Secoiridoids from Dogwood (Cornus officinalis) Potentiate Progesterone Signaling

Abstract

The use of botanical dietary supplements for the alleviation of conditions such as hot flashes, premenstrual syndrome, and fertility, is prolific worldwide. Estrogen and progesterone receptors (ER and PR) and their corresponding steroid hormones are critical for the relief of hot flashes, the treatment of patients who develop endometriosis, and these pathways can influence the development of endometrial, ovarian and breast cancers. However, few studies have investigated or identified the natural product components in herbal supplements that act on the PR. In the current study, a new secoiridoid, demethoxy-cornuside (1), along with six known secoiridoids (2–7) were isolated from the twigs of dogwood (Cornus officinalis) by bioassay-guided isolation with a progesterone response element (PRE)-luciferase (Luc) reporter assay in Ishikawa cells. Four phytoprogestins (1, 2, 6, 7) potentiated the effect of progesterone in the PRE/Luc assay. This study demonstrates that C. officinalis components might potentiate progesterone signaling in the presence of progesterone, which could modify progesterone receptor action in hormone responsive tissues such as the uterus and mammary gland.

Graphical Abstract

The use of botanical dietary supplements for women’s health is prolific worldwide and spans from use in premenstrual syndrome, fertility, and alleviation of hot flashes.¹ An estimated 9.60 billion dollars is spent annually by Americans on alternative products, including on health supplements marketed specifically toward women.² Estrogen and progesterone receptors (ER, PR), and their corresponding steroid hormones (estradiol and progesterone), act through complex mechanisms to regulate biological processes in women’s health. These receptors are critical for the treatment of patients who develop endometriosis, the relief of hot flashes, and both hormones can impact the risk for developing endometrial, breast and ovarian cancers.^3,4 However, few studies have investigated or identified the natural product components found in herbal supplements that act on the PR when compared to literature that focuses on its estrogenic counterpart, even though drugs commonly used in women’s health act through PR-mediated mechanisms.¹ Given the established role of ER modulators in botanicals, ligands for other hormone receptors, such as PR, are likely present and bioactive in supplements used for women’s health.⁵ Since the health of millions of women worldwide may currently be affected, investigations into botanicals as a source of progestins is significant, particularly given that several ubiquitously distributed natural product structural classes may interact with PR receptor binding.^5-7

One botanical supplement with reported hormone modifying activity is dogwood [Cornus officinalis Sieb. et Zucc. (Cornaceae)], which is also commonly used as a traditional medicine to nourish the liver and kidney in several east Asian countries. In modern pharmacological studies, the extract of C. officinalis and its individual specialized metabolites have shown a broad range of biological activities: renal and hepatic protective, antidiabetic, cardioprotective, antioxidative, anti-inflammatory, antiaging, neuroprotective, and antibacterial effects.^8,9 To date, more than 300 compounds have been reported from C. officinalis including flavonoids, iridoids, tannins, saccharides, terpenes, organic acids, and essential oils.^8-10 Although the popular usage of C. officinalis is not related to gynecological symptoms, the extract of the fruit of C. officinalis inhibited progesterone induced activation of luciferase in a PRE/Luc assay in T47D breast cancer cells.⁶ This result suggested that compounds that act like progesterone may be present in dogwood. In the current study, we employed a progesterone response element (PRE)-luciferase reporter assay in Ishikawa cells to characterize the agonist, antagonist, or potentiation effects of the fruits and twigs of dogwood. This resulted in the isolation of a new secoiridoid, demethoxy-cornuside (1), along with six known secoiridoid glycosides, cornuside (2), secologanoside (3), secoxyloganin (4), sweroside (5), (7α)-7-O-methyl morroniside (6), and (7β)-7-O-methyl morroniside (7). Four of these compounds (1, 2, 6, and 7) potentiated the activity of progesterone, increasing the PRE/Luc activity 42%, 43%, 29%, 18% higher than progesterone alone. This study demonstrates that components of C. officinalis might potentiate progesterone signaling, which could modify progesterone mediated action in reproductive tissues such as the uterus and mammary tissue.

In order to identify phytoprogestins from C. officinalis, the fruit and the twigs were provided from the Botanical Center at the University of Illinois at Chicago. Previously, the extract of the fruit of C. officinalis was tested using a PRE/Luc assay in T47D breast cancer cells and inhibited progesterone induced activation of luciferase.⁶ In this study, Ishikawa cells that stably express progesterone receptor B (PR-B) were used to screen for compounds that activate or inhibit the receptor. Initially, extracts of the fruit and the twigs of C. officinalis were tested at 20 μg/mL in the PRE-luciferase assay and did not induce significant PRE/Luc activity (Figure 1A). However, when each extract was combined with 1 nM of progesterone, only the twig extract mixture resulted in a 122% increase in PRE/Luc activity over that induced by 1 nM progesterone alone, indicating that components of the twig extract were potentiating progesterone activity (Figure 1B). To determine which constituents were responsible for this, bioassay-guided fractionation was performed. Liquid-liquid partitioning, followed by reversed-phase HPLC separation of a BuOH layer led to the isolation of a new secoiridoid, demethoxy-cornuside (1), along with six known secoiridoid glycosides, cornuside (2), secologanoside (3), secoxyloganin (4), sweroside (5), (7α)-7-O-methyl morroniside (6), and (7β)-7-O-methyl morroniside (7) (Figure 2). Structures and their corresponding bioactivities were analyzed.

Figure 1.

The extract of the twigs of C. officinalis potentiated the effect of 1 nM progesterone on PRE/Luciferase induction in a reporter assay. (A) Progesterone-response element (PRE/Luc) activity in Ishikawa cells stably expressing progesterone receptor B with the extract of the dogwood fruit and twigs. (B) PRE/Luc activity after treating Ishikawa cells with the extract of the dogwood fruit and twigs in the presence of 1 nM progesterone. Extracts were tested at 20 μg/mL concentration, n≥2. Data represent mean ± SEM.

Figure 2.

Structures isolated from the twigs of C. officinalis.

The molecular formula of demethoxy-cornuside (1) was C₂₃H₂₈O₁₄ as determined by HRESIQTOFMS analysis. The ¹³C and ¹H NMR data of this compound were similar to those of 2 and implied a secoiridoid glycoside.¹¹ In the ¹H NMR spectrum, a terminal double bond was deduced from the resonances at δ_H 5.80 (1H, ddd, J = 17.3, 10.4, 9.2 Hz), δ_H 5.29 (1H, dd, J = 17.3, 1.7 Hz), and 5.23 (1H, dd, J = 10.3, 1.7 Hz). In addition, the singlet at δ_H 7.14 and 7.06 suggested a trisubstituted double bond and a galloyl group, respectively. The ¹³C NMR spectrum contained signals corresponding to a carbonyl carbon, δ_C 168.6 (a carboxyl signal at δ_C 174.2 was deduced through analysis of the HMBC spectrum), and a sugar unit (characteristic signals at δ_C 99.7 and 78 – 62 and δ_H 4.7 – 3.2) (Table 1). Overall, these NMR features indicated that 1 was an analogue of cornuside (2).

Table 1.

¹H NMR (900 MHz) and ¹³C NMR (225 MHz) Data for Compound 1 in CD₃OD.

Position	δH (J in Hz)	δC
1	5.43, d (5.0)	97.2
3	7.14, s	148.4
4		116.1
5	3.00, dt (10.6, 5.5)	30.5
6	2.24, m	29.6
	1.80, m
7	4.28, m	64.3
8	5.80, ddd (17.3, 10.4, 9.2)	136.1
9	2.69, dt (9.9, 5.2)	45.4
10	5.29, dd (17.3, 1.7)	119.4
	5.23, dd (10.3, 1.7)
11		174.2a
1′	4.66, d (7.9)	99.7
2′	3.21, dd (9.3, 7.9)	74.7
3′	3.36, dd (9.1, 8.6)	77.8
4′	3.28, dd (9.6, 8.4)	71.5
5′	3.29, m	78.3
6′	3.88, dd (12.0, 2.4)	62.7
	3.66, dd (12.1, 5.8)
1″		121.6
2″, 6″	7.06, s	110.1
3″, 5″		146.4
4″		139.7
7″		168.6

^aDeduced from HMBC spectrum.

The complete flat structure of 1 was determined by a combination of 2-D NMR experiments. First, the presence of a sugar moiety was secured by a combination of COSY and gHSQC data for the signals (Figure 3, COSY correlations). Similarities in the chemical shifts and splitting patterns of these signals with those of 2 suggested the presence of a glucose group. The pyranose nature of this sugar was supported by long-range correlations from H-5′/C-1′ in the gHMBC data. In addition, the β-configuration at the anomeric center was also assigned by its carbon chemical shift (δ_C 99.7) and large ¹H/¹H coupling constant (J_1′,2′ = 7.9 Hz). In order to determine the D- or L-configuration of the glucose group, acid hydrolysis of 1 followed by derivatization with L-cysteine methyl ester and phenyl isothiocyanate was carried out.^12,13 The UPLC retention time and MS data of the hydrolysate derivative of 1 were compared with those of D- and L-glucose standards; the data were identical to those from D-glucose. Taken together, these data indicate the presence of a β-D-glucopyranose unit on compound 1.

Figure 3.

COSY (blue lines) and key HMBC (red arrows) correlations for compound 1.

The structure of the aglycone was also determined by combined NMR analysis. Based on COSY (H-5–H-10, H-9–H-1) and gHMBC (H-1/C-3; H-3/C-1, C-5, C-11; H-5/C-3, C-11), the presence of 3-vinyl-4-oxyethyl-3,4-dihydropyran-5-carboxylic acid was revealed (Figure 3). A galloyl group was confirmed by analysis of chemical shifts from ¹³C, ¹H, and gHMBC (H-2″, H-5″/C-1″–C7″) experiments. The chemical shift of C-7 at δ_C 64.3 and the HMBC correlation H-7/C-7″ indicated a galloyl moiety at this location via an ester linkage. The connection of a glucose moiety with the secoiridoid aglycone was established through observation of the downfield shift of C-1 at δ_C 97.2, the gHMBC correlations (H-1/C-1′ and H-1′/C-1), and the ROESY cross-peak at H-1/H1′. Thus, compound 1 was identified as demethoxy-cornuside.

To determine absolute configurations of three stereogenic centers at C-1, C-5, and C-9 in the secoiridoid moiety, the electronic circular dichroism (ECD) and coupling constants were implied. The ECD spectrum of 1 was compared to the ECD spectrum of authentic cornuside (2), purchased from Adooq Bioscience (Irvine, CA, USA) (Figure S7, Supporting Information). The ECD curve of 1 was highly comparable to that of 2. Thus, we concluded that C-5 and C-9 had 5S and 9R configurations, which were identical with the corresponding configurations of 2. For the configuration at C-1, the observed ¹H/¹H coupling constant (J_1,9 = 5.0 Hz) was compared to similar classes of secoiridoids reported in literature.^11,14-22 According to the literature, in the ¹H NMR spectra of several naturally occurring secoiridoid derivatives and their synthetic derivatives, the ¹H/¹H coupling constant (J_1,9) for a trans-oriented center exhibits values in the range of 4.5 – 8.8 Hz.^11,14-22 Conversely, the coupling constant for a cis-oriented center has been recorded as 1.5 – 2.5 Hz.^21,22 Therefore, C-1 has a 1S configuration.

PRE/Luc activities of 1 – 7 at 10 μM were evaluated. None of compounds activated PRE/Luc activity alone (Figure 4A). However, in the presence of 1 nM progesterone, compounds 1, 2, 6, and 7 exhibited potentiation effects, increasing the PRE/Luc activity 42%, 43%, 29%, and 18% higher than progesterone alone (Figure 4B). Compounds 1 and 2 exhibited higher potency than other compounds and similar potency to each other, which indicated that a methoxy or demethoxy moiety at C-11 didn’t affect the activity. Based on the structural similarity between 1 and 3 or 2 and 4, the galloyl group likely contributes significantly to the potentiation activity. This is the second finding of potentiators mediating PR signaling after our previous report of the bioactivity of irilone from red clover.²³ Previous reported natural product PR signaling mediators were steroidal, phenolic acid, or flavonoid structures.^6,7,23-26 Therefore, it is possible that the secoiridoids from C. officinalis might provide a distinct mechanism of action for PR signaling, though further experiments are required to confirm this. Further, there is a chance that these secoiridoids could amplify endogenous progesterone signaling because some of the secoiridoids from C. officinalis potentiated progesterone signaling at a biologically relevant concentration. The fruit of C. officinalis, which is used as a dietary supplement, is also known to contain secoiridoids. In our study, we did not observe significant potentiation effects on PR mediated signaling with the fruit extract, but we found that the nonpolar layer of the fruit extract exhibited antagonist activity in the PRE/Luc assay. Therefore, the potentiation effect of the fruit extract might be masked by antagonists in the nonpolar layer (Figure S8, Supporting Information). Further work is needed to elucidate the effect of each of these secoiridoids in progesterone-responsive tissues and characterize antagonists in the fruit extract so that women can make more informed decisions in regard to the safe use of supplements derived from C. officinalis.

Figure 4.

Compound 1, 2, 6, and 7 potentiated the activity of progesterone. (A) PRE/Luc activity in Ishikawa PR-B cells with 1, 2, 6, and 7. (B) PRE/Luc activity after treating Ishikawa cells with 1, 2, 6 and 7 in the presence of 1 nM progesterone. Compounds were tested at 10 μM concentration, n≥3. Data represent mean ± SEM. Significantly different from 1 nM progesterone, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

EXPERIMENTAL SECTION

General Experimental Procedures.

Optical rotation was measured on a PerkinElmer 241 polarimeter using a 100 mm cell. UV spectra were recorded on a Varian Cary 5000 spectrophotometer. IR spectra were acquired on a Thermo-Nicolet 6700 FT-IR equipped with a SMART iTR sampling accessory. ECD spectra were recorded on a JASCO J-815 spectrometer. NMR spectra were obtained on a Bruker 800 MHz AVANCE NMR spectrometer equipped with a 5 mm TXI room temperature inverse probe with xyz-axis pfg and Topspin version 1.3 operating software, and a Bruker 900 (226.2) MHz AVANCE NMR spectrometer equipped with a 5 mm TCI cryogenic inverse probe with z-axis pfg and TopSpin version 2.1 operating software, at the University of Illinois at Chicago Center for Structural Biology. Chemical shifts (δ) are given in ppm and coupling constants (J) are reported in Hz. ¹H and ¹³C NMR chemical shifts were referenced to the CD₃OD (δ_H 3.31 ppm and δ_C 49.0 ppm). High resolution mass spectra were obtained on a Bruker COMPACT ESIQTOF mass spectrometer at the University of Illinois at Chicago. HPLC was performed on a Shimadzu LC-20AB equipped with a SPD-20A UV/Vis detector and a Waters Prep LC4000 system equipped with a Water486 UV/Vis detector. All solvents were spectroscopic grade.

Plant Material.

Dried plant twigs and fruits of Cornus officinalis were provided by the UIC/NIH Center for Botanical Dietary Supplements (plant material code BC362 and BC282).

Extraction and Isolation.

Dried plant material (197.5 g) of Cornus officinalis was grinded and repeatedly extracted with MeOH (1L × 3) and CH₂Cl₂ (1L × 3). The combined extracts (16.3 g) were successively partitioned between H₂O and hexanes; the former fraction was sequentially repartitioned with EtOAc and n-BuOH (2.34 g). An aliquot of the n-BuOH layer (1.50 g) was fractionated using RP-C₁₈ preparative HPLC (10 mL/min, gradient of MeOH:H₂O with 0.1% formic acid (FA) from 30:70 to 65:35 over 20 min and from 65:35 to 100:0 over 5 min, followed by an isocratic flow of 100% MeOH with 0.1% FA for 10 min), to afford fractions F1 to F16. F6, F8, F10, and F12 were rich with specialized metabolites. F6 and F8 were purified using RP-C₁₈ semi-preparative HPLC (2.0 mL/min, gradient of MeCN-H₂O with 0.1% FA from 10:90 to 65:35 over 25 min and from 65:35 to 100:0 over 5 min, followed by an isocratic flow of 100% MeCN with 0.1% FA for 5 min) to afford secologanoside (3) and sweroside (5). F10 was purified using RP-C₁₈ analytical HPLC (0.8 mL/min, gradient of MeCN-H₂O with 0.1% FA from 5:95 to 25:75 over 40 min, followed by an isocratic flow of 100% MeCN with 0.1% FA for 5 min) to afford secoxyloganin (4), (7β)-7-O-methylmorroniside (7), (7α)-7-O-methylmorroniside (6), and demethoxy-cornuside (1). F12 was purified using RP-C₁₈ semi-preparative HPLC (2.0 mL/min, gradient of MeCN-H₂O with 0.1% FA from 1:9 to 1:0 over 30 min, followed by an isocratic flow of 100% MeCN with 0.1% FA for 5 min) to afford cornuside (2).

Demethoxy-cornuside (1).

An off-white amorphous powder; [α]_D²⁰ −1.2 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 275 (5.16) nm; ECD (MeOH) λ_max (Δε) 227 (−7.5), 248 (+2.6) nm; IR (neat) ν_max 3309, 2940, 2829, 1447, 1022 cm⁻¹; ¹H NMR (900 MHz, CD₃OD) and ¹³C NMR (225 MHz, CD₃OD), see Table 1; HRESIMS m/z 551.1395 [M + Na]⁺ (calcd for C₂₃H₂₈NaO₁₄, 551.1377).

Hydrolysis and UPLC-MS Analysis of Compound 1.

A solution of 1 (2.0 mg) in 3 N HCl (500 μL) was stirred at 90 °C for 1 h. After hydrolysis, the reaction mixture was concentrated, diluted with H₂O, neutralized with 3N NH₄OH, and dried. The residue was re-dissolved in 500 μL of a solution of L-cysteine methyl ester (50 mg/mL) and pyridine. After mixing thoroughly, the reaction mixture was incubated at 60 °C for 1 h. 500 μL of reagent consisting of phenyl isothiocyanate and pyridine was added and heated for another hour. The solution was filtered and injected into an UPLC-MS/MS system (0.5 mL/min, isocratic of MeCN-H₂O at 15:85 with 0.1% FA) with C₁₈ UPLC column (Agilent, 2.1 × 50 mm). A peak corresponding to a hydrolysate sugar derivative was detected at 4.90 min. The same procedures were carried out with L- and D-glucose. The retention time and MS data of the detected peak in the test sample were compared with those of standard D- (t_R = 4.98 min) and L-glucose (t_R = 4.48 min), respectively.

Authenticated Compounds.

Cornuside (2) (A14599, Adooq Bioscience, Irvine, CA, USA), progesterone (P0130-25G, Sigma-Aldrich, St. Louis, MO, USA), D-glucose (G8270-100G, Sigma-Aldrich, St. Louis, MO, USA), and L-glucose (G5500-250MG, Sigma-Aldrich, St. Louis, MO, USA) were purchased from commercial sources.

Cell Culture.

Ishikawa cells stably expressing PR-B were treated as described previously.²³

Progesterone Response Element/Luciferase (PRE/Luc) Activity Assay.

PRE/Luc activity assay were performed as previously described.²³

Statistical Analysis.

All data are presented as means ± the standard error of the mean (SEM) with n≥3. Due to the large number of samples, screening data of extracts, partitioned layers, and HPLC fractions of C. officinalis were not statistically analyzed. All other data were analyzed by ANOVA followed by a Tukey’s or Dunnett’s post hoc test, with p < 0.05 considered significant. Analysis was performed using Prism version 7.0a.