Ceramide kinase regulates acute wound healing by suppressing 5-oxo-ETE biosynthesis and signaling via its receptor OXER1

Abstract

The sphingolipid, ceramide-1-phosphate (C1P), has been shown to promote the inflammatory phase and inhibit the proliferation and remodeling stages of wound repair via direct interaction with group IVA cytosolic phospholipase A₂, a regulator of eicosanoid biosynthesis that fine-tunes the behaviors of various cell types during wound healing. However, the anabolic enzyme responsible for the production of C1P that suppresses wound healing as well as bioactive eicosanoids and target receptors that drive enhanced wound remodeling have not been characterized. Herein, we determined that decreasing C1P activity via inhibitors or genetic ablation of the anabolic enzyme ceramide kinase (CERK) significantly enhanced wound healing phenotypes. Importantly, postwounding inhibition of CERK enhanced the closure rate of acute wounds, improved the quality of healing, and increased fibroblast migration via a "class switch" in the eicosanoid profile. This switch reduced pro-inflammatory prostaglandins (e.g., prostaglandin E2) and increased levels of 5-hydroxyeicosatetraenoic acid and the downstream metabolite 5-oxo-eicosatetraenoic acid (5-oxo-ETE). Moreover, dermal fibroblasts from mice with genetically ablated CERK showed enhanced wound healing markers, while blockage of the murine 5-oxo-ETE receptor (oxoeicosanoid receptor 1) inhibited the enhanced migration phenotype of these cell models. Together, these studies reinforce the vital roles eicosanoids play in the wound healing process and demonstrate a novel role for CERK-derived C1P as a negative regulator of 5-oxo-ETE biosynthesis and the activation of oxoeicosanoid receptor 1 in wound healing. These findings provide foundational preclinical results for the use of CERK inhibitors to shift the balance from inflammation to resolution and increase the wound healing rate.

Keywords: 5-HETE; 5-oxo-ETE; arachidonic acid; ceramide kinase; ceramide-1-phosphate; eicosanoids; group IVA phospholipases A2; inflammation; lipidomics.

Fig. 1

SYR382141 decreases ceramide-1-phosphate levels in cells. A: WT pDFs pretreated with SYR382141 (100 nM), NVP-231 (100 nM), or DMSO (0.01%) for 30 min received mechanical trauma via asterisk pattern scratch across the plate. Cells were collected 2 h post-injury and analyzed for C1P levels via UPLC ESI-MS/MS. Values expressed as fold change to DMSO controls (∗P < 0.05, ∗∗P < 0.01, ∗∗∗∗P < 0.0001; n = 3, pDFs collected from three different mice; two-way ANOVA with Dunnett's multiple comparisons test). B: HUVECs were treated with SYR382141 (100 nM), NVP-231 (300 nM), or DMSO (0.01%) for 24 h. Cells were collected and analyzed for C1P levels via UPLC ESI-MS/MS. Values expressed as fold change to DMSO controls. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; n = 3; one-way ANOVA with Tukey post-hoc test).

Fig. 2

Inhibition of ceramide kinase increases the closure rate of acute wounds in mice. A: Wound closure rate of 5 mm biopsy wound on dorsum of CERK-KO or WT mice treated with carboxymethyl cellulose (CMC) control (1% CMC) or SYR382141 (60 mg/kg), orally twice daily beginning 1 day post-injury (1 wound per mouse repeated on two separate occasions). B: Graph depicting acute wound closure rate quantified as percent of initial wound size over 10 days. Two-way ANOVA with Dunnett's multiple comparisons test, ∗P < 0.05, ∗∗∗∗P < 0.0001; n = 5 wounds per genotype, 1 per mouse.

Fig. 3

Inhibition and genetic ablation of ceramide kinase improve wound quality. A: Wound tissue harvested 10 days post-injury from WT control (1% CMC), CERK-KO, and SYR382141-treated (60 mg/kg) WT mice under H&E (cell infiltration), Masson’s Trichrome (collagen deposition), FAP (fibroblast activation protein), and type I collagen staining. B: Graph depicting quantification of infiltrating cells, FAP area, and type I collagen area, analyzed via ImageJ cell counter and Fiji ImageJ bundle area tool, contrast enhanced (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; n = 3 samples per treatment group, 1 wound per mouse; two-way ANOVA with Dunnett’s multiple comparisons test).

Fig. 4

Inhibition or genetic ablation of ceramide kinase enhances the migration of dermal fibroblasts and HETE biosynthesis. A: pDFs from WT, CERK-KO, and cPLA₂α-KI mice treated with DMSO (0.001%), SYR382141 (100 nM) or NVP-231 (100 nM). Still images from time points 0- and 24-h. Brightness enhanced; lines added for emphasis. Cells were observed using a live cell incubation chamber mounted on a Keyence BZ-X710 microscope which took images every 3 min for 24 h (n = 4; pDFs taken from two separate animals per genotype). B: Graph depicting migration velocities of pDFs treated with CERK inhibitors SYR382141 (100 nM) or NVP-231 (100 nM), calculated using the Keyence VW-9000 motion analysis software (Dunnett's multiple comparisons test; n = 4; pDFs taken from two separate animals per genotype). C: C1P (C:16 (C16:0) and C:24 (C24:0)) production in wound tissue from CERK-KO mice compared to WT (n = 4 per genotype; one wound per mouse). D: Eicosanoids from WT, CERK-KO, and cPLA₂α-KI pDFs pretreated with SYR382141 or DMSO control collected 2 h after mechanical injury (two-way ANOVA with Dunnett’s multiple comparisons test; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; n = 3, pDFs taken from three separate mice per genotype).

Fig. 5

Inhibition of ceramide kinase enhances 5-HETE and 5-oxo-ETE biosynthesis in acute wounds. A: Heatmap representation of eicosanoid profile of wound tissue harvested 10 days post-injury from WT mice treated with SYR382141 (60 mg/kg) or control (1% CMC) (n = 3 per treatment group; 1 wound sample per mouse). B: Graphical comparison of C1P profile of wound tissue harvested 10 days post-injury (n = 3 per treatment; 1 wound sample per mouse; two-way ANOVA with Šídák's multiple comparisons test). C: Graph depicting eicosanoid profile of wound tissue harvested 10 days post-injury. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; n = 3 per treatment group; 1 wound sample per mouse; two-way ANOVA with Šídák's multiple comparisons test).

Fig. 6

A: Graph depicting pDF migration velocity of WT, CERK-KO, and cPLA₂α-KI pDFs treated with combinations of FLAP inhibitor MK886 (7.5 nM), OXER1 antagonist Gue1654 (10 μM), ceramide kinase inhibitor SYR382141 (100 nM). All values compared to WT DMSO control (n = 3 per treatment; pDFs taken from three separate mice per genotype; two-way ANOVA with Dunnett's multiple comparisons test; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001). B: Graph depicting pDF migration velocity of WT, CERK-KO, and cPLA₂α-KI pDFs treated with combinations of 5-HETE (100 nM), and 5-oxo-ETE (1 nM) treatments in combination with MK886 (7.5 nM) and Gue1654 (10 μM). All values compared to panel (A) WT DMSO control (n = 3 per treatment; pDFs taken from three separate mice per genotype; two-way ANOVA with Dunnett's multiple comparisons test; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001). C: Representative microscope images of 5-oxo-ETE rescue and Gue1654 suppression of pDF migration from data graphed in panels A and B. Contrast enhanced, lines added for emphasis.

Fig. 7

Inhibition of CERK-induced C1P elevates 5-HETE and its conversion to 5-oxo-ETE and subsequent action on OXER1-like receptor resulting in increased fibroblast activity and expedited wound healing. The presence of CERK results in elevated C1P, leading to an eicosanoid shift from HETEs to PGE₂ resulting in delayed wound healing via reduced fibroblast migration.