Antidepressants Target the ST3GAL5-GM3 Lipid Pathway to Suppress Microglial Inflammation

Abstract

Major depression (MD) is associated with chronic inflammation and impaired neuroplasticity; however, the cellular mechanisms underlying antidepressant action remain incompletely understood. We performed transcriptomic profiling and functional validation in human microglia treated with venlafaxine (VEN) and vortioxetine (VOR), or with stable ST3GAL5 overexpression (ST3GAL5OE). Differential expression analysis, enrichment studies, and functional assays using NF-κB-RE-NlucP and SIE-NlucP reporter lines were conducted to assess the impact on inflammatory signaling. Microarray analysis identified 41 genes consistently upregulated and 316 consistently downregulated across VEN, VOR, and ST3GAL5OE conditions. Upregulated genes were enriched for synaptic organization, whereas downregulated genes were associated with nitric oxide biosynthesis and pro-inflammatory pathways, including Rap1, MAPK, and PI3K-Akt signaling. Functional assays confirmed that VEN and VOR suppressed cytokine-induced NF-κB and STAT3 activation, effects that were recapitulated by exogenous GM3 treatment and ST3GAL5 overexpression. Chronic exposure to VEN or VOR produced more modest, pathway-specific suppression, supporting convergence on the ST3GAL5-GM3 axis. These findings extend the conventional monoaminergic model of antidepressant action by highlighting the ST3GAL5-GM3 lipid remodeling axis as a novel regulatory pathway that attenuates microglial inflammatory signaling. Although validation in primary microglia and in vivo models is required, our results suggest that this axis could serve as both a therapeutic target and a candidate biomarker for inflammation-associated MD.

Keywords: GM3 ganglioside; ST3GAL5; antidepressants; cytokine signaling; inflammation; major depression; microglia; venlafaxine; vortioxetine.

Figure 1

Commonly regulated genes across VEN, VOR, and ST3GAL5OE. Venn diagrams showing overlap of upregulated (left) and downregulated (right) genes (|log2FC| ≥ 0.3) shared among venlafaxine (VEN), vortioxetine (VOR), and ST3GAL5 overexpression (ST3GAL5OE). A total of 41 genes were commonly upregulated and 316 were commonly downregulated across all three conditions.

Figure 2

Functional enrichment of commonly regulated genes. (A) GO Biological Process enrichment of commonly upregulated genes, highlighting synapse-related pathways such as regulation of synapse organization, synaptic vesicle cycle, and neuron cell-cell adhesion. (B) GO Biological Process enrichment of commonly downregulated genes, including nitric oxide biosynthetic process and reactive nitrogen species metabolism, suggesting suppression of pro-inflammatory mediator production. Adjusted p-values are shown.

Figure 3

Reporter assays confirm cytokine-induced NF-κB and STAT3 activation. (A) Dose-response of NF-κB-RE-NlucP reporter cells stimulated with TNF-α (0–10 ng/mL). (B) Dose-response of SIE-NlucP reporter cells stimulated with IL-6 (0–50 ng/mL). (C,D) ATP viability assays showing minimal cytotoxicity under these conditions. Methods: MG cells stably expressing NF-κB-RE-NlucP or SIE-NlucP were seeded into 96-well white plates at 1 × 10³ cells/well in 50 µL medium. After 24 h, cells were stimulated with TNF-α (NF-κB-RE-NlucP) or IL-6 (SIE-NlucP) at the indicated concentrations (maximum 100 ng/mL for TNF-α; 50 ng/mL for IL-6) using 5-fold serial dilutions. Following 24 h incubation, NanoLuc activity was measured using the Nano-Glo^® Luciferase Assay System (A,B), and ATP levels were assessed with the CellTiter-Glo^® assay (C,D). Data represent mean ± SD (n = 3).

Figure 4

Exogenous GM3 suppresses cytokine-induced NF-κB and STAT3 activation. (A) GM3 (5 µM) reduced TNF-α-induced NF-κB activation. (B) GM3 reduced IL-6-induced STAT3 activation. (C,D) ATP viability assays confirmed no additional cytotoxicity with GM3 treatment. Methods: MG cells stably expressing NF-κB-RE-NlucP or SIE-NlucP were seeded into 96-well white plates at 5 × 10³ cells/well in 50 µL medium. GM3 (5 µM) was added at seeding. After 24 h, cells were stimulated with TNF-α (NF-κB-RE-NlucP) or IL-6 (SIE-NlucP) at the indicated concentrations (maximum 10 ng/mL for TNF-α; 50 ng/mL for IL-6) using 2-fold serial dilutions. Following a further 24 h incubation, NanoLuc activity was measured using the Nano-Glo^® Luciferase Assay System (A,B), and ATP levels were assessed with the CellTiter-Glo^® assay (C,D). Data represent mean ± SD (n = 3).

Figure 5

ST3GAL5 overexpression attenuates inflammatory signaling. (A) Western blot confirming ST3GAL5-DYK expression in transfected cells. Cell lysates from ST3GAL5-DYK stably expressing cells (ST3GAL5OE) were subjected to Western blotting, and the same membrane was sequentially probed with anti-DYK and anti-β-actin antibodies. (B,C) Reporter assays showing suppression of TNF-α-induced NF-κB and IL-6-induced STAT3 activity in ST3GAL5OE cells compared with vector controls. (D) ATP levels were unaffected, confirming no cytotoxicity. Methods: MG cells stably transfected with ST3GAL5-DYK (ST3GAL5OE) or vector control were seeded into 96-well white plates at 5 × 10³ cells/well in 50 µL medium. After 24 h, TNF-α (maximum 10 ng/mL) or IL-6 (maximum 50 ng/mL) was added with 2-fold serial dilutions. Following an additional 24 h incubation, NanoLuc activity was measured using the Nano-Glo^® Luciferase Assay System (B,C), Data represent mean ± SD (n = 3).

Figure 6

Chronic antidepressant exposure phenocopies ST3GAL5 effects. (A) VEN200 cells (cultured for >7 days in the presence of 200 µM venlafaxine) exhibited reduced TNF-α-induced NF-κB activation. (B) VOR40 cells (cultured for >7 days in the presence of 40 µM vortioxetine) exhibited reduced IL-6-induced STAT3 activation. Methods: MG cells were chronically pretreated with venlafaxine (200 µM) or vortioxetine (40 µM) for at least one week and designated as VEN200 and VOR40, respectively. These cells were seeded into 96-well white plates at 5 × 10³ cells/well in 50 µL medium. After 24 h, TNF-α (maximum 10 ng/mL, 2-fold serial dilutions) or IL-6 (maximum 50 ng/mL, 2-fold serial dilutions) was added. Following an additional 24 h incubation, NanoLuc activity was measured using the Nano-Glo^® Luciferase Assay System. Data are presented as mean ± SD (n = 3).