Sex-mediated elevation of the specialized pro-resolving lipid mediator levels in a Sjögren’s syndrome mouse model

Abstract

Our previous results showed that the specialized pro-resolving mediator (SPM) Resolvin D1 (RvD1) promotes resolution of inflammation in salivary glands in non-obese diabetic (NOD)/ShiLtJ, a mouse model for Sjögren’s syndrome (SS). Additionally, mice lacking the RvD1 receptor ALX/FPR2 show defective innate and adaptive immune responses in salivary glands. Particularly, ALX/FPR2 KO mice exhibit exacerbated inflammation in their salivary glands in response to systemic LPS treatment. Moreover, female ALX/FPR2 KO mice show increased autoantibody production and loss of salivary gland function with age. Together, these studies suggest that an underlying SPM dysregulation could be contributing to SS progression. Therefore, we investigated whether SPM production is altered in NOD/ShiLtJ using metabololipidomics and enzyme-linked immunosorbent assay (ELISA). Our results demonstrate that SPM levels were broadly elevated in plasma collected from NOD/ShiLtJ female mice after disease onset, whereas these drastic changes did not occur in male mice. Moreover, gene expression of enzymes involved in SPM biosynthesis were altered in submandibular glands (SMG) from NOD/ShiLtJ female mice after disease onset, with 5-LOX and 12/15-LOX being downregulated and upregulated, respectively. Despite this dysregulation, the abundances of the SPM products of these enzymes (ie, RvD1 and RvD2) were unaltered in freshly isolated SMG cells suggesting that other cell populations (eg, lymphocytes) may be responsible for the overabundance of SPMs that we observed. The elevation of SPMs noted here appeared to be sex mediated, meaning that it was observed only in one sex (females). Given that SS primarily affects females (roughly 90% of diagnosed cases), these results may provide some insights into the mechanisms underlying the observed sexual dimorphism.

1 |. INTRODUCTION

Sjögren’s syndrome (SS) is an autoimmune disorder where salivary gland function is progressively diminished.¹ This disease affects 0.1%−0.6% of the population² with a nine times greater prevalence in females.^3,4 This disease has been classified as either primary when it occurs by itself or secondary when it is present along with other diseases.⁵ SS is characterized by immune cell infiltration into the salivary glands leading to destruction of the normal tissue architecture and secretory dysfunction which is often associated with oral thrush, periodontitis, and caries.⁶ Current treatment regimens include saliva substitutes as well as secretory agonists both of which are only palliative as there is no cure for this disease.⁷ Moreover, while both genetic as well as environmental factors have been implicated as possible etiological factors, the cause for SS is still unknown.⁸ Nonetheless, a hallmark of the disease is chronic inflammation in the salivary glands, a process that is not completely understood.²

Uncontrolled inflammation has been recognized as a unifying component in many chronic diseases including metabolic syndrome,⁹ neurological diseases,¹⁰ neuropathic pain,¹¹ and more recently in SS.^12–15 Specialized pro-resolving lipid mediators (SPMs) limit uncontrolled inflammation while promoting natural healing processes by blocking inflammatory cytokine signaling and stimulating tissue repair.^14,16,17 Among known members of the SPM family are resolvins (Rv) including the E series (Resolvin E1–3 [RvE1–3], derived from eicosapentaenoic acid or EPA), D series (Resolvin D1 [RvD1]-6, derived from docosahexanoic acid or DHA and [AT-RvD 1–5], which are their aspirin-triggered [AT] forms), maresins (MaR1 and MaR2), lipoxin A₄ (LXA₄) and lipoxin B₄ (LXB₄), epi-lipoxin, 15-epi-LXA₄ and 15-epi-LXB₄, and protectin D1 (PD1).^12,18–20 The biosynthesis and stereochemistry of each SPM have been elucidated^21–26 with potent actions demonstrated in many tissues, including but not limited to airways,²⁷ vasculature,²⁸ skin,²¹ conjunctiva,^29,30 bone,³¹ lung,³² heart,³³ and periodontium^34–36 showing expression of these pathways throughout the body.

Our previous studies demonstrated that RvD1 and AT-RvD1 prevent TNFα-mediated disruption of salivary gland epithelial integrity while promoting survival in rat parotid Par-C10 cells³⁷ as well as in primary mouse submandibular glands (SMG) cells.³⁸ Moreover, in vitro studies showed that RvD1 biosynthetic pathways are expressed and active in primary mouse SMG cells as well as in human minor salivary glands.³⁸ Specifically, RvD1 is produced from DHA by the action of 12/15-LOX and 5-LOX via the generation of the intermediary 17 hydroxy-docosahexaenoic acid (17-HDHA) metabolite and its oxygenation. A summary of these pathways is shown in Figure 1.

FIGURE 1.

Biosynthetic pathways of specialized pro-resolving mediators in salivary glands. Arachidonic acid (AA), docosahexaenoic acid (DHA), and eicosapentaenoic acid (EPA) are released from the cell membrane by the action of the enzyme phospholipase A2 (PLA₂). Fatty acids are subsequently oxygenated and converted into their respective intermediate metabolites with further epoxidation of the intermediate metabolite leading to the production of each specialized pro-resolving mediator

Studies using the non-obese diabetic (NOD)/ShiLtJ SS mouse model indicate that treatment with AT-RvD1 before disease onset maintains the integrity and function of the salivary gland epithelium thereby preventing the progression of the disease in vivo.^39,40 Treatment with AT-RvD1 after disease onset restores salivary gland function and causes a reduction of Th17 cell infiltration in both female and male mice.⁴¹ Mice lacking the RvD1 receptor ALX/FPR2 display impaired innate and adaptive immune responses in salivary glands (exacerbated inflammatory reaction and increased apoptosis) in response to systemic LPS treatment.⁴² Finally, female ALX/FPR2 KO mice show altered antibody production and a progressive SS-dependent loss of salivary gland function.⁴³ Together, these results suggest that an underlying SPM dysregulation could be contributing to SS progression. Therefore, we investigated, in this study, whether SPM production is altered in the NOD/ShiLtJ SS mouse model.

2 |. MATERIALS AND METHODS

2.1 |. Animals

Mouse studies were performed under protocols approved by the University of Utah Institutional Animal Care and Use Committee. NOD/ShiLtJ SS-like mouse model at 7, 16, and 20 weeks of age were used for this study. The animals were fed Irradiated Teklad Global Soy Protein-Free Extruded Rodent Diet (catalog number: 2920X, Envigo, Indianapolis, IN) with the following lipid content: Total fat—6.5%, Total Saturated—0.8%, Total Monosaturated—1.1%, Total Polyunsaturated—2.9%, C16:0 Palmitic—0.6%, C18:0 Stearic—0.1%, C18:1ω9 Oleic—1.1%, C18:2ω6 Linoleic—2.6%, and C18:3ω3 Linolenic—0.3%.

2.2 |. Saliva collection

Mice were anesthetized with ketamine/xylaxine (100 mg/kg ketamine and 5 mg/kg xylazine), and injected with pilocarpine-HCl/PBS (Sigma) at 10 mg/kg via i.p to stimulate saliva secretion. Then, saliva was collected using a 200 μL pipette and placed immediately on ice in the presence of a protease inhibitor cocktail (Sigma).

2.3 |. Plasma collection

Mice were euthanized with CO₂ followed by cervical dislocation. Blood samples were collected via abdominal exsanguination into heparin-coated tubes (BD Microtainer #365965, BD Biosciences). The blood was then centrifuged at 14 000 RPM for 10 minutes at 4°C to obtain supernatant plasma samples.

2.4 |. Targeted metabololipidomics

Targeted metabololipidomics using liquid chromatography-tandem mass spectrometry (LC-MS/MS) was performed to quantify levels of the pro-resolving SPMs (D-series resolvins, E-series resolvins, protectins, and maresins) and the pro-inflammatory eicosanoids as described previously.⁴⁴ Briefly, specimens were placed in 1 mL of ice-cold methanol (containing deuterium-labeled internal standards [d4-LTB₄, d4-PGE₂, d5-LXA₄, d5-MaR2, d5-RvD2, d8–5-HETE; 500 pg each]) as shown in Table S1 to facilitate quantification and calculation of recovery. Samples were held at −80°C for 1 hour to allow for protein precipitation and centrifuged. Supernatants were collected and concentrated to less than 1 mL of methanol content under nitrogen flow in an automated evaporation system (TurboVap LV, Biotage). SPMs were extracted using solid-phase C18 columns, brought to dryness using the TurboVap and immediately resuspended in methanol-water (50:50 v/v) for automated injected into an LC-MS/MS system comprised of a Shimadzu Nexera XR liquid chromatography system (Shimadzu, Kyoto, Japan) hyphenated to a SCIEX QTRAP 6500 mass spectrometer (Sciex, Framingham, MA) equipped with an Ion Drive Turbo V ESI source. A Poroshell 120 EC-C18 (4.6×100 mm, 2.7 micron particle size) column (Agilent Technologies, Santa Clara, CA, USA) guarded by a Security Guard Ultra C18 precolumn (Phenomenex, Torrance, CA, USA) were used for HPLC separation, see Table S2. After 2 minutes at 50% of solvent B (A: 0.01% acetic acid in water; B: 0.01% acetic acid in methanol), analytes were separated by a linear gradient up to 80% of solvent B over 9.0 minutes and the concentration of solvent B was maintained at 80% for 3.5 minutes. The column was subsequently washed for 5.5 minutes at 98% B and re-equilibrated with 80% of solvent A over 3.0 minutes. The flow rate was 0.5 mL/min and the column were heated to 50°C. To monitor and quantify the levels of each lipid mediator, a multiple reaction method (MRM) was developed with signature ion fragments for each molecule, see Table S3. Confirmatory identification was conducted using published criteria of at least six diagnostic fragment ions.⁴⁵ Calibration curves for each analyte were obtained at 0.1, 1, 10, and 100 pg using injection volumes of 1, 2, and 5 μL with r² values in the range .98-.99. Quantification was carried out based on peak area of the MRM transition and the linear calibration curve for each compound.

2.5 |. Primary SMG cell isolation

Both female and male NOD/ShiLtJ mice at 7 weeks of age (pre-disease stage) and 12 weeks post-disease stage) were euthanized with CO₂ and their SMG were dissected as previously described.³⁹ Then, glands were dissociated using a GentleMACS Tissue Dissociator (Miltenyi Biotec Inc, Sand Diego, CA) followed by enzymatic digestion with a commercially available mouse tumor dissociation kit (#130–096-730, Miltenyi Biotec Inc, San Diego, CA) with a subsequent filtration step using a 70 μm of cell strainer as previously described.³⁷ Next, freshly isolated SMG cell suspensions were cultured in defined DMEM-Ham’s F12 (1:1) medium containing 2.5% of fetal bovine serum, 0.1 μM of retinoic acid, 80 ng/mL of epidermal growth factor, 2 nM of triiodothyronine, 5 mM of glutamine, 0.4 μg/mL of hydrocortisone, 5 μg/mL of insulin, 5 μg/mL of transferrin, 5 ng/mL of sodium selenite, and 50 μg/mL of gentamicin. Finally, cells were cultured in 96-well plates at 37°C in a humidified atmosphere of 5% of CO₂.

2.5.1 |. Enzyme-linked immunosorbent assay

SMG freshly isolated cells were treated with and without DHA at (0–200 μM, #90310, Cayman Chemical, Ann Arbor, MI, USA) and with or without 17-HDHA (10 μM, #33650, Cayman Chemical, Ann Arbor, MI, USA). Then, cell culture supernatants were collected at various time points (0–48 hours) and RvD1 and RvD2 levels quantified using Enzyme-linked immunosorbent assay (ELISA) kits (#500380 and #501120, respectively, Cayman Chemical, Ann Arbor, MI, USA) according to manufacturer’s instructions. Note that the manufacturer indicates no or very little cross-reactivity between RvD1 and RvD2 with DHA (less than 0.05% and less than 0.01%, respectively). Moreover, results were similar when using defined medium (eg, DMEM/F12 with 2.5% FBS all the supplements as described above) or complete medium (eg, DMEM/F12 with 2.5%).

2.5.2 |. Gene expression

Briefly, frozen SMG were lysed, and mRNA obtained using the using Qiagen RNeasy microkit (QIAGEN Inc, Valencia, CA). Then, cDNA was prepared using Biorad iScript Reverse Transcription Supermix (Bio-Rad, Hercules, CA) and was diluted at 1:50 ratios and used for qPCR. The reactions were then carried out by adding the following reagents: 2.5 μL of each primer listed in Table 1, stock 10 μM (final primer concentration of 1.25 μM), 5 μL of 1:50 cDNA dilutions and 10 μL of 2X SYBR Green master mix (Bio-Rad, Hercules, CA). Finally, qPCR experiments were performed on 96 well plates, relative fold changes of gene expression were normalized using β-actin as previously described⁴² using Biorad CFX Connect Real-Time PCR Detection System.

TABLE 1.

List of primers used in this study

Gene	FW	RV
β-actin	GTAACAATGCCATGTTCAAT	CTCCATCGTGGGCCGCTCTAG
5-LOX	ACTACATCTACCTCAGCCTCATT	GGTGACATCGTAGGAGTCCAC
5-LOXAP	GCCGGACTGATGTACCTGTT	CGCTTCCGAAGAAGAAGATG
12/15-LOX	ACCCCACCGCCGATTTT	AGCTTCGGACCCAGCATTT
ALX/FPR2	GCCAGGACTTTCGTGGAGAGAT	GATGAACTGGTGCTTGAATCACT
GPR18	ACAGTCTCCTCCGAGGGCAG	TGTGGAAGGGCACGAAGCAG

2.6 |. Western blot analyses

Total proteins from freshly dissected mouse SMG were obtained using T-PER™ Tissue Protein Extraction Reagent (Thermo Fisher Scientific) containing a cocktail of protease inhibitors (Sigma) and dissociated with a Fisher Scientific Sonic Dismembrator (model FB-120; microtip; output level, 5; duty cycle, 50%; Thermo Fisher Scientific). Samples were centrifuged at 14 000 r.c.f. for 10 minutes to remove cellular debris. Then, samples were denatured by boiling in SDS sample buffer (Bio-Rad) 5 minutes. Fifty μg of total protein was loaded in each lane of a gradient 4%−15% SDS-PAGE gel (Bio-Rad), and then, the electrophoresed samples were transferred to a nitrocellulose membrane (Bio-Rad). Subsequently, membranes were blocked for 1 hour at RT using 5% of BSA in Tris-buffered saline [0.137 M NaCl, 0.025 M Tris (hydroxymethyl)-aminomethane, pH 7.4] containing 0.1% of Tween-20 (TBST) and immunoblotted overnight at 4°C with the following primary antibodies in TBST containing 5% of BSA: mouse-anti-β-actin (Abcam; Cambridge, MA) 1:2500 dilutions, rabbit-anti-FPR2 (Thermo Fisher Scientific) 1:1000 dilutions and rabbit anti-G protein-coupled receptor 18 (GPR18) (Abcam; Cambridge, MA) 1:1000 dilutions. After incubation, membranes were washed three times for 5 minutes each with TBST and incubated with peroxidase-linked goat-anti-rabbit or goat-anti-mouse IgG antibody (Cell Signaling Technology) diluted 1:5000 in TBST containing 5% of BSA for 1 hour at RT. The membranes were washed three times for 5 minutes each with TBST and treated with a Bio-Rad Clarity detection reagent (Bio-Rad). The protein bands were visualized using Bio-Rad ChemiDoc MP imager and band intensities were quantified using ImageJ software. β-actin was used for protein normalization and the relative protein expression level was presented as the ratio of ALX/FPR2 or GPR18 to actin.

2.7 |. Statistical analyses

Data are means ± SD of results from at least three experiments. Then, statistical comparison between two groups was assessed using Student’s t test where P values <.05 represent significant differences.

3 |. RESULTS

Our previous work demonstrated that systemic treatment with AT-RvD1 prevents disease onset in the NOD/ShiLtJ mice.³⁹ Moreover, a similar treatment delivered after disease onset, diminished SS-like features.⁴¹ Together, these studies indicate a functional role for SPM biosynthesis in SS pathogenesis. To determine whether alteration of SPM production occurs in the NOD/ShiLtJ mouse model, we detected SPM levels in plasma and saliva collected from pre-disease and post-disease female and male mice using LC-MS/MS-based metabololipidomics.

We observed that all DHA-derived SPMs had an approximately >9-fold increased level of production in plasma from post-disease female mice as compared to pre-disease female mice (Figure 2A) where specific SPM levels were significantly altered (eg, 17S-HDHA, MaR1, PD1, PDX, RvD1, RvD3–5) (Figure 2B). The only exception was RvD2, which showed an approximately 1.6-fold increase that did not reach significance thresholds (Figure 2B). EPA-derived SPMs had a >100-fold increases in plasma from post-disease female mice as compared to their pre-disease counterparts (Figure 2A) where specific SPM levels were statistically significantly upregulated (eg, 18 hydroxy-eicosapentaenoic acid [18-HEPE]) while RvE1 was not (Figure 2B). Similarly, all arachidonic acid (AA)-derived products had an approximately >40-fold increase in plasma from post-disease female mice as compared to their pre-disease counterparts (Figure 2A) where specific SPM levels were statistically significantly upregulated (eg, leukotriene B₄ [LTB₄], Prostaglandin D₂ [PGD₂], Prostaglandin E₂ [PGE₂], Figure 2B). Together, these results indicate that SPMs were broadly elevated in plasma of female NOD/ShiLtJ mice after disease onset. Interestingly, this SS-dependent SPM dysregulation was also sex-specific as SPMs in post-disease male mice did not show such broad elevations when compared to their pre-disease counterparts (Figure 2A). Particularly, among the DHA-derived SPMs, PD1 was increased by approximately 6-fold in plasma from post-disease male mice as compared to their pre-disease counterparts (not statistically different, Figure 2C). Moreover, Mar1, 17-HDHA, PDX, and RvD1 levels were relatively unchanged in plasma from post-disease male mice as compared to their pre-disease counterparts (not statistically different, Figure 2C), whereas 14-HDHA and RvD2 levels were significantly decreased in plasma from post-disease male mice as compared to their pre-disease counterparts (Figure 2C). Regarding the EPA-derived SPMs, 18-HEPE demonstrated a modest 1.6-fold increase in abundance in post-disease male mice as compared to their pre-disease counterparts, whereas the levels of RvE1 were exhibited an approximately 10-fold reduction in post-disease male mice, (Figure 2A) however, there were no significant differences (Figure 2C). Moreover, all the AA-derived products showed only a modest increase in post-disease male mice as compared to their pre-disease counterparts with LTB₄ and LXB₄ showing an approximately a 10-fold increase, whereas PGD₂ and PGE₂ demonstrated a 5-fold increase in post-disease male mice (Figure 2A) with no significant differences observed (Figure 2C).

FIGURE 2.

Broad elevation of lipid mediators occurred in female NOD/ShiLtJ mice after disease onset. Serum samples were collected from pre-disease (7 weeks old) and post-disease onset (20 weeks old) female and male NOD/ShiLtJ mice. Lipid mediator levels were then analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Lipid mediator levels detected in post-disease onset mice were normalized relative to pre-disease sex-matched mice to demonstrate age-related dysregulation in serum SPM levels (A). Individual Lipid mediators detected in serum and saliva samples obtained from female (B) and male (C) mice pre- and post-disease onset. Lipid mediator levels in saliva samples obtained pre- and post-disease onset from female and male mice were normalized relative to their pre-disease counterparts (D). Identification of lipid mediators in blood serum of female mice. The figure shows LC-chromatograms (left) and the corresponding product ion spectra (right) from a representative female mouse specimen together with the structural formula of RvD1 (A), RvD2 (B), PD1 (C), and Mar1 (D). Lipid mediators were identified, quantified, and matched using the retention times and MSMS-spectra of pure synthetic standards together with at least six diagnostic ions of the product ion spectra collected from each blood serum sample (E). A total of five mice were used for each experimental group with *P < .05 indicating a significant difference from 7 weeks old controls

In saliva, the DHA-derived SPMs, RvD1, and RvD2 exhibited an approximately 1.5-fold increase in post-disease female mice as compared to their pre-disease counterparts while additional DHA-derived SPMs were undetectable (Figure 2D). Interestingly, only 14-HDHA, RvD2 and RvD3 were significantly increased (Figure 2B). In contrast, males SPM levels in saliva were not significantly different between post-disease as compared to the pre-disease counterparts (Figure 2D). Regarding AA-derived products, LTB₄, LXB₄, PGD₂, and PGE₂ showed approximately a 2-fold increase in post-disease female mice as compared to their pre-disease counterparts; however, these differences were not statistically significant (Figure 2D). In males, LTB₄ and LXB₄ were approximately 0.3-fold decreased in post-disease mice as compared to pre-disease mice with no statistically significant changes while PGD₂ and PGE₂ could not be detected (Figure 2D). Finally, representative MRM chromatograms of plasma and saliva samples and selected AA- and DHA-derived metabolites at the two time points together with the quantification and recovery of analytes are shown in Figure 2E. Collectively, these results indicate that SPM levels are broadly elevated in plasma and saliva from NOD/ShiLtJ female mice after disease onset, whereas these drastic changes are not observed in male mice.

As indicated above, our previous studies demonstrated that intravenous treatment with AT-RvD1 in female NOD/ShiLtJ mice either before or after disease onset reverses SS-like features.^39,41 Interestingly, in the current study we observed that levels of RvD1 are elevated in both plasma and saliva collected from post-disease female mice as compared to their pre-disease counterparts. Despite of this elevation, however, inflammation persisted. Specifically, it is possible that this disease persists through a yet to be discovered deregulated adaptive immune response, with antibody production, and as such the endogenous pathway centered on SPMs is unable to overcome this aberrant immune activation. Furthermore, inflammation was controlled by introduction of additional RvD1 where there was already an apparent excess. This raised the question of whether NOD/ShiLtJ mouse SMG had a defect in endogenous resolvin production that would limit their ability to mediate a resolution of inflammation. To address this, we investigated whether RvD1 biosynthesis was altered in primary SMG cell cultures derived from the NOD/ShiLtJ SS mouse model before disease onset. To that end, we isolated SMG cells from pre-disease female and male NOD/ShiLtJ mice and measured RvD1 production in response to addition of its primary precursor DHA. Both female and male SMG cells were capable of producing RvD1 in response to DHA in a dose dependent manner where a maximal RvD1 production occurred at 200 μM (Figure 3A). These results indicate that the RvD1 biosynthetic machinery appears to be intact in SMGs of pre-disease NOD/ShiLtJ mice.

FIGURE 3.

RvD1 production occurred in both female and male submandibular gland cells in a dose dependent manner. Freshly isolated submandibular gland cells from 7 weeks old pre-disease male and female NOD/ShiLtJ mice were obtained by mechanical disruption and enzymatic digestion as described in Materials and Methods. Subsequently, cells were incubated with various concentrations of DHA and after 24 hours, cell culture supernatants were collected followed by RvD1 (A) and RvD2 (B) detection using ELISA. Data were compiled from three separate experiments and presented as mean ± SD

Since RvD1 biosynthesis was not affected in SMG from pre-disease NOD/ShiLtJ mice, we investigated whether biosynthesis of other D-series resolvins were altered. To this end, we isolated SMG cells from pre-disease female and male NOD/ShiLtJ mice and measured RvD2 production in response to DHA. Both female and male SMG cells were capable of producing RvD2 in response to DHA (Figure 3B). These results indicate that the RvD2 like the RvD1 biosynthetic machinery in SMGs appears to be intact in pre-disease NOD/ShiLtJ mice.

In order to determine whether RvD1 production increases over time in pre-disease SMG cells we measured the time course of RvD1 production. As shown in Figure 4A,B, both female and male SMG cells produced RvD1 in response to DHA starting at 2 hours and maintaining this production at steady levels for the period studied (ie, 48 hours). These results suggest that the biosynthetic pathway for RvD1 in vitro is intact in SMG cells derived from pre-disease NOD/ShiLtJ mice.

FIGURE 4.

RvD1 production occurred in both female and male SMG cells in a time dependent manner. Freshly isolated submandibular gland cells from 7 weeks old pre-disease male and female NOD/ShiLtJ mice were obtained by mechanical disruption and enzymatic digestion as described in Materials and Methods. Then, cells were incubated with DHA (100 μM) or 17-HDHA (10 μM) at various time points and cell culture supernatants were collected followed by RvD1 detection using ELISA. Shown are the time-course of RvD1 production obtained in females (A) and males (B). Data were compiled from three separate experiments and presented as mean ± SD

Since female mice exhibited a broad dysregulation of SPMs after disease onset as compared to their pre-disease counterparts, we investigated whether their SMGs were defective in their ability to synthesize SPMs. Therefore, we detected RvD1 and RvD2 production in SMGs collected from pre-disease and post-disease female NOD/ShiLtJ SMGs in response to DHA or 17-HDHA stimulation. Surprisingly, we observed no significant differences in RvD1 production: cells isolated from the two mice groups produced similar RvD1 levels at 2 (Figure 5A) and 24 hours (Figure 5C). Likewise, no differences were observed in RvD2 levels at the same time points (Figure 5B,D). These results suggest that the machinery to biosynthesize either RvD1 or RvD2 in vitro is intact in SMG cells from female mice after disease onset.

FIGURE 5.

RvD1 and RvD2 biosynthetic machinery was intact in SMG cells from female mice. Freshly isolated submandibular gland cells from 7 weeks old pre-disease and 16 weeks old post-disease onset female NOD/ShiLtJ mice were obtained by mechanical disruption and enzymatic digestion as described in Materials and Methods. Then, cells were incubated with DHA (100 μM) or 17-HDHA (10 μM) and supernatants collected at 2 hours (A, B) and 24 hours (C, D) followed by RvD1 (A, C) and RvD2 (B, D) detection using ELISA. Data were compiled from three separate experiments and presented as mean ± SD

Since the biosynthetic RvD1 and RvD2 machinery appeared to be intact in NOD/ShiLtJ mice, we verified the gene expression levels of the enzymes involved in their biosynthesis using SMGs from both sexes at pre-disease state as well as after disease onset. Our results showed that SMGs from post-disease female mice had an upregulation of 12/15-LOX expression while 5-LOX was downregulated as compared to their pre-disease counterparts (Figure 6A). In males, no significant differences in the expression of these biosynthetic enzymes were observed between the post-disease and pre-disease groups (Figure 6A). Together, these results indicate that expression of SPM-biosynthetic enzymes is dysregulated in SMG of females but not male NOD/ShiLtJ mice after disease onset.

FIGURE 6.

Dysregulation of SPM biosynthetic enzymes and SPM receptors occurred in female NOD/ShiLtJ mice after disease onset. mRNA was isolated from whole SMG (A) and SMG cells in culture (B) from 7 weeks old pre-disease and 16 weeks old post-disease male and female NOD/ShiLtJ mice as described in Materials and Methods. Gene expression levels of SPM biosynthetic enzymes and SPM cell surface receptors were subsequently analyzed via qRT-PCR using β-actin for normalization. Data were compiled from four separate experiments and presented as mean ± SD where *P < .05 indicates a significant difference from pre-disease controls. C, SMG whole tissue lysates from NOD/ShiLtJ mice at pre-disease and post-disease stages were obtained as described in Materials and Methods and processed for ALX/FPR2 and GPR18 detection using Western blot analysis. Data are representative of three or more experiments and presented as mean ± SD where *P < .05 indicates a significant difference from pre-disease controls

Previous studies showed that AT-RvD1 treatment improved saliva secretion and reduced salivary gland inflammation in female mice after disease onset.⁴¹ This raised the question whether alterations in SPM receptor expression cause a need for additional amounts of RvD1 for receptor activation. Using qPCR as well as Western blots we were able to show that the RvD1 receptor ALX/FPR2 and the RvD2 receptor GPR18 were significantly upregulated in SMG collected from post-disease female mice as compared to their pre-disease counterparts at both gene expression and protein levels (Figure 6A,C). However, no significant differences were observed in ALX/FPR2 or GPR18 expression in the male counterparts (Figure 6A,C). These results could explain why additional RvD1 amounts enhance the resolution of inflammation in female mice after disease onset. Interestingly, these sex-dependent changes in SPM biosynthetic enzymes and receptor expression in post-disease female mice were only seen in mRNA isolated from whole SMG tissue. No differences in SPM biosynthetic enzymes and receptor expression were seen when SMG cells were cultured in vitro (Figure 6B). These results further highlight that SPM biosynthetic machinery is not inherently defective in post-disease female mice.

4 |. DISCUSSION

We demonstrated that SPM production is elevated in plasma and saliva collected from female NOD/ShiLtJ mice after disease onset but not in their male or younger female counterparts (Figure 2A–D). Our results are consistent with previous studies showing SPM elevation in inflammatory diseases. RvD1 levels are elevated in plasma in mice infected with Trypanosoma cruzi⁴⁶ and in synovial fluid of osteoarthritis patients.⁴⁷ LXA₄ levels are increased in plasma collected from humans infected with Toxoplasma gondii,⁴⁸ and in exudates condensates of asthma patients⁴⁹ and in leukocytes and kidneys of patients with post-streptococcal glomerulonephritis.⁵⁰ However, other studies have shown a marked decrease in SPM levels during disease pathology. Specifically, RvD1 and LXA₄ levels are reduced in cerebrospinal fluid and hippocampus of Alzheimer’s disease patients⁵¹ and RvD3 plasma levels are decreased in a murine model of arthritis.⁵² Moreover, a reduced capacity to biosynthesize lipoxins, maresins, protectins, and resolvins was observed in cystic fibrosis patients.⁵³ Finally, RvD1 levels are decreased in plasma samples derived from patients suffering from systemic lupus erythematosus.⁵⁴ Together, these studies support the notion that SPM dysregulation is a common feature among inflammatory diseases. The sex dimorphism that we observed regarding SPM dysregulation in the NOD/ShiLtJ mouse model after disease onset could be further explored by investigating which individual cell populations may be responsible for this alteration and by validating these results in minor salivary glands biopsies obtained from patients diagnosed with SS. Additional evidence that sex hormones may be involved comes from the observation that castrated NOD/ShiLtJ male mice display SS-like clinical features similar to females while ovariectomized females show the opposite effect.^55,56 Testosterone treatment of female mice has been shown to reduce salivary gland inflammation.⁵⁷ Finally, Th17 cells from a C57BL/6.NOD-Aec1Aec2 SS-like mouse model show increased proliferation only in females.⁵⁸ Together, these studies indicate a potential role for sex hormones in SS that could be further explored by quantifying SPMs in castrated and ovariectomized mice.

The RvD1 and RvD2 biosynthetic machinery appeared to be intact in SMG in NOD/ShiLtJ mice at all stages (Figures 3–5). Specifically, production of these SPMs was elicited by DHA in a dose and time dependent manner. Moreover, SPM generation levels were similar to those observed in healthy C57BL/6 mice (Figure S1). However, RvD1 levels in NOD/ShiLtJ female mice after disease onset were elevated in plasma and to a lesser extent in saliva. Together, these results indicate that SMG cells appear to not be responsible for the increased RvD1 levels observed in this group and suggest that other cell types may be responsible for the dysregulation. Particularly, it is possible that immune cells from NOD/ShiLtJ female mice may have a specific defect in SPM production, while this has not been reported in SS, such defects have been observed in other autoimmune diseases. For instance, alveolar macrophages isolated from severe asthma patients show reduced LXA₄ and increased LTB₄ production.⁵⁹ Furthermore, eosinophils derived from chronic rhinosinusitis patients had a reduced biosynthesis of 12/15-LOX-derived mediators.⁶⁰ Finally, eosinophils derived from severe asthma patients showed a reduction in PD1 biosynthesis.⁶¹ Future studies to investigate the biosynthetic capabilities of specific cell populations within NOD/ShiLtJ mouse SMGs as well as human minor salivary glands are warranted.

We detected an increase in the expression of ALX/FPR2 and GPR18 receptors for RvD1 and RvD2, respectively. While these results may suggest that more receptors are available for binding by their specific ligands, we did not measure actual receptor binding nor activation of specific downstream intracellular pathways or receptor internalization as such. However, previously we have shown that AT-RvD1 treatment blocks the progression of SS-like features in pre-disease mice and restores saliva secretion if given at disease onset, suggesting that ALX/FPR2-mediated signaling must be active in these mice.³⁹

A limitation of this experiment is the likelihood of a different situation when the treatment occurs during in vivo conditions. Particularly, previous studies demonstrated that intravenous AT-RvD1 treatment of NOD/ShiLtJ at disease onset decreases SMG inflammation while restoring saliva secretion.^39,41

Regarding the differential expression of SPM-biosynthetic enzymes, our results are consistent with previous studies that show that the activity of 5-LOX and 12/15-LOX are influenced by sex hormones.^62,63 Moreover, we observed an SS-dependent difference in the expression of 5-LOX and 12/15-LOX between female NOD/ShiLtJ mice after disease onset as compared their young counterparts. Our results show that even though the in vitro production of RvD1 was not compromised in SMG cells derived from female mice after disease onset, under in vivo conditions, these mice demonstrated sex-dependent dysregulation of SPM receptors and SPM biosynthetic enzymes. In summary, there was a sex-dependent dysregulation of SPMs, their receptors as well as their biosynthetic enzymes in female but not male NOD/ShiLtJ mice after disease onset.

One of the primary clinical features of SS is that the disease is closely linked to sex, with roughly 90% of cases being diagnosed among females^3,4; however, no underlying sexual differences have been identified that might account for the degree to which females are more vulnerable to the disease. Specifically, only vague and largely descriptive references to hormonal differences have been offered to date.^64–66 The data shown here offers an indication of potential underlying mechanisms that may help shed light on the vastly differing disease patterns between sex. Nonetheless, caution should be exercised in interpreting these results (ie, dysregulation of SPM levels and gene expression of enzymes involved in SPM biosynthesis in females post disease onset), with further investigations warranted to replicate these findings to better understand their implications for determining SS pathogenesis.

Finally, it is worth noting that the NOD/ShiLtJ mouse line is primarily a model of autoimmune Type 1 diabetes and further studies are needed to establish the potential implication of sex-specific changes in plasma SPM levels in disease pathogenesis.

Abbreviations:

17-HDHA

17 hydroxy-docosahexaenoic acid

18-HEPE

18 hydroxy-eicosapentaenoic acid

5-LOXAP

5-LOX-associated protein

AT

aspirin-triggered

DHA

ocosahexanoic acid

ELISA

enzyme-linked immunosorbent assay

EPA

eicosapentaenoic acid

FPR2

formyl peptide receptor 2

GPR18

G protein-coupled receptor 18

LOX

lipoxygenase

LTB₄

leukotriene B₄

LXA₄

lipoxin A₄

LXB₄

lipoxin B₄

MaR1

maresin R1

MaR2

maresin R2

NOD

non-obese diabetic

PD1

protectin D1

PGD₂

prostaglandin D₂

PGE₂

prostaglandin E₂

RvD1

resolvin D1

RvE1–3

resolvin E1–3

SMG

submandibular glands

SPM

specialized pro-resolving mediator

SS

Sjögren’s syndrome