Effects of chemosynthetic choline plasmalogens on gonadotropin secretion from bovine gonadotrophs

Abstract

Ethanolamine plasmalogens (EPls) and choline plasmalogens (CPls), unique glycerophospholipids may play important roles in milk production and reproduction in postpartum dairy cows. While CPls are more abundant in bovine blood, EPls are predominant in the brain. Brain EPls are the only recognized ligands of G protein-coupled receptor 61 (GPR61), a receptor that co-localizes with GnRH receptors on gonadotrophs. We hypothesized that chemosynthetic CPls stimulate gonadotropin secretion from bovine gonadotrophs, similar to the reported effects of chemosynthetic EPls. Anterior pituitary cells from healthy, post-pubertal heifers, were cultured for 3.5 days and then treated with increasing concentrations (0, 0.7, 7, 70, or 700 pM) of EPl with vinyl-ether-bonded stearic acid and ester-bonded oleic acid (C18:0-C18:1EPl) as a positive control, or CPls with vinyl-ether-bonded stearic acid and ester-bonded oleic acid (C18:0-C18:1CPl), arachidonic acid (C18:0-C20:4CPl), or docosahexaenoic acid (C18:0-C22:6CPl). After 2 h, the medium samples were harvested for FSH and LH assays. C18:0-C18:1EPl (7–700 pM) stimulated basal FSH and LH secretion (P < 0.01). None of the tested CPl concentrations stimulated LH secretion. Only 700 pM of C18:0-C18:1CPl, but not lower concentrations, stimulated FSH secretion (P < 0.05), an effect that was inhibited by a SMAD pathway inhibitor. However, both C18:0-C18:1CPl and C18:0-C20:4CPl synergized with GnRH to stimulate FSH secretion. In silico molecular-docking simulations using the deep-learning algorithm ColabFold revealed that CPls bind to the three-dimensional structural model of GPR61. In conclusion, C18:0-C20:4CPl stimulated FSH secretion exclusively in the presence of GnRH, whereas C18:0-C18:1CPl weakly stimulated FSH secretion and showed potential interaction with the GnRH signaling pathways.

Ethanolamine plasmalogens (EPls) and choline plasmalogens (CPls) are unique glycerophospholipids. Brain EPls are the only recognized ligands of G protein-coupled receptor 61 (GPR61), a newly identified receptor that co-localizes with GnRH receptors (GnRHRs) in lipid rafts on the gonadotroph surface [1,2,3]. Even in the absence of GnRH, EPls extracted from the brains of young heifers, but not old cows, can stimulate the secretion of FSH and LH from bovine gonadotrophs via Sma and Mad (SMAD) pathway [3, 4]. Intriguingly, in the human brain, EPl levels diminish with age, and this decline induces several age-related diseases, including Alzheimer’s disease [2, 5]. Therefore, EPls may function as a molecular link in age-related infertility by acting on GPR61 in gonadotrophs.

EPls contain ethanolamine as the head group bonded to the glycerol backbone at the sn-3 position, whereas CPls contain choline as the head group bonded at the sn-3 position. Both EPls and CPls contain “two legs”: a fatty alcohol bonded to the sn-1 position through a vinyl-ether bond and a fatty acid bonded to the sn-2 position through an ester bond. Based on the various possible combinations of fatty alcohols and acids, the bovine brain contains at least 20 EPl molecular species [4], and specific combinations of EPls appear to be critical for their effects [3, 4]. Chemical synthesis of specific EPl or CPl molecular species is extremely challenging, and only one chemosynthetic EPl and a few chemosynthetic CPl are commercially available. Chemosynthesis of EPl: 1-(1Z-octadecenyl)-2-oleoyl-sn-glycero-3-phosphoethanolamine (syn. PLAPE or phosphatidylethanolamine-plasmalogen-oleic acid; C18:0-C18:1EPl) strongly stimulates both FSH and LH secretion, even in the absence of GnRH [6], and the amount of endogenous C18:0-C18:1EPl is higher in young brains than in old brains [4]. Blood concentrations of EPls and CPls change dramatically around parturition and the first postpartum ovulation, correlating with important parameters of milk production and reproduction in dairy cows [7]. Therefore, blood plasmalogens may play an important role in postpartum dairy cows.

Concentrations of CPls are approximately 4.6-fold higher than those of EPls in cow blood [7], while the amounts of EPls are approximately 27.5-fold higher than those of CPls in cow brains [7]. These differences may be due to the differential regulation of plasmalogen biosynthesis in peripheral tissues and the brain [8], and blood CPls may be transformed into brain EPls [9]. Gonadotrophs in the anterior pituitary (AP) gland receive both central (e.g., GnRH) and peripheral regulation (e.g., ovarian hormones) [10]. However, no previous studies have investigated whether CPls stimulate FSH and LH secretion. Follicular fluid contains CPls, which may be indicators of ovarian aging [11]. Further studies are required to determine whether CPls affect FSH and LH secretion.

In this study, we tested the hypothesis that commercially available chemosynthetic CPls could stimulate FSH and LH secretion from bovine gonadotrophs. The activity of 1-(1Z-octadecenyl)-2-oleoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen-oleic acid; C18:0-C18:1CPl), 1-(1Z-octadecenyl)-2-arachidonoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen-arachidonoic acid; C18:0-C20:4CPl), 1-(1Z-octadecenyl)-2-docosahexaenoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen- docosahexaenoic acid; C18:0-C22:6CPl), and C18:0-C18:1EPl were used as positive controls. In cultured bovine AP cells, LH secretion is stimulated by increasing amounts of GnRH, with a peak at 0.1 nM or 1 nM GnRH [12]. However, excess GnRH (> 1 nM) suppresses LH secretion by bovine gonadotrophs [12]. Similarly, a high concentration of C18:0-C18:1EPl exerts a comparable suppressive effect on GnRH-induced FSH and LH secretion [6]. Therefore, we evaluated the effects of CPls on FSH and LH secretion in the presence and absence of GnRH.

In our previous study, we constructed a three-dimensional (3D) structural model of bovine GPR61 using a recently developed method with the highest accuracy, ColabFold (an online implementation of AlphaFold2) [13, 14]. The results of in silico molecular-docking simulations revealed the presence of three binding sites for C18:0-C18:1EPl, located in the extracellular, transmembrane, and cytoplasmic regions of GPR61. Therefore, we conducted molecular-docking simulations of the CPl molecular species with the binding sites of the predicted bovine GPR61 structural model constructed using ColabFold.

Materials and Methods

Animals

All experiments were performed in accordance with the Guiding Principles for the Care and Use of Experimental Animals in the Field of Physiological Sciences (Physiological Society of Japan) and were approved by the Committee on Animal Experiments of Yamaguchi University (approval number 301).

All cattle were managed by a contracted farmer in western Japan. The farm had open free-stall barns with free access to water. The cattle were fed twice daily with a total mixed ration, according to the Japanese feeding standard [15]. All cattle were non-lactating, non-pregnant, and had no follicular cysts, luteal cysts, or other ovarian disorders, as observed by macroscopic examination of the ovaries [16].

Pituitary cell culture and analysis of the effects of EPl and CPl on FSH and LH secretions

We obtained pituitary glands from post-pubertal (25.4 ± 0.5 months old, n = 8) Japanese Black heifers at a local abattoir. Briefly, heifers were stunned using a captive bolt pistol and subsequently exsanguinated by cutting the throat. The heads were placed on ice within 5 min of slaughter. Pituitary glands were isolated within 15 min of slaughter, immediately stored in HEPES (+) buffer (NaCl, 137 mmol/L; KCl, 3 mmol/l; Na₂HPO₄, 0.7 mmol/l; n-2-hydroxyethyl piperazine ethanesulfonic acid [HEPES], 25 mmol/l; glucose, 10 mmol/l; CaCl₂, 360 μmol/l; pH 7.2), and transported to the laboratory. The heifers were in the middle luteal phase, that is, 8 to 12 days after ovulation, as determined by macroscopic examination of the ovaries and uterus [17]; the endometrium was pink to red in color without mucus and stroma edema, while the corpus luteum was compact, soft, and 15–25 mm in diameter, with a luteal tissue surface, and the color on the inside was tan or orange. The AP glands show the highest LH, FSH, and GnRHR concentrations during this phase [18]. Blood samples could not be obtained because of farm and slaughterhouse regulations.

The large size of the bovine pituitary gland facilitates easy differentiation of the anterior lobe (light brown or pink color and hard portion) from the posterior and intermediate lobes (brown color and softer parts) following sagittal dissection along the midline [19]. AP cell dissociation and culture were performed as the previously described method [20]. Briefly, the removed AP glands were diced (1 mm³) using scissors and washed several times with HEPES (+) buffer. The AP tissues were then incubated with 0.4% collagenase (032-22364, FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) dissolved in HEPES (+) buffer containing 1% bovine serum albumin (010-25783, BSA; FUJIFILM Wako Pure Chemical Corporation) for 45 min at 37°C. After centrifugation (400 × g for 5 min), the supernatant was discarded, and the cell fraction was washed once in HEPES (−) buffer (NaCl, 137 mmol/l; KCl, 3 mmol/l; Na₂HPO₄, 0.7 mmol/l; HEPES, 25 mmol/l; glucose, 10 mmol/l; pH 7.2). The AP cells were then incubated with 0.25% pancreatin (from porcine pancreas, P3292, Sigma-Aldrich Inc., St. Lous, MO, USA) dissolved in HEPES (−) buffer for 10 min at 37ºC. After incubation, 10-ml Dulbecco’s Modified Eagle’s Medium (Thermo Fisher Scientific, Waltham, MA, US) plus serum (DMEMS), supplemented with 10% horse serum (16050-130, Thermo Fisher Scientific), 2.5% fetal calf serum (10437-028, Thermo Fisher Scientific), 1 × MEM nonessential amino acids (11140050, Thermo Fisher Scientific), 50 μg/ml streptomycin (32204-92, Nacalai Tesque, Kyoto, Japan), and 100 IU/ml penicillin (26239-42, Nacalai Tesque), was added to hamper the pancreatin reaction. After washing and centrifugation, the cell suspension was filtered through a fine nylon mesh (100-μm meshes, Asone, Osaka, Japan) into another plastic centrifuge tube, and the cell fraction was pelleted by centrifugation. After treating the cell fraction with lytic buffer (TRIzma base, 17 mM; NH₄Cl, 140 mM), the cells were washed three times with DMEMS. After the final wash, an aliquot of the cells was counted using a hemocytometer. Using this method, cell viabilities greater than 90% were obtained, as determined by trypan blue exclusion. The total cell yield was 19.7 × 10⁶ ± 0.8 × 10⁶ cells per pituitary gland. The dispersed cells were then suspended in DMEMS. The cells (2.5 × 10⁵ cells/ml, total 0.3 ml) were plated in 48-well culture plates (MS-80480, Sumitomo Bakelite, Tokyo, Japan) and maintained at 37ºC in a humidified atmosphere with 5% CO₂ for 3.5 days. Each experiment was repeated eight times with eight different pituitary glands, using four wells per treatment. Recombinant human activin A (final concentration, 10 ng/ml; 582-97904, R&D Systems, Minneapolis, MN, USA) was used to stimulate FSH synthesis 24 h before testing. Bovine activin A (National Center for Biotechnology Information [NCBI] reference sequence of bovine activin A is NP_776788.1) and ovine activin A (NP_001009458.1) have 100% sequence homology with that of humans (CAA40805.1), and a 24-h culture with the same concentration of recombinant human activin A stimulates FSH expression in cultured ovine AP cells [21].

To evaluate the effect of EPL and CPls in the absence of GnRH, the initial medium was replaced with 295-µl DMEM containing 0.1% BSA and 10 ng/ml activin A and incubated for 2 h. Treatment was conducted by adding 5-µl DMEM alone or 5-µl DMEM containing various concentrations (final concentrations of 0, 0.7, 7, 70, or 700 pM) of C18:0-C18:1EPl (852758P, Sigma-Aldrich Inc.), C18:0-C18:1CPl (852467C, Sigma-Aldrich Inc.), C18:0-C20:4CPl (852469C, Sigma-Aldrich Inc.), or C18:0-C22:6CPl (852472C, Sigma-Aldrich Inc.). After incubation for another 2 h, the medium from each well was collected for the FSH and LH radioimmunoassays (RIAs). This incubation time was selected because the expected FSH and LH concentrations were within the detectable range for each RIA and because the same treatment time was utilized in previous studies on EPl [3, 4, 6]. The precise physiological concentrations of EPls and CPls in ruminant blood were not determined in this study. However, in our previous studies, 500 pM bovine brain EPl demonstrated significant and strong effects on FSH and LH secretion from cultured bovine AP cells [3, 4]. C18:0-C18:1EPl (7–700 pM) stimulates basal FSH and LH secretion [6]. Therefore, these CPl concentrations were used in the present study. “GnRH” wells were incubated with 0.1 nM GnRH, whereas “Control” wells were incubated with DMEM only. After incubation for 2 h, the medium from each well was collected for FSH and LH RIAs.

To evaluate the effect of EPL and CPls in the presence of GnRH, AP cells obtained from a different set of post-pubertal Japanese Black heifers (n = 8, the middle of the luteal phase as determined by macroscopic examination, 26.1 ± 0.5 months old) were cultured in the medium described in the previous section for 3.5 days. Each experiment was repeated eight-fold with each of the eight pituitary glands, using four wells per treatment. Recombinant human activin A (final concentration) was used to stimulate FSH synthesis 24 h before testing. The initial medium was replaced with 290-µl DMEM containing 0.1% BSA and 10 ng/ml activin A and incubated at 37 °C for 2 h. Pretreatment was performed by adding 5-µl DMEM alone or 5-µl DMEM containing various concentrations (final concentration, 0, 0.7, 7, 70, or 700 pg/ml) of EPL and CPls. The cells were incubated while gently shaking for 5 min and then were treated with 5-µl 6 nM GnRH (Peptide Institute Inc., Osaka, Japan) (final concentration, 0.1 nM) dissolved in DMEM for 2 h to stimulate FSH and LH secretion.

Effects of SMAD pathway inhibitor on CPl-induced FSH secretion

We evaluated the effect of the SMAD pathway inhibitor LDN212854 (Cayman Chemical, Michigan, USA) on CPl-induced FSH and LH secretion from AP cells. AP cells obtained from a different set of post-pubertal Japanese Black heifers (n = 6, the middle of the luteal phase as determined by macroscopic examination, 26.1 ± 0.5 months old) were cultured in the medium described in the previous section for 3.5 days. Each experiment was repeated six-fold with each of the six pituitary glands, using four wells per treatment. Recombinant human activin A (final concentration) was used to stimulate FSH synthesis 24 h before testing. The initial medium was replaced with 290-µl DMEM containing 0.1% BSA and 10 ng/ml activin A for 2 h. Cells were pretreated with 5-µl DMEM alone or with 5-µl DMEM containing LDN212854 (1,000 nM, final concentration [3]). After 30 min of incubation, either 5-µl DMEM alone or 5-µl DMEM containing C18:0-C18:1CPl (final concentration, 700 pM, which demonstrated the stimulatory effect on the basal FSH secretion) was added to each culture well. The cells were incubated for 2 h, after which the medium was collected for RIAs.

RIAs to measure gonadotropin concentration in culture media

The FSH concentrations in the culture media were assayed in duplicate using double-antibody RIA with 125I-labeled bFSH, reference-grade bFSH, anti-oFSH antiserum (AFP5318C, AFP5346D, and AFPC5288113, National Hormone & Pituitary Program of the National Institute of Harbor-UCLA medical center Torrance, CA, USA), and goat anti-rabbit immunoglobulin (Shibayagi Co., Ltd., Gunma, Japan) to precipitate the first antibody–antigen complex. The limit of detection was 0.20 ng/ml. At 4.00 ng/ml, the intra- and interassay coefficients of variation (CV) were 4.2% and 7.4%, respectively.

The LH concentrations in the culture media were assayed in duplicate by double-antibody RIA using ¹²⁵I-labeled bLH, reference-grade bLH, anti-oLH-antiserum (AFP11118B, AFP11743B, and AFP192279), and goat anti-rabbit immunoglobulin. The limit of detection was 0.40 ng/ml. At 2.05 ng/ml, the intra- and interassay CV values were 3.6% and 6.4%, respectively.

Statistical analysis

Data were analyzed using StatView version 5.0 for Windows (SAS Institute, Inc., Cary, NC, USA). The Shapiro–Wilk’s W test or Kolmogorov–Smirnov Lilliefors test was used to evaluate the normality or log-normality of the distribution of each variable. All variables were normally distributed. The Grubb’s test verified the absence of outliers for any of the variables. Differences in LH or FSH concentrations were analyzed using one-factor analysis of variance with post hoc comparisons using the Tukey–Kramer test. The level of significance was set at P < 0.05. Data are expressed as raw mean ± SEM.

Structure modeling and docking simulation

In this study, all applications used were included in Schrödinger Release 2023–1 (Schrödinger, LLC, New York, NY, USA). We constructed a 3D structural model of bovine GPR61 using ColabFold [13] to perform molecular-docking simulations to clarify the binding sites and states of each of the aforementioned EPl or CPl molecular species (C18:0-C18:1EPl, C18:0-C18:1CPl, C18:0-C20:4CPl, or C18:0-C 22:6CPl) on GPR61. Briefly, the FASTA sequence of bovine GPR61 (NP_001033660) was used in conjunction with ColabFold to construct the target protein structure. After removing the unreliable regions appearing in the constructed model, the bond orders and hydrogenations for the GPR61 model were assigned using Maestro. Hydrogen-bond optimization was performed using PROPKA [22]. Energy minimization was conducted in Maestro using the OPLS4 force field [23]. The loop structure between transmembrane region 4 (TM4) and TM5 (D185–H201) was modified using the knowledge-based mode of Prime [24]. The number of output modified loops was set to 10, and clustering was performed using the root-mean-square distance of the heavy atoms belonging to the modified loops. The centroid structure of the cluster with the largest population was used for the docking simulations. The structures of the tested EPl and CPl molecular species were created using the 2D sketcher in Maestro. All ligands used for docking were converted to 3D ionized structures using LigPrep, and the same prepared structures were used in every calculation. For the docking simulation, the binding sites were detected using SiteMap [25, 26] and used as the centroid for the grid setting. Glide [27, 28] was used for docking simulations using an enhanced sampling mode and an expanded-sampling mode in an advanced setting. The maximum number of outputs for docking poses was set to 100. Furthermore, binding free‑energy analyses of the docking poses at each site were conducted with the molecular mechanics–generalized Born surface area (MM‑GBSA) method. An implicit‑membrane environment was assigned to the modeled GPR61 using Prime, after which binding free energies were predicted with Prime MM‑GBSA.

Results

EPl- or CPl-mediated stimulation of gonadotropin secretion from AP cells

In the absence of GnRH, treatment of AP cells with 7, 70, and 700 pM C18:0-C18:1EPl stimulated the basal secretion of both FSH and LH (Fig. 1A). C18:0-C18:1CPl (700 pM), but not lower concentrations, stimulated FSH secretion (Fig. 1B). The tested concentrations of C18:0-C18:1CPl (Fig. 1B), C18:0-C20:4CPl (Fig. 1C), and C18:0-C22:6CPl (Fig. 1D) stimulated LH secretion, unlike C18:0-C18:1EPl (Fig. 1A).

Fig. 1.

Effect of addition of numerous concentrations of a chemosynthetic ethanolamine plasmalogen (EPl) or three chemosynthetic choline plasmalogen (CPl) in medium lacking GnRH on hormone secretion from cultured anterior pituitary cells. The FSH and LH concentrations in control cells (cultured in medium lacking EPls, CPls, and GnRH) were averaged and set as 100%; the mean FSH and LH concentrations in each treatment group are expressed as percentages of the control value. Bars labeled with distinct letters (A, B, and C): significantly different stimulatory effects (P < 0.05); bars labeled with same letter: similar stimulatory effect. C18:0-C18:1EPl: 1-(1Z-octadecenyl)-2-oleoyl-sn-glycero-3-phosphoethanolamine (syn. PLAPE, or phosphatidylethanolamine-plasmalogen-oleic acid). C18:0-C18:1CPl: 1-(1Z-octadecenyl)-2-oleoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen-oleic acid). C18:0-C20:4CPl: 1-(1Z-octadecenyl)-2-arachidonoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen-arachidonoic acid). C18:0-C22:6CPl: 1-(1Z-octadecenyl)-2-docosahexaenoyl-sn-glycero-3-phosphocholine (syn. phosphatidylcholine-plasmalogen- docosahexaenoic acid).

Conversely, in the presence of GnRH, 700 pM of C18:0-C18:1CPl (Fig. 2B) and 70 pM of C18:0-C20:4CPl (Fig. 2C) acted synergistically with GnRH to stimulate FSH secretion.

Fig. 2.

Effect of addition of various concentrations of the chemosynthetic EPl or three CPl in medium containing GnRH on hormone secretion from cultured anterior pituitary cells. The FSH and LH concentrations in control cells (cultured in medium lacking EPls, CPls, and GnRH) were averaged and set as 100%; the mean FSH and LH concentrations in each treatment group are expressed as percentages of the control value. Bars labeled with distinct letters (A, B, and C): significantly different stimulatory effects (P < 0.05); bars labeled with same letter: similar stimulatory effect.

Figure 3 shows that 700 pM C18:0-C18:1CPl stimulated basal FSH secretion and that pretreatment with LDN212854 suppressed this stimulation.

Fig. 3.

Effect of the SMAD pathway inhibitor, LDN212854 on C18:0-C18:1CPl-induced FSH secretion from cultured AP cells of post-pubertal heifers. The mean FSH or LH concentration for each treatment group is expressed as a percentage of the control value. Different letters indicate statistically significant differences (P < 0.05).

In silico molecular docking

Figure 4 shows a 3D structural model of bovine GPR61 and the inferred binding sites after the removal of unreliable and unstructured regions (M1–G30 and A363–S451). The binding sites for EPl and CPl were inferred at three locations: the extracellular, cytoplasmic, and transmembrane regions of GPR61, and docking simulations were performed for all three sites.

Fig. 4.

Three-dimensional structure of bovine GPR61 modeled using AlphaFold2. The three binding sites detected as targets in the docking simulation, the extracellular (site 1), cytoplasmic (site 2), and transmembrane (site 3) sites (indicated by colored spheres). TM, transmembrane region.

Figures 5A, 5B, 5C, and 5D show the docking of EPl or CPl molecular species onto the extracellular site; the binding sites of both sn-1 and sn-2 side chains are located at the center and the hydrophobic regions between TM1 and TM7. Moreover, the binding sites of phosphoric acid and amines are located near TM1 and TM2.

Fig. 5.

Docking pose of C18:0-C18:1EPl (A, E, I), C18:0-C18:1CPl (B, F, J), C18:0-C20:4CPl (C, G, K), and C18:0-C22:6CPl (D, H, L) in the extracellular site (site 1) (A, B, C, D), the cytoplasmic site (site 2) (E, F, G, H), or the transmembrane site (site 3) (I, J, K, L) of three-dimensional (3D) structure of bovine GPR61 modeled using ColabFold. The conformations shown feature the smallest docking scores. Red, blue, and yellow surfaces represent regions of the binding site that are suitable for ligand hydrogen-bond acceptors, ligand hydrogen-bond donors, and hydrophobic groups, respectively. Red arrows represent the head group and purple and brown side-chains represent fatty alcohols bonded to a glycerol backbone at sn-1 position through a vinyl-ether bond and fatty acids bonded to sn-2 position through an ester bond, respectively. TM, transmembrane region.

Figure 5E, 5F, 5G, and 5H show the docking of EPl or CPl molecular species onto the cytoplasmic site; the binding site of the sn-1 side chain is located between TM5 and TM6, whereas the binding site of sn-2 is located closer to TM6 and TM7.

Figure 5I, 5J, 5K, and 5L reveals the docking of EPl or CPl molecular species onto the transmembrane site; the binding of the sn-2 side-chain is located at a narrow hydrophobic region between TM4 and TM5.

Figure 6 shows the distribution of the docking scores and the binding free energy based on MM‑GBSA of EPl or CPl molecular species onto the three binding sites, in the extracellular (A, D), cytoplasmic (B, E), and transmembrane (C, F) regions, in the 3D structure of bovine GPR61, modeled using ColabFold. The docking score indicates the binding free energy of each molecular species. One hundred docking poses were calculated for EPl and CPl molecular species at the extracellular, cytoplasmic, and transmembrane sites (Sites 1, 2, and 3, respectively). The docking scores of EPl and CPl showed smaller variances (indicating stronger binding) for the extracellular and cytoplasmic sites than for the transmembrane sites. However, there were no apparent trends in the binding free energies of the three sites (Figs. 6D–F).

Fig. 6.

Distributions of docking scores (A, B, and C) and the binding free energy based on the MM-GBSA method (D, E, and F), for C18:0-C18:1EPl, C18:0-C18:1CPl, C18:0-C20:4CPl, and C18:0-C22:6CPl across three binding sites: extracellular (site 1) (A, D), cytoplasmic (site 2) (B, E), and transmembrane (site 3) (C, F). The 3D structure of bovine GPR61 was modeled using ColabFold. All ligands used in docking were converted to three‑dimensional, ionized structures, and these prepared structures were consistently employed across all calculations. The docking score represents the estimated binding free energy for each molecular species, reflecting the free energy difference between the bound and unbound states. For each molecular species, 100 poses were generated during docking onto the sites. The bolded and italicized values represent the mean and population variance in the docking scores and the binding free energy across the 100 poses.

Discussion

The stimulatory effects of C18:0-C18:1CPl, C18:0-C20:4CPl, and C18:0-C22:6CPl on basal LH secretion were not statistically significant. However, in the presence of GnRH, both C18:0-C18:1CPl and C18:0-C20:4CPl synergistically stimulated FSH secretion. Furthermore, in the absence of GnRH, 700 pM C18:0-C18:1CPl stimulated FSH secretion via the SMAD pathway. Therefore, this mechanism appears to be receptor mediated. GPR61 co-localizes with GnRHR in lipid rafts on bovine gonadotroph surfaces [1]. Moreover, in silico simulations were used to estimate binding between GPR61 and CPls. Therefore, GPR61 may act a receptor. A comparison of the bovine GPR61 sequence (NP_001033660) with the human sequence (NP_001380836.1) revealed 94% identity and 95% similarity, whereas a comparison with the mouse sequence (NP_001292390.1) showed 96% identity and 97% similarity. These findings indicate that GPR61 is highly conserved, indicating that the CPl-binding site is preserved across these species. Therefore, it is important to discuss the importance of these observations.

The difference in the molecular structure lies in the head group, ethanolamine of EPls, or choline of CPls. Both C18:0-C18:1EPl and C18:0-C18:1CPl contained vinyl-ether-bonded stearic acid (C18:0) and ester-bonded oleic acid (C18:1). However, the stimulatory effects of bovine gonadotrophs on FSH and LH secretion are very different. Therefore, the present study demonstrated the importance of the head group in stimulatory effects. Compared with C18:0-C18:1CPl, C18:0-C20:4CPl and C18:0-C22:6CPl had longer ester-bonded side chains at sn-2. The stimulatory effects of C18:0-C20:4CPl and C18:0-C22:6CPl were not significant in the absence of gonadotropin-releasing hormone. In our previous study [3], LEPI, which has no acyl group at the sn-2 position, had no effect on FSH and LH secretion. Therefore, the present study also revealed the importance of the ester-bonded acyl group at the sn-2 position in controlling FSH and LH secretion.

In silico simulations were performed to estimate the binding of GPR61 to chemosynthetic EPl and CPls in this study. Accurate protein structure prediction using the AlphaFold2 algorithm is important for drug discovery and fundamental chemical biology [29]. Therefore, a 3D model of GPR61 and plasmalogens may provide important clues for clarifying the mechanisms of FSH and LH secretion. However, the GPCR models predicted by AlphaFold2 may not be perfect when compared to the experimental structures [30]. A recent study by another group needed to utilized both AlphaFold2 and cryo-electron microscopy due to the difficulty in preparing appropriate molecules including GPR61 attached to both Gαs and inverse agonist [31]. Although the GPR61 models predicted by AlphaFold2 appeared imperfect, an inverse agonist was observed [31]. Therefore, caution must be exercised when discussing the 3D model of GPR61 and docking scores.

We also included site 3 in the docking simulations because ligand binding at this site has been demonstrated in previous crystal structures (Protein Data Bank IDs: 5TZY [32] and 6C1R [33]). However, the Glide scoring function used in the present study penalized hydrophobic groups exposed to solvents and polar groups buried in hydrophobic protein regions. Moreover, we did not find any apparent trends in the binding free energies among the sites (Figs. 6D–F). Therefore, the scores at site 3 may not be directly comparable to those at other sites, although the smaller variance in docking scores for EPl or CPls suggests stronger binding at the extracellular and cytoplasmic sites than at the transmembrane site. The docking scores of C18:0-C22:6CPl showed a larger distribution (indicating weaker binding) for the extracellular site than those of C18:0-C18:1EPl and C18:0-C18:1CPl. The inverse agonist is bound to the cytoplasmic site of the 3D model of GPR61 predicted by AlphaFold2 to inhibit binding of Gαs [31]. There are no previous studies for the effects of CPl on gonadotropin secretion in other species. However, little is known regarding GPR61 expression. However, GPR61 requires Gαs protein to activate cytoplasmic pathway in HEK293 cells with heterologous expression of GPR61 [34]. The N-terminal domain of GPR61 is essential for its constitutive activity in vitro [35]. Therefore, further studies are required to clarify whether the extracellular site is the important site to bind EPl and CPl for the stimulatory effects on FSH and LH secretion, while the cytoplasmic site is the important site to bind Gαs.

However, it was impossible to explain all the observed stimulatory effects solely based on the calculated docking scores and binding free energy. Particularly, we could not incorporate other factors into the 3D model of GPR61 using the AlphaFold2 in this study. It is well known that GPCR proteins can form functionally active homomers and heteromers with different receptors. Heterodimerization among paralogs of GnRHRs of a protochordate result in modulation of the ligand-binding affinity, signal transduction, and internalization [36]. Moreover, GPR61 forms heteromers with other GPCRs [2]. Therefore, it is possible that GPR61 forms a heteromer with GnRHR, affecting ligand-binding affinity, signal transduction, and internalization of GnRHR, and thus the synthesis and secretion of FSH and LH from the AP cells.

The brain contains the highest amount of EPl compared to other organs [5]. EPl is also present in the human cerebrospinal fluid and blood [37]. Therefore, the brain may be the main source of EPls, which may arrive from the brain via the hypophyseal portal blood or systemic circulation to the AP. There is another possible explanation for the effects of CPls on gonadotophin secretion in vivo. Oral administration of plasmalogens has shown promising health benefits in patients with Alzheimer’s [38] and Parkinson’s [39] disease [39]. The amounts of CPls are approximately 4.6-fold higher than those of EPls in cow blood [7], while the amounts of EPls are approximately 27.5-fold higher than those of CPls in cow brain [7]. Blood plasmalogen concentrations were increased by dietary plasmalogen absorption from the small intestine in rats [40]. Some but not all dietary CPls are adsorbed into the lymph ducts of the small intestine [41] and are released into the portal vein [42]. Dietary DHA-enriched phosphatidylcholine increases DHA-containing EPls in the mouse brain [9]. Therefore, blood CPls may be converted into EPls in the brain and reach the AP via the pituitary portal blood, affecting gonadotrophs. Further studies are required to determine the direct and indirect mechanisms of action of CPls on FSH and LH secretion in vivo.

Because of our limited facilities and skills, we could not utilize X-ray crystallography or cryoelectron microscopy to experimentally determine the structures. GPCRs transmit specific external stimuli into cells by changing their conformation to couple and activate G proteins to initiate signal transduction [43]. Because of our limited facilities and skills, we could not utilize nuclear magnetic resonance to examine GPCR plasticity and conformational dynamics or atomic force microscopy to explore the spatiotemporal dynamics and kinetic aspects of GPCRs [44].

It should be noted that CPls play various important roles in the brain and other organs, apart from GPR61. C18:0-C18:1CPl protects against arachidonic acid-induced cytotoxicity in human neuroblastoma SH-SY5Y cells [45]. C18:0-C18:1CPl have preventive effects against Alzheimer’s disease by reducing factors involved in the amyloid β-associated pathogenesis in SH-SY5Y cells [46]. The length and saturation of CPIs alters the aggregation rate of α-synuclein in vitro [47]. However, further studies are required to elucidate the role of CPls in animals.

In conclusion, C18:0-C20:4CPl stimulated FSH secretion exclusively in the presence of GnRH, whereas C18:0-C18:1CPl weakly stimulated FSH secretion and showed potential interaction with the GnRH signaling pathways.

Conflict of interests

The authors have nothing to declare.

Data availability

The data supporting this study will be shared upon reasonable request from the corresponding author.