cell line THP-1 (ATCC) cultured in RPMI Medium 1640 supplemented with 10% FCS, 1% Sodium Pyruvate (Thermo Fisher Scientific), and Penicillin-Streptomycin-L-Glutamine (Corning); Tris/Glycine and Tris/Glyine/SDS buffers and blotting-grade blocker for western blot (Bio-Rad); all-trans-retinoic acid (retinoic acid) and isotretinoin (13-cis-retinoic acid [13cRA]) (Sigma-Aldrich) reconstituted in dimethyl sulfoxide (Thermo Fisher Scientific).

Results expressed as means±standard error (SE) for the number of separate experiments indicated in each Fig. 1, 2, 3, 4, 5. Post hoc two-tailed Student's t-test used for comparison between two groups. For Fig. 4, 5B, to address increased likelihood of Type I errors when making comparisons over multiple groups, p-values adjusted using Bonferroni correction. For Fig. 5A, post hoc two-tailed paired t-test performed to ensure analysis of paired samples. Significant differences considered for those with p-values ≤0.05.

To test G2A expression under different conditions, we examined the effects of 9(S)-HODE and PA on THP-1 cells and primary human monocytes. G2A mRNA expression or the population of cells expressing G2A in either THP-1 or monocytes was unchanged after treatment with different concentrations of 9(S)-HODE (Fig. 1A). In contrast, PA exposure increased the population of THP-1 cells expressing G2A approximately 35-fold compared to isotype control (Fig. 1B), and increased the gene expression of GPR132 in both THP-1 and monocytes by approximately 50% and 40%, respectively (Fig. 1A, p<0.05). We next assessed the effect of TLR2, an important mediator of innate immune system in response to PA17. We observed a significant 30% reduction in GPR132 expression in monocytes pre-treated with TLR2 neutralizing antibody prior to addition of PA, as compared to the isotype control (Fig. 1C, p<0.01).

9-HODE is a known ligand of G2A12. We sought to characterize the baseline levels of this ligand and related oxidized lipids produced from human monocytes as well as their inducibility with PA. Monocytes were treated with PA for 24 hours; then cell lysates were purified and analyzed with liquid chromatography and tandem mass spectrometry. We measured the oxidized forms of linoleic acid, 9-HODE and 13-HODE, as well as three oxidized forms of another FFA, arachidonic acid, 5-, 12-, and 15-HETE. Both 9-HODE and 13-HODE were found in significantly higher concentrations (5-fold and 10-fold, respectively) than all three HETEs. Exposing monocytes to PA increased the concentration of 9-HODE by approximately 2-fold (Fig. 2, p<0.001), whereas 13-HODE was not significantly altered. Of the three HETEs measured, only 15-HETE was found to be significantly upregulated by PA (p<0.05); however, its concentration was notably less than the concentrations of both the HODEs.

To understand the immunomodulatory functions of G2A, we transfected THP-1 cells using siRNA oligonucleotides specific to G2A mRNA (siG2A), then measured gene expression and protein secretion of inflammatory cytokines IL-6, IL-8, and IL-1β, following exposure to PA. Our transfection methods consistently yielded approximately 50% knockdown of G2A mRNA expression when compared to the non-targeting oligonucleotide control (siCTRL) (Supplementary Fig. 1, p<0.05). Knockdown of the G2A gene in THP-1 cells led to an increase in inflammatory cytokines both in response to PA and G2A ligand 9(S)-HODE. A 24 hours incubation of siG2A transfected THP-1 cells with live PA resulted in increases of approximately 1.8-fold for IL-6 (Fig. 3A, p<0.01) and 1.3-fold for IL-8 (Fig. 3C, p<0.05), but no change in IL-1β; when compared to control (Fig. 3E). At the mRNA level, G2A knockdown in THP-1 cells stimulated with PA resulted in approximately 2.2-fold increase of IL6 (Fig. 3B, p<0.01), 3.9-fold increase of IL8 (Fig. 3D, p<0.01), and 1.9-fold increase of IL1B (Fig. 3F, p<0.05) mRNA when compared to control. G2A knockdown THP-1 cells treated with 9(S)-HODE exhibited increases of approximately 2.7-fold for IL6 (Fig. 3B, p<0.05), 1.2-fold increase for IL8 (Fig. 3D, p<0.05), and 1.4-fold increase for IL1B (Fig. 3F, p<0.001) when compared to control. At the protein level, there was nearly a 2.3-fold increase of IL-8 (Fig. 3C, p<0.05) when comparing PA-stimulated cells transfected with siG2A to siCTRL, while levels of IL-6 and IL-1β were not significantly changed.

The MAPK and NF-κB pathways are thought to be integral in the regulation of inflammatory cytokines IL-6 and IL-8. To elucidate the signaling pathways that are activated by G2A to regulate inflammatory response, we measured the phosphorylation of two MAPK pathways, ERK1/2 (p44/42) and p38 MAPK, and NF-κB p65 using western blot. Densitometric analysis of blot bands showed that the G2A knockdown in monocytes stimulated with PA exhibited approximately 2.3-fold increase in the phosphorylation of the ERK1/2 MAPK pathway compared to control (Fig. 4A, p<0.01), and nearly a 1.5-fold increase in NF-κB p65 subunit (Fig. 4C, p<0.05). PA induced approximately a 50% increase in the phosphorylation of p38 MAPK in the scramble control group (Fig. 4B, p<0.05); however, this was not found to be significantly modulated by the G2A knockdown.

To explore the potential clinical implications of G2A in acne, we compared the mRNA profile of acne lesions to normal skin. Total RNA samples from punch biopsies of five acne lesions and matched non-lesional controls (normal skin) were obtained from D. Thiboutot15. mRNA expression profiles were assessed using Affymetrix gene expression microarray system. For each of the donors, G2A mRNA expression (GPR132) was higher in the acne lesion biopsies than in the matched controls. When the expression values of donors were averaged, the mean GPR132 expression was doubled (Fig. 5A, p<0.001) in acne lesions compared to normal skin. Comparison of secretoglobin SCGB1D2 (p<0.01) and human beta-defensin 4 DEFB4 (p<0.05) from the current data set corroborates previously published microarray data obtained from the same lesional and control samples using a previous version of the Affymetrix human array (U133A 2.0)15.

Retinoids, a class of vitamin A-derived compounds that bind various members of the retinoic acid receptor181920, affect keratinocyte maturation and anti-inflammatory activity2122. Retinoids also downregulate TLR2 expression in monocytes23. We investigated whether 13cRA and ATRA, two widely used therapeutics for acne2425 could regulate G2A expression in monocytes. Monocytes were pre-treated with 13cRA and ATRA for 1 hour prior to a 24 hours exposure to live PA. These monocytes were then labeled with anti-G2A antibodies and analyzed using flow cytometry. Compared to control, ATRA alone did not induce a significant change in the percentage of cells expressing G2A but 13cRA induced a 10-fold increase (Fig. 5B, p<0.001). When exposed to PA, monocytes exhibited a 100-fold increase in the percentage of cells expressing G2A compared to untreated cells (p<0.0001). Concurrent treatment with ATRA did not significantly modulate the PA-induced upregulation of G2A whereas 13cRA increased G2A expression on monocytes by approximately 30% compared to PA alone (p<0.05).