Dopamine receptor 2 activation inhibits ovarian vascular endothelial growth factor secretion in vitro: implications for treatment of ovarian hyperstimulation syndrome with dopamine receptor 2 agonists  Hortensia Ferrero, Ph.D., Carmen M. García-Pascual, Ph.D., Raúl Gómez, Ph.D., Francisco Delgado-Rosas, Ph.D., Omar Cauli, Ph.D., Carlos Simón, M.D., Francisco Gaytán, M.D., Antonio Pellicer, M.D.  Fertility and Sterility  Volume 101, Issue 5, Pages 1411-1418.e2 (May 2014) DOI: 10.1016/j.fertnstert.2014.01.031 Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Figure 1 The effects of dopamine receptor 2 agonists (D2-Ag) and antagonists (D2-Ant) on vascular endothelial growth factor (VEGF) secretion by luteinized granulosa cells (GCs). Graph shows a dose-response experiment in which luteinized GCs (n = 13) were cultured in the presence of 5 IU hCG with 0 (vehicle control), 20, 40, 60, 80, or 100 μM doses of the D2-ag cabergoline, the D2-ant L-741.626, or a combination of both compounds for 72 hours. The amounts of VEGF in the supernatant were measured at the end point. Each point on the graph represents the mean of 13 patients at each of the D2-Ag and/or D2-Ant doses assayed. A VEGF dose-response inhibition is observed in luteinized GCs cultured with D2-ag alone. The decreases in VEGF secretion initiated by cabergoline were partially prevented when similar doses of L-741.625 were administered concomitantly. ∗P<.05, ∗∗P<.01, compared with the vehicle. Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Figure 2 Linear regression between the amount of dopamine receptor 2 (D2) and the percentage of luteinized granulosa cell (GC) vascular endothelial growth factor (VEGF) inhibition in response to increasing doses of D2 agonists. Graph shows the correlation curves between D2 luteinized GC-surface expression (n = 13; cultured with 0 [vehicle control], 20, 40, 60, 80, or 100 μM doses of D2 agonists [cabergoline] in the presence of 5 IU/mL hCG for 72 hours; X-axis) and the extent of VEGF secretion inhibition (the presence of VEGF in the supernatant; Y-axis). The percentage of VEGF inhibition accomplished with each of the D2 agonist doses assayed was determined by using the formula: 100 − [(VEGF measured in cabergoline treated condition × 100)/VEGF measured in cabergoline untreated conditions]. Each different symbol (Δ, ■, ○, ▴, □) represents each of the five D2 agonist doses at which the same luteinized GCs (n = 13) were assayed. When cultured with 20–60 μM cabergoline the efficiency of VEGF secretion inhibition by D2 agonists depends on the luteinized GC-surface D2 density, as shown by the correlation between these parameters. When cultured with 80–100 μM cabergoline the efficiency of VEGF secretion inhibition by D2 agonists does not depend on the amount of D2 expressed by luteinized GCs as shown by the lack of correlation between these parameters. ∗R2 > 0.3 was considered a positive correlation. Abs = absorbance. Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Figure 3 The effects of the dopamine receptor 2 agonist cabergoline (Cb2) on vascular endothelial growth factor (VEGF) messenger RNA (mRNA) production and protein secretion by luteinized granulosa cell (GC) over a time course. The VEGF mRNA expression (expressed as arbitrary fluorescence units [AFU]) in the presence (A) or absence (C) of 5 IU/mL hCG, and VEGF protein secretion levels (expressed as picograms per milliliter) in the presence (B) or absence (D) of 5 IU/mL hCG in luteinized GCs (n = 8) cultured with/without Cb2 (100 μM) at different time points (24, 48, or 72 hours). Cabergoline affected VEGF protein levels in treated versus control luteinized GCs but not VEGF mRNA expression. ∗∗P<.01, compared with Cb2 untreated conditions. Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Figure 4 Immunohistochemical detection and quantification of dopamine receptor 2 (D2) in ovarian luteinized granulosa cell (GC) sections throughout the luteal phase. Images represent immunohistochemical detection of D2 (brown) in ovarian sections obtained from women undergoing surgery during the early (A), mid (B), or late (C) luteal phases. Intense D2 immunostaining was present in theca layer (white arrows) throughout the luteal phase. Macrophages were observed in the theca layer, by immunostaining of parallel sections with the macrophage marker CD68 (black arrows in the inset in A). Neither stromal nor thecal ovarian blood vessels (asterisks in A) stained positive for D2. Variation in luteinized GC D2 staining intensity (tissue outlined in green) was observed throughout the luteal phase. The D2 immunostaining was clearly positive in luteinized GCs during early luteal phase (A) but became faint during the midluteal phase (B), and disappeared in GC-lutein cells during the late luteal phase (C). Graph below the images (D) corresponds to quantitative image analysis of D2 staining throughout the luteal phase (expressed as energy units/pixel [eu/pix]) in the green area of outlined luteinized GCs. Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Supplemental Figure 1 Time course and dose-response effects of hCG on vascular endothelial growth factor secretion and apoptosis in luteinized granulosa cells (GCs). Graph in (A) shows time course experiments in which luteinized GCs (n = 6) were cultured for 1, 3, 5, or 7 days, in the presence of 10 IU/mL of hCG, to define the peak in vascular endothelial growth factor (VEGF) secretion. Graph in (B) shows a dose-response experiment in which luteinized GCs (n = 6) were cultured with 0, 1, 5, 10, or 50 IU/mL of hCG to identify which dose induced the highest level of VEGF secretion after 72 hours. The VEGF secretion peaked after 72 hours. The VEGF levels increased in response to hCG in a dose-dependent fashion, reaching a plateau at 5 IU/mL. ∗P<.05, ∗∗P<.01, compared with control conditions (day 1, hCG at 0 IU). Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions

Supplemental Figure 2 Dose-response evaluation of the toxic effects exerted by dopamine receptor 2 agonists and antagonists in luteinized granulosa cells (GCs). Luteinized GCs (n = 8) were cultured with 5 IU/mL hCG plus cabergoline (Cb2) (dopamine receptor 2 agonist) or L-741.626 (dopamine receptor 2 antagonist) at 0, 20, 40, 60, 80, 100, 200, 400 μM, or an ethanol (EtOH) vehicle, for 72 hour. (A) Cell viability: luteinized GCs were incubated with Hoechst 33342 and propidium iodide and the specific absorbance of each dye was assessed with an In-Cell analyzer to provide automatic quantification of cell viability. (B) Apoptosis (as the percentage of the total number of cells): luteinized GCs were fixed using 4% formaldehyde and apoptosis was assessed using the ApopTag ISOL Dual Fluorescence kit. The amount of vehicle required to dilute each drug dose is expressed as the final percentage (vol:vol) of EtOH in the cultures within its corresponding bar. There was a slight but insignificant decrease in cell viability observed starting at 200 μM, but dramatic decreases were observed at 400 μM doses of both dopamine receptor 2 agonist and dopamine receptor 2 antagonist, and this decrease was paralleled by similar decreases when corresponding amounts of the vehicle alone was used. A similar pattern to that observed for cell mortality was seen during the analysis of apoptosis. A one-way analysis of variance (ANOVA) followed by Tukey post-hoc analysis was used to compare apoptosis and viability rates in treated luteinized GCs. ∗P<.05, compared with the corresponding control group (0 μM, 0 EtOH). Fertility and Sterility 2014 101, 1411-1418.e2DOI: (10.1016/j.fertnstert.2014.01.031) Copyright © 2014 American Society for Reproductive Medicine Terms and Conditions