Dysregulation of Arachidonic Acid Metabolism Drives Inflammatory Lipid Production in Localized Provoked Vulvodynia

Abstract

Background/Objectives: Localized provoked vulvodynia (LPV) is characterized by chronic vulvar pain upon light touch to the vestibule, a specialized ring of tissue immediately surrounding the vaginal opening. LPV affects about 14 million people in the US, yet the etiopathology of the disease is unknown. In LPV, the vestibule expresses elevated levels of the pro-nociceptive pro-inflammatory mediators prostaglandin E₂ (PGE2) and interleukin-6 (IL-6), which corresponds to lower pain thresholds. Previous studies have shown reduced amounts of arachidonic acid (AA)-derived pro-resolving lipid mediators in tissue biopsies from LPV patients that might impede the resolution of inflammation. AA is obtained from dietary linoleic acid, pointing to a defect in the metabolism of dietary polyunsaturated fatty acids in LPV. We aimed to further explore the involvement of AA metabolism in LPV, which appears dysregulated in the vestibule of LPV patients and culminates in chronic inflammation and chronic pain. Methods: Vestibular and vulvar tissue biopsies obtained from LPV and non-LPV patients were used to generate fibroblast strains and assessed for COX/LOX expression using qRT-PCR. Fibroblast strains were treated with inflammatory stimuli, and then COX-1 and COX-2 expression was assessed using Western blot analysis. Pro-inflammatory mediator production was assessed using enzyme-linked immunosorbent assays (ELISAs). ALOX5 and ALOX12 expression was assessed using qRT-PCR. Finally, lipidomic analysis was carried out to screen for 143 lipid metabolites following inflammatory challenge. Results: Tissue and fibroblasts from LPV patients exhibited altered expression of COX/LOX enzymes and production of AA-derived lipid mediators compared to non-LPV patients. Conclusions: Lipid profiles of tissue and vestibular fibroblasts from LPV patients differed from non-LPV patients, and this difference was attributed to differential COX/LOX expression and activity, which metabolizes AA derived from dietary linoleic acid. This dysregulation fosters chronic inflammation and reduced resolution capacity in LPV patients, causing chronic pain. While further work is needed, these findings suggest that dietary modifications could impact the LPV mechanism.

Keywords: 12-LOX; COX-1; COX-2; arachidonic acid; chronic pain; inflammation; lipid metabolism; localized provoked vulvodynia; specialized pro-resolving mediators.

Figure 1

LPV vestibular tissue displays altered expression of cyclooxygenase (COX) and lipoxygenase (LOX) enzymes. (A) Illustration of the Q-tip test used to diagnose LPV. Light touch to the tissue surrounding the vaginal opening (the vulvar vestibule, referred to as “Vest” throughout) is immensely painful, whereas touching the adjacent external vulva (referred to as “Vulv”) elicits little pain. Moreover, 6 mm punch biopsies of vestibular and vulvar tissue were collected from LPV patients and healthy controls during routine gynecologic surgeries. Biopsies were cut into three pieces, with one piece being used to assess expression of COX/LOX enzymes. (B–G) Individual plots of qPCR results. COX-2 (PTGS2) and 5-LOX (ALOX5) mRNA levels were significantly elevated in the painful LPV vestibule compared to non-painful controls, while expressions of COX-1 (PTGS1) and 15-LOX-2 (ALOX15B) were decreased. 12-LOX (ALOX12) expression was enhanced in the vestibule of control patients with no significant changes identified in LPV tissue. No significant differences were observed in 15-LOX-1 (ALOX15) expression across the four groups. All data points were normalized to 18S rRNA expression, and statistical significance was determined utilizing linear mixed effects models. LPV/control and vest/vulva sampling sites were treated as fixed effects and biological replicates as random effects. Data are represented with mean ± SEM, n = 17 LPV and n = 10 controls. p < 0.05 was used as the threshold for significance, with * designating pairwise significance.

Figure 2

COX-1 levels decline in LPV cases with IL-1β and AA treatment. Vulvar fibroblasts were incubated with vehicle, IL-1β [500 pg/mL], arachidonic acid (AA) [1 µM], or a combination of IL-1β and AA treatments for 48 h prior to being lysed for Western blot analyses. Membranes were probed with COX-1 rabbit mAB (Abclonal # A4301) at a dilution of 1:2000 and the goat anti-rabbit IgG Starbright Blue 700 secondary antibody (Bio-Rad #12004161) at a dilution of 1:2500. COX-1 expression levels for (A) LPV vestibular, (B) LPV external vulvar, (C) control vestibular, and (D) control external vulvar fibroblasts were assessed for three cases and three controls. Stimulation with AA alone and in combination with IL-1β decreased expression of COX-1 in case vestibular and vulvar fibroblasts. No statistically significant expression changes were identified in controls across the four conditions. Cropped Western blot images used for analysis appear below each graph. COX-1 bands were identified at 72 kDa. Full blot images are available in Figure S3. Statistical significance was determined utilizing linear mixed-effect models where cell treatments were assigned as fixed effects, and patient ID was treated as a random effect to account for biological replicates. p < 0.05 was used as the threshold for significance, with * designating significant changes between the four treatments. ns denotes comparisons that did not reach significance (ns = not significant).

Figure 3

Arachidonic acid induces COX-2 expression in vulvar fibroblasts. Cells were incubated with vehicle, IL-1β [500 pg/mL], arachidonic acid (AA) [1 µM], or a combination of IL-1β and AA treatments for 48 h and then were collected for Western blot analyses. Membranes were probed with COX-2 mouse mAb (Invitrogen #35-8200) at a dilution of 1:500 and peroxidase AffiniPure goat anti-mouse secondary antibody (Jackson ImmunoResearch #115-035-003) at a dilution of 1:20,000. COX-2 expression levels for (A) LPV vestibular, (B) LPV external vulvar, (C) control vestibular, and (D) control external vulvar fibroblasts were assessed for three cases and three controls. Addition of AA alone and in combination with IL-1β stimulated COX-2 expression in LPV and control vestibular cells. Addition of AA with IL-1β also significantly increased COX-2 expression in LPV and control external vulvar fibroblasts. Cropped Western blot images used for analysis appear below each graph. COX-2 bands at 80 kDa were used for analysis. Full blot images are available in Figure S3. Statistical significance was determined utilizing linear mixed-effect models where cell treatments were assigned as fixed effects, and patient IDs were treated as random effects to account for biological replication. p < 0.05 was used as the threshold for significance, with * designating significant changes between the four treatments.

Figure 4

Pro-inflammatory cytokine production is enhanced with arachidonic acid treatment in vulvar fibroblasts. Cells were incubated with IL-1β [500 pg/mL], arachidonic acid (AA) [1 µM], and a combination of treatments for 48 h, and supernatants were collected for ELISA analyses. (A) Levels of prostaglandin-E2 (PGE₂) and (B) interleukin-6 (IL-6) in vulvar cells after 48-h vehicle, IL-1β [500 pg/mL], arachidonic acid (AA) [1 µM], and combination treatments. Addition of AA significantly elevated production of PGE₂ in LPV vestibular cells, and addition of AA in combination with IL-1β significantly stimulated production of PGE₂ in all strains. IL-1β alone and in combination with AA also induced IL-6 production in both LPV and control fibroblasts. Statistical significance was determined utilizing linear mixed effects models treating LPV/control, location, and treatments as fixed effects and technical replicates as random effects. p < 0.05 was used as the threshold for statistical significance, and in the graphs, “†” denotes significant change from vehicle, “‡” denotes significant change from IL-1β, and “#” denotes significant change from arachidonic acid for a particular group (i.e., LPV vest). Data were measured in triplicate and represented with mean ± SEM, n = 3 LPV, and three controls per treatment. p < 0.05 was used as the threshold for statistical significance denoted by *.

Figure 5

Lipoxygenase expression and enzymatic activity in vulvar fibroblasts. Cells were incubated with vehicle, IL-1β [500 pg/mL], arachidonic acid (AA), and AA in combination with IL-1β for 48 h prior to extracting RNA for gene expression analyses. Levels of (A) ALOX5 and (B) ALOX12 were assessed utilizing RT-qPCR. No significant changes in ALOX5 were observed, whereas ALOX12 was enhanced upon treatment with IL-1β for LPV vestibular cells. The addition of arachidonic acid and IL-1β reduced ALOX12 expression for both LPV and control vestibular fibroblasts. No changes in external vulvar ALOX12 expression were observed. ALOX15 and ALOX15B were undetectable in vulvar fibroblast cells. (C) Lipoxygenase activity assay of vulvar fibroblasts revealed reduced enzymatic activity of LOX enzymes in vestibular fibroblasts compared to cells cultured from the external vulva. Specific activity was calculated by subtracting the activity of samples incubated with a LOX inhibitor from uninhibited samples. One unit is defined as the amount of lipoxygenase that oxidizes 1 µmol of LOX probe per minute at a pH of 7.4 at room temperature. Statistical significance was determined utilizing linear mixed effects models with fixed effects for LPV/control, location, and treatment groups and random effects for patient ID. p < 0.05 was used as the threshold for statistical significance denoted by *, and in the graphs, “†” denotes significant change from vehicle, “‡” denotes significant change from IL-1β, and “#” denotes significant change from arachidonic acid for a particular group. Data are represented with mean ± SEM, n = 5 LPV, and six controls measured in triplicate for qPCR assays, and n = 3 for the LOX activity assay.

Figure 6

Targeted metabololipidomics of arachidonic acid-derived lipid mediators. LC-MS lipidomic analysis of >150 lipids in vulvar fibroblasts treated with arachidonic acid (AA) [1 µM], IL-1β [500 pg/mL] in combination with arachidonic acid, and Poly(I:C) [50 pg/mL] in combination with arachidonic acid. Principal component analysis (PCA) was performed on samples from VEHICLE, AA, IL-1β + AA, and Poly(I:C) + AA (shown in different colors). (A) 2D and (B) 3D plots of the samples along the first principal components were created to illustrate the major differences between the tissue types and denote LPV vest tissue with a different symbol than the other tissues. (C) Heatmaps of lipids separated by general eicosanoid type. The LPV vest is the fourth column in each group (denoted with a large arrow at the top); the LPV vest has increases in most HDoHEs and HEPEs under both endogenous and exogenous inflammatory stimuli. Heatmaps for the lipids were sorted by average quantity of all vehicle-treated samples. Each lipid (row) was scaled by the maximum quantity for that lipid so the relative quantities of lipids in different treatments are apparent. Heatmaps were constructed to visualize trends across groups/treatments, and comparison of quantities between lipids is not possible in these figures. Therefore, selected lipids are graphed in Figure 7. n = 3 cases and three controls run in triplicate.

Figure 6

Targeted metabololipidomics of arachidonic acid-derived lipid mediators. LC-MS lipidomic analysis of >150 lipids in vulvar fibroblasts treated with arachidonic acid (AA) [1 µM], IL-1β [500 pg/mL] in combination with arachidonic acid, and Poly(I:C) [50 pg/mL] in combination with arachidonic acid. Principal component analysis (PCA) was performed on samples from VEHICLE, AA, IL-1β + AA, and Poly(I:C) + AA (shown in different colors). (A) 2D and (B) 3D plots of the samples along the first principal components were created to illustrate the major differences between the tissue types and denote LPV vest tissue with a different symbol than the other tissues. (C) Heatmaps of lipids separated by general eicosanoid type. The LPV vest is the fourth column in each group (denoted with a large arrow at the top); the LPV vest has increases in most HDoHEs and HEPEs under both endogenous and exogenous inflammatory stimuli. Heatmaps for the lipids were sorted by average quantity of all vehicle-treated samples. Each lipid (row) was scaled by the maximum quantity for that lipid so the relative quantities of lipids in different treatments are apparent. Heatmaps were constructed to visualize trends across groups/treatments, and comparison of quantities between lipids is not possible in these figures. Therefore, selected lipids are graphed in Figure 7. n = 3 cases and three controls run in triplicate.

Figure 7

Individual lipid plots of metabolites of interest. (A–I) LC-MS lipidomic analysis of >150 lipid species for vulvar fibroblasts treated with vehicle, arachidonic acid (AA) [1 µM], IL-1β [500 pg/mL] in combination with arachidonic acid, and Poly(I:C) [50 pg/mL] in combination with arachidonic acid. Straight lines denote differences between two treatment groups at either end of the line. Bars with ticks denote significant differences between LPV/control locations under the ticks. Data were measured in triplicate and displayed with mean ± SEM, n = 3 LPV and three controls per treatment, run in triplicate. Statistical significance was determined utilizing linear mixed effects models with fixed effects for LPV/control, location, and treatment groups and random effects for patient ID. p < 0.05 was used as the threshold for significance, with * designating significant differences between groups.

Figure 7

Individual lipid plots of metabolites of interest. (A–I) LC-MS lipidomic analysis of >150 lipid species for vulvar fibroblasts treated with vehicle, arachidonic acid (AA) [1 µM], IL-1β [500 pg/mL] in combination with arachidonic acid, and Poly(I:C) [50 pg/mL] in combination with arachidonic acid. Straight lines denote differences between two treatment groups at either end of the line. Bars with ticks denote significant differences between LPV/control locations under the ticks. Data were measured in triplicate and displayed with mean ± SEM, n = 3 LPV and three controls per treatment, run in triplicate. Statistical significance was determined utilizing linear mixed effects models with fixed effects for LPV/control, location, and treatment groups and random effects for patient ID. p < 0.05 was used as the threshold for significance, with * designating significant differences between groups.

Figure 8

Simplified overview of SPM biosynthesis. Arachidonic acid, eicosapentaenoic acid, and docosahexaenoic acid metabolism by LOX and COX enzymes is depicted. Prostaglandins are inflammatory, while specialized pro-resolving mediators resolve inflammation.