Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 24.
Published in final edited form as: J Nat Prod. 2025 Jan 14;88(1):183–190. doi: 10.1021/acs.jnatprod.4c01257

Cytotoxic and Noncytotoxic Steroidal Constituents of Cryptolepis dubia

Yulin Ren , Elizabeth N Kaweesa , Ruoheng Zhou §, Yue Liu §, Kongmany Sydara , Mouachanh Xayvue , Djaja D Soejarto ‡,, Sijin Wu §, Xiaolin Cheng , Joanna E Burdette , A Douglas Kinghorn †,*
PMCID: PMC11792172  NIHMSID: NIHMS2052132  PMID: 39808736

Abstract

(−)-Cryptanoside A (1) was identified previously as a major cytotoxic component of the stems of Cryptolepis dubia collected in Laos, which mediates its activity by targeting Na+/K+-ATPase (NKA), with hydrogen bonds formed between its 11- and 4′-hydroxy groups and NKA being observed in its docking profiles. In a continuing investigation, 1 and its 17-epimer, (−)-17-epi-cryptanoside A (2), and the new (+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3) and the known [(+)-2,21-dihydroxypregna-4,6-diene-3,20-dione or 2-hydroxy-6,7-didehydrocortexone (4)] pregnane-type steroids were isolated from C. dubia. In addition, (−)-11,4′-di-O-acetylcryptanoside A (1a) has been synthesized from the acetylation of 1. The structures of these compounds were determined by analysis of their spectroscopic data, with their cytotoxic and NKA inhibitory activities being evaluated. In contrast to 1 that exhibited potent activities, the other compounds were largely inactive. Molecular docking profiles indicated that 13 and 1a bind to NKA, but some subtle differences were observed in their interactions with NKA, which may contribute to their differential cytotoxic and NKA inhibitory potency.

Graphical Abstract

graphic file with name nihms-2052132-f0001.jpg


Steroids are an important group of natural products involved in ubiquitous cellular processes and biotechnological applications and thus exhibit a wide range of physiological activities.1 As major signaling chemicals, these compounds play a key role in intercellular communication, biosynthesis, and other biological functions of plants that are involved with their growth and development.2 Thus far, several steroids have been used as hormone therapy for the treatment of cancer and other diseases, of which the synthetic pregnane corticosteroid, dexamethasone, is a widely used chemotherapy for the treatment of infections and cancer,3 while cyproterone acetate has been used clinically for palliative care in prostate cancer and other androgen-dependent conditions.4,5

Cardiac glycosides are a group of steroid lactones that have attracted wide interest due to their therapeutic use in treating congestive heart disease through inhibition of Na+/K+-ATPase (NKA), and several of these compounds have been evaluated in cancer clinical trials. For example, digoxin was found to show antitumor activity by targeting directly HIF-1α, NKA, and NF-κB, while oleandrin targets multiple proteins and signaling pathways to mediate its activity.6

Cryptolepis dubia (Burm.f.) M.R. Almeida (Apocynaceae) is an evergreen liana used widely in folk medicine in Southeast Asia for the treatment of infections and other diseases.7 In a continuing search for anticancer agents from higher plants,8 C. dubia collected in Laos was found to exhibit cytotoxic effects against several different human cancer cell lines, from which a cardiac glycoside epoxide, (−)-cryptanoside A (1), was identified previously as a major active component. (−)-Cryptanoside A (1) binds to NKA and inhibited its activity, but its binding pose and NKA inhibitory potency are different from those observed for digoxin, indicating that this compound could be useful for the design of new antitumor agents that target NKA.9 To provide a preliminary structure-activity relationship (SAR) study, 1 and several additional compounds were isolated from C. dubia in the present investigation. These include (−)-17-epi-cryptanoside A (2), [(+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3), and (+)-2,21-dihydroxypregna-4,6-diene-3,20-dione or 2-hydroxy-6,7-didehydrocortexone (4), with (−)-11,4′-di-O-acetylcryptanoside A (1a) being synthesized from 1. In addition, the structural determination and cytotoxic and NKA inhibitory activities of these compounds obtained are reported herein, with the structure of a clinically used analogue of 3 and 4, cyproterone acetate (5), being presented as a reference.

graphic file with name nihms-2052132-f0006.jpg

RESULTS AND DISCUSSION

A sample of the stems of C. dubia collected in Laos was extracted previously with MeOH, followed by partitioning with n-hexane and CH2Cl2, from which (−)-cryptanoside A (1) was isolated as a major active component.9 Additional phytochemical work on this plant sample resulted in the isolation of a further quantity of 1 and several other steroidal compounds 24, with 1a being prepared from 1. The complete structure of 1 was determined previously by analysis of its spectroscopic and the single-crystal X-ray diffraction data, following which the complete structures of 1a and 24 have been identified by analysis of their spectroscopic data and by comparison of these data with those reported for 1 and other analogues.

Compound 2 was purified as an amorphous colorless powder, [α]20D −10 (MeOH) and −23 (CHCl3), with a molecular formula of C30H42O10, the same as 1, as shown by the protonated molecular peak at m/z 563.2825 observed in its HRESIMS spectrum, in conjunction with the 1H and 13C NMR spectroscopic data (Table 1).10 The UV (λmax 218 nm) and IR (νmax 3457, 1744, 1709, and 1623 cm−1) spectra showed absorptions consistent with the presence of hydroxy and carbonyl groups and an unsaturated lactone unit, and the NMR spectra indicated the characteristics of a cardiac glycoside.9

Table 1.

1H and 13C NMR Spectroscopic Data of 2 and 3

position 2a 2b 3a 3b

1 32.77, CH2 α, 1.37, dd (13.8, 4.0)
β, 2.29, dt (13.9, 3.6)
42.63, CH2 α, 1.66, m
β, 2.40, dd (12.5, 5.7)
2 27.25, CH2 β, 1.56, m
α, 1.67, m
69.91, CH β, 4.43, dd (13.2, 5.6)
3 70.59, CH α, 3.85, m 199.65, C
4 32.89, CH2 β, 1.30, dd (9.2, 2.7)
α, 1.52, m
120.83, CH 5.76, br s
5 35.86, C β, 1.71, m 165.33, C
6 27.13, CH2 α, 1.56, m
β, 2.38, dd (15.8, 6.3)
127.72, CH 6.16, br s
7 52.13, CH α, 3.36, d (6.2) 141.48, CH 6.16, br s
8 63.09, C 37.58, CH β, 2.26, t (10.5)
9 35.90, CH α, 2.19, m 51.04, CH α, 1.34, m
10 35.22, C 38.10, C
11 73.63, CH β, 4.65, dd (12.5, 4.8) 20.58, CH2 α,1.44, m
β, 1.66, m
12 212.02, C 37.86, CH2 α, 1.34, m
β, 2.17, m
13 64.80, C 44.95, C
14 82.09, C 52.97, CH α, 1.36, m
15 33.60, CH2 α, 2.06, m
β, 1.95, m
24.08, CH2 β, 1.93, m
α, 1.48, m
16 25.43, CH2 β, 1.88, m
α, 2.12, m
23.74, CH2 α, 2.20, m
β, 1.93, m
17 46.28, CH β, 3.13, dd (11.4, 7.2) 54.25, CH α, 2.47, t (9.0)
18 19.16, CH3 β, 1.40, s 13.44, CH3 β, 0.83, s
19 23.39, CH3 β, 1.23, s 17.23, CH3 β, 1.23, s
20 170.48, C 175.05, C
21 73.51, CH2 5.16, dt (16.9, 1.6)
4.87, dd (17.0, 1.6)
22 116.21, CH 5.86, dd (3.4, 1.6)
23 174.11, C
1’ 95.65, CH α, 4.95, br d (3.2)
2’ 34.65, CH2 β, 2.22, m
α, 1.54, m
3’ 78.45, CH β, 3.55, m
4’ 76.41, CH α, 3.17, br t (9.1)
5’ 67.87, CH β, 3.71, dd (9.3, 6.2)
6’ 17.92, CH3 α, 1.26, d (6.2)
OH-11 α, 3.52, d (4.8)
OH-14 β, 2.65, d (1.4)
OH-20 2.37, t (7.6)
OCH3-3’ 56.61, CH3 α, 3.41, s
OH-4’ β, 2.45, br s
a

Spectroscopic data (δ) were measured in CDCl3 at 176.02 MHz (Bruker AVE II) and referenced to the solvent residual peak at δ 77.16.10 Assignments of chemical shifts are based on the analysis of 1D- and 2D-NMR spectra. The 13C NMR carbon types were determined by DEPT 90, DEPT 135, and HSQC experiments.

b

Spectroscopic data (δ, Hz) were measured in CDCl3 at 400.13 MHz for 2 and 700.03 for 3 and referenced to the solvent residual peak at δ 7.26.10 Assignments of chemical shifts are based on the analysis of 1D- and 2D-NMR spectra.

Comparison of the NMR spectroscopic data of 2 with those of 1 implied that these two compounds contain an identical saccharide moiety and similar steroidal A and B rings, but the structural characteristics near the C-17 position were found to be different. The 1H NMR chemical resonances at δH 3.13 (dd, J = 11.4 and 7.2 Hz, H-17), δH 1.40 (s, H-18), and δH 5.86 (dd, J = 3.4 and 1.6 Hz, H-22) for 2 were shifted to those at δH 3.95 (t, J = 8.5 Hz), δH 1.06 (s), and δH 6.00 (d, J = 0.8 Hz) for these respective protons of 1, while the 13C NMR resonances at δC 64.80 (C-13), δC 82.09 (C-14), 33.60 (C-15), 25.43 (C-16), δC 46.28 (C-17), and 116.21 (C-22) observed for 2 were replaced by those at δC 63.33, 81.41, 35.95, 28.69, 42.58, and 118.92 for these respective carbons of 1 (Table 1 and Tables S1 and S2, Supporting Information). These values indicated that 2 is the C-17 epimer of 1, as supported by its 2D HMBC NMR spectrum (Figure 1),11 which was comparable with that of 1.

Figure 1.

Figure 1.

COSY (Inline graphic, 1H→1H) (C-7/8 changed to normal bond, 010625) and key HMBC (Inline graphic, 1H→13C) and selected NOESY (Inline graphic, 1H→1H) correlations of and 2 and 3. The conformations shown for the NOESY correlations were generated based on the crystal structures of (−)-cryptanoside A (1)9 for 2 and (12R,16R,17S)-16,17-epoxy-12-hydroxypregna-4,6-diene-3,20-dione for 3.11

The overview NOESY profiles indicated that both 1 and 2 have the same configuration for their respective glycose moiety and the A-, B-, and C-ring system, but those for their D-ring proved to be different (Figure 1).9 The 1H NMR resonances at δH 3.52 and δH 2.65 for the OH-11 and OH-14 groups of 2, respectively, were assigned by the HMBC correlations observed between OH-11 and C-9, C-11, and C-12 and between OH-14 and C-8, C-14, and C-15. The NOESY correlations between H-11 and H-18 and H-19 and between OH-14 and H-18 of 2 indicated an α-oriented OH-11 and a β-oriented OH-14, while the NOESY correlations between H-17 and OH-14 and H-18 of 2 suggested a β-oriented H-17 (Figure 1). This configurational assignment for 2 was further supported by its ECD spectrum, which showed a positive Cotton effect (CE) around 291 nm and three negative CEs around 244, 226, and 203 nm (Figure 2).

Figure 2.

Figure 2.

ECD (left column) and UV (right column) spectra of compounds 1 (green), 1a (blue), 2 (orange), 3 (red), and 4 (purple). The ECD data were obtained in HPLC grade MeOH as the average of three scans corrected by subtracting a spectrum of the appropriate solution in the absence of the samples recorded under identical conditions. Each scan in the range 200–400 nm was obtained by taking points every 0.1 nm for 19 and every 1 nm for other compounds, with a 50 nm/min scanning speed and a 1 nm band width.

The positive CE around 291 nm for 2 is consistent with that observed for 1, which could result from the 7,8β-epoxy group and the 11α-hydroxy-12-oxo substitution and consequent conformational effects on the ECD spectrum. This indicated that both 1 and 2 have the same absolute configuration at C-5, C-7–C-11, C-13, and C-14. A 17α-lactone unit could be postulated for 2 from the negative ECD band around 244 nm, which is reversed from the positive CE around 241 nm observed for 1 that was proposed to correlate to its 17β-lactone unit.9 These ECD bands around 241 and 244 are indicative of exciton coupling arising from the α,β-unsaturated carbonyl chromophore, as supported by the UV absorption maxima at ca. 215 and ca. 218 nm (Figure 2).9,12 Thus, a (3S, 5S, 7S, 8R, 9S, 10S, 11S, 13R, 14R, 17S, 1’R, 3’S, 4’S, 5’S) absolute configuration could be assigned for 2, (−)-17-epi-cryptanoside A.

Compound 3 was isolated as an amorphous colorless powder, [α]20D +33 (MeOH) and +25 (CHCl3), with a molecular formula of C20H26O4, as shown by the protonated molecular peak at m/z 331.1901 observed in its HRESIMS spectrum. The IR spectrum of 3 indicated the presence of hydroxy (3408 cm−1), carbonyl (1704 cm−1) and conjugated carbonyl (1670 cm−1) and olefinic (1615 cm−1) groups, as supported by the absorption at 282.5 nm in its UV spectrum, which was suggestive of a conjugated dienone chromophore.13 The 1H and 13C NMR spectra revealed the presence of the NMR spectroscopic resonances at δH/δC 0.83 (s)/13.44 and 1.23 (s)/17.23 for two angular methyl groups, at δH/δC 4.43/69.91 for an oxygen-bearing carbon atom, at δH/δC 5.76/120.83, 6.16/127.72, and 6.16/141.48 for a dienone unit, and at δC 175.05 for a carboxyl group (Table 1). The methyl groups (C-18 and C-19) were placed at the C-13 and C-10 positions, respectively, as indicated by the HMBC correlations observed between H-18 and C-12 and C-17 and between H-19 and C-1 and C-5, while the oxygen-substituted carbon could be assigned as C-2, as supported by the HMBC correlations between H-2 and C-1 and C-3 and between H-4 and C-2, C-6, and C-10. The dienone unit (C-3–C-7) was located between C-2 and C-8 positions, as evidenced by the HMBC correlations observed between H-1 and C-3, H-6 and C-8, and H-7 and C-9 and C-14, while the carboxyl group was attached to the C-17 position, as implied by the HMBC correlations observed between H-17 and C-13, C-16, and C-20 (Figure 1). These NMR spectroscopic data are similar to those reported for 12β-hydroxypregna-4,6-diene-3,20-dione and 21-hydroxypregna-4,6-diene-3,20-dione (6,7-didehydrocortexone),14 suggesting a pregnane skeleton for 3.

Compound 4, (+)-2,21-dihydroxypregna-4,6-diene-3,20-dione, is a known dienone-containing pregnane, but its NMR spectra were not reported by any previous investigation. Its NMR spectra are closely similar to those of 6,7-didehydrocortexone, except for its 13C NMR spectroscopic resonance at δC 69.86 for C-2, which was reported at δC 34.1 for C-2 of 6,7-didehydrocortexone (Tables S1 and S2, Supporting Information),14 indicating that 4 should be 2-hydroxy-6,7-didehydrocortexone. Comparison of the 13C NMR spectroscopic data of 3 and 4 suggested that these compounds contain an identical steroid core (A–D rings). However, the 13C NMR spectroscopic resonance at δC 69.59 for C-21 of 4 was absent for 3, which thus showed the resultant 13C NMR spectroscopic resonances at δC 54.25 and 175.05 for C-17 and C-20 rather than those observed at δC 58.89 and 210.01 for these respective carbons of 4. Further comparison of the NMR spectroscopic data of 3 with those of 12β-hydroxypregna-4,6-diene-3,20-dione (cybisterol) implied that the substituents at the C-2, C-12, and C-20 positions of these compounds are different. A 13C NMR spectroscopic resonance at δC 30.2 for C-21 of cybisterol was missing for 3, while the resonances at δC 33.9 (C-2), 77.2 (C-12), 68.7 (C-17), and 213.8 (C-20) observed for cybisterol were substituted by chemical shifts at δC 69.91, 37.86, 54.25, and 175.05 for these respective carbons of 3 (Table 1).15 These indicated that 3 is a 21-nor pregna-4,6-dieno-3,20-dione, namely, (+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid. This conclusion was supported by the consistent NMR spectroscopic data reported for the 21-norpregnane, (17S)-4,6-androstadiene-3,12-dione-17-carboxylic acid.11

Steroids are non-flat molecules, with their A–C rings assuming a trans-fused chair conformation, which leads to a β orientation for both the C-18 and C-19 methyl groups.1 The NOESY correlations observed between H-2 and H-19, H-19 and H-8, and H-8 and H-18, as well as between H-17 and H-9 and H-14, indicated the β-oriented H-8, and C-18–C-20 and the α-oriented H-9, H-14, and OH-2 of 3 (Figure 1). Such a configurational assignment for 3 could then be further supported by its ECD spectrum, which showed a positive CE around 286 nm and a negative CE around 244 nm (Figure 2). The overview ECD curve of 3 is consistent with that reported for (17S)-4,6-androstadiene-3,12-dione-17-carboxylic acid, which shifted around 40 nm to the respective CEs around 350 and 280 nm, probably due to its C-12 carbonyl group.11 Thus, a (2R, 8S, 9S, 10R, 13S, 14S, 17S) absolute configuration could be assigned to both 3 and 4, based on their consistent ECD and 2D NOESY NMR spectra.

Previously, the cytotoxic (−)-cryptanoside A (1) was found to bind to NKA, with hydrogen bonding between its 11- and 4′-hydroxy groups and the enzyme being observed.9 To test the importance of these hydroxy groups in mediating the cytotoxic and NKA inhibitory activities of 1, its 11,4′-diacetate (1a) was prepared. Compound 1a gave a molecular formula of C34H46O12, as shown by the protonated molecular peak at m/z 647.3056 [M + H]+ observed in its HRESIMS spectrum. This molecular formula indicated the presence of two acetyl groups, as supported by the 1H NMR spectroscopic resonances at δH 2.10 and 2.20. In addition, these acetyl groups could attach to the C-11 and C-4′ position, as suggested by the resonances at δH 4.69 (H-4′) and δH 5.65 (H-11), which were different from those observed at δH 3.17 and 4.72 for these respective protons of 1. Also, the specific rotation value and the UV, 2D NOESY NMR, and ECD spectra of 1a were all consistent with those of 1. Thus, 1a was identified as (−)-11,4′-di-O-acetylcryptanoside A.

Compounds 14 and 1a were evaluated for their cytotoxicity against human HT-29 colon, MDA-MB-231 breast, and OVCAR3 and OVCAR5 ovarian cancer, and MDA-MB-435 melanoma cells, with paclitaxel as the positive control. Compound 1 exhibited potent cytotoxicity toward all cancer cell lines tested,9 which is comparable with that of digoxin,12 while 2 exhibited a selective activity toward OVCAR3 cells (IC50 7.0 μM), but all other compounds were inactive (Table 2). These indicated that the 17β-lactone unit and 11 and 4′ hydroxy groups are all important for 1 to mediate its cytotoxicity. Thus, the absence of the 17β-lactone unit, the C-3 glycose moiety, and 11 and 14 hydroxy groups in 3 and 4 may contribute to their non-cytotoxicity observed.

Table 2.

Cytotoxicity and NKA Inhibitory Activity of Compounds 1–4 and 1aa

compound HT-29 MDA-MB-231 OVCAR3 OVCAR5 MDA-MB-435 NKA docking scoreb

1 c 0.22 ± 0.23 0.50 ± 0.05 0.14 ± 0.05 0.44 ± 0.08 0.20 ± 0.12 1.2 ± 0.1 −9.2
1a >10 >10 >10 >10 >10 >100 −9.0
2 >10 >10 7.0 ± 0.1 >10 >10 40 ± 0.1 −9.2
3 >10 >10 >10 >10 >10 >100 −8.6
4 >10 >10 >10 >10 >10 NT d NT d
digoxinc 0.28 ± 0.03 0.31 ± 0.03 0.10 ± 0.01 NT d 0.17 ± 0.02 0.23 ± 0.02 −12.0
Paclitaxele 3.4 ± 0.1 3.8 ± 0.1 4.9 ± 0.1 7.3 ± 0.1 1.9 ± 0.1 NT d NT d
a

IC50 values are the concentrations (μM) required for 50% inhibition of viability of the human HT-29 colon, MDA-MB-231 breast, OVCAR3 and OVCAR5 ovarian cancer and MDA-MB-435 melanoma cell lines and of NKA activity for a given test compound. To test the cytotoxicity, the cells were treated by these compounds for a 72 h, and IC50 values were calculated using nonlinear regression analysis with measurements performed in triplicate and representative of three independent experiments.

b

Docking scores (kcal/mol) were obtained from binding between compounds selected and NKA calculated by AutoDock Vina (minimal, PDB: 4RET).

c

Data reported previously.9,12

d

NT: Not tested.

e

Positive control (IC50 nM).

Dienone-containing pregnanes are a small group of steroids, of which 12β-hydroxypregna-4,6-diene-3,20-dione and 20R-hydroxypregna-4,6-diene-3,20-dione (analogues of 3 and 4), along with several cardiac glycosides, have been isolated from the leaves of Nerium oleander,16 showing that these compounds can co-occur with cardiac glycosides. The structure of 12β-hydroxypregna-4,6,16-triene-3,20-dienone has been determined by analysis of its single-crystal X-ray diffraction data,17 while analysis of the cytotoxicity of some analogues showed the importance of their C-16 and C-17 double bond in mediating this type of activity.11,18 Consistently, 3 and 4 have a saturated C-16 and C-17 bond, and thus they were non-cytotoxic against a small panel of human cancer cell lines. Interestingly, an analogue, cyproterone acetate (trade name, Androcur) (5), has been used clinically for palliative care in prostate cancer and other androgen-dependent conditions.5

As many cardiac glycosides are well-known NKA inhibitors, compound 1 was reported previously for its potent NKA inhibitory activity.9 Following this, compounds 1a, 2, and 3 were tested using this enzyme-based bioassay in the present study, but they all were inactive (IC50 >20 μM). The NKA inhibitory potency of these compounds was consistent with their cytotoxicity observed, indicating that these compounds may exhibit their activity against human cancer cells through inhibition of NKA activity. Interestingly, compounds 1a and 3 seem to activate NKA at a high concentration (>100 μM) (Figure 3), suggesting that the NKA modulatory property of steroids may be enhanced by structural modification.

Figure 3.

Figure 3.

Inhibition of Na+/K+-ATPase by 1a, 2, and 3 and digitoxin.

Previously, compound 1 was found to bind to NKA, with the binding being not as strong as that observed for digoxin. However, the overall binding poses of 1 and digoxin are different, which could contribute to their different cytotoxic potency observed. Following these observations, the docking profiles for 1a, 2 and 3 in NKA have been investigated by AutoDock Vina, using a previously published procedure.9 Somewhat similar binding poses were observed for these compounds, while many subtle differences were found in their interactions with NKA (Figure 4), due to their various substituent groups.

Figure 4.

Figure 4.

Docking profile for 1a (left), 2 (middle), and 3 (right) in Na+/K+-ATPase (NKA).

When compared with 1, a similar binding pose was found for 2. However, the C-17 lactone unit of 2 adopts a different orientation, which prevents it from penetrating deeply in the active binding pocket of NKA, while some hydrogen bonds between 2 and NKA as found for 1 are missing. However, the orientation of the C-17 lactone unit of 2 causes rotation of its steroidal D-ring towards a hydrophobic surface formed by Leu793 and Ile800, which leads to the formation of a hydrogen bond between the C-14 hydroxy group and Thr797 and a π-π stacking interaction between the steroidal C-ring and Phe783. In addition, the β-face of 2 orients towards the hydrophilic surface formed by the αM1 and αM2 helices (Gln111, Glu117, Asp121, Asn122), and the saccharide moiety also interacts with several polar or charged residues of NKA, including Glu116, Glu312 and Arg880. Thus, 2 can bind to NKA, with the docking score being −9.2 kcal/mol, the same as 1 (Table 2).

The steroidal core of 1a adopts a similar binding pose as digoxin within the NKA pocket,12 with its α-face facing the hydrophobic surface formed by the αM4 and αM6 helices of NKA. However, the large C-11 substituent of 1a causes its steroid core to move toward the αM5 and αM6 helices to form a hydrogen bond between the C-14 hydroxy group and Thr797 of NKA and a hydrophobic interaction between its C-ring and Phe783 of NKA. In addition, the C-17 lactone unit of 1a can reach deep into the NKA pocket to form an interaction with Glu327, while its saccharide moiety has a chance to interact with several polar or charged residues of NKA, including Glu116, Glu312 and Arg880. Thus, 1a was found to bind to NKA (docking score, −9.0 kcal/mol) (Table 2).

The steroidal core of 3 adopts a binding pose similar to digoxin,12 but its small C-17 carbonyl group is not able to form sufficient interaction with residues at the bottom of the NKA pocket, including Glu327. However, a hydrogen bond formed between the C-20 carboxylic hydroxy group of 3 and the main chain of Gly319 of NKA would compensate its binding to NKA. Thus, this compound (3) binds to NKA, showing a docking score of −8.6 kcal/mol (Table 2).

All compounds 1a, 2, and 3 bind to NKA, with the docking scores being same or closely similar with that of 1, but these compounds were not found to exhibit cytotoxic and NKA inhibitory properties, as 1 did. While the overall docking poses for 1a, 2, and 3 are similar, but they are different from that observed for 1. In addition, the interactions of these compounds with specific residues in the NKA binding pocket vary, which may explain their differing bioactivities observed.

As minor steroids, 4,6-dienone-containing pregnanes have been isolated mainly form the Apocynaceae family, including the species Adenium obesum (Forssk.) Roem. & Schult., Apocynum venetum L. var. basikurumon Hara, Macrosiphonia petraea (A. St.-Hil.) Kuntze, Nerium oleander L., and Strophanthus divaricatus (Lour.) Hook. et Arn.,11,1318 of which the C-21-norpregnanes are unusual. A compound, namely, (17S)-4,6-androstadiene-3,12-dione-17-carboxylic acid, was isolated previously from the leaves and twigs of Strophanthus divaricatus, and its complete structure was determined by comparison of its spectroscopic data with those of the analogue, of which the structure has been determined by analysis of the single-crystal X-ray diffraction data.11 However, no bioactivities (included cytotoxicity) of this previously isolated C-21-norpregnane has been reported. Thus, in the present study, isolation of the new C-21-norpregnane, (+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3), from the stems of C. dubia, along with evaluation of its activities toward several different huma cancer cell lines and NKA, could be supportive of future investigations for this type of unusual steroids.

An early investigation showed that several 4,6-dienone-containing pregnanes exhibited toxic activities towards goldfish.13 Recently, this type of compounds were found to show cytotoxicity toward several human cancer cell lines, for which the existence of a Δ16 moiety seems to be necessary.11,18 Consistently, a Δ16 moiety is absent in 3 and 4, and these compounds were thus non-cytotoxic. As mentioned above, the synthetic analogue, cyproterone acetate (5), has been used clinically for palliative care in prostate cancer and other diseases. It has a dual mechanism of action, in being a strong antiandrogen but also exerting glucocorticoid effects, and its effectiveness has been established in monotherapy in comparison to standard forms of treatment.5 This indicates that further investigations on the 4,6-dienone-containing pregnanes, including 3 and 4, may be helpful for the development of new hormone therapy and palliative care for the treatment of cancer and other diseases.

EXPERIMENTAL SECTION

General Experimental Procedures.

The specific rotations were measured at room temperature on a MCP 150 polarimeter (Anton Paar). The ultraviolet (UV) spectra were recorded on a Hitachi U2910 ultraviolet spectrophotometer. Electronic circular dichroism (ECD) spectroscopic measurement was performed using a JASCO J-810 spectropolarimeter. The infrared (IR) spectra were recorded on a Nicolet 6700 FT-IR spectrometer (Thermo Scientific). The NMR spectra were run at room temperature on a Bruker Avance III 400 or a Bruker Avance III HD 700 MHz NMR spectrometer. ESIMS and HRESIMS data were collected on a Bruker Maxis 4G Q-TOF mass spectrometer in the positive-ion mode. Column chromatography was conducted using silica gel (65 × 250 or 230 × 400 mesh, Sorbent Technologies). Analytical TLC was performed on precoated silica gel 60 F254 plates (Sorbent Technologies). Sephadex LH-20 was purchased from Amersham Biosciences. For visualization of TLC plates, H2SO4 was used as a spray reagent. All procedures were carried out using solvents purchased from commercial sources and employed without further purification. Digitoxin and paclitaxel were obtained from Sigma-Aldrich (purity ≥98%).

Plant Material.

The stems of Cryptolepis dubia (Burm.f.) M.R. Almeida (Apocynaceae) (acquisition number AAA07194) was identified by D.D.S. and collected in November 2018 by D.D.S., K.S., and M.X. (voucher specimen: DDS 15365). This plant sample was collected under a collaborative arrangement between the University of Illinois at Chicago (USA) and the Institute of Traditional Medicine, Ministry of Health, Vientiane, Laos.9

Extraction and Isolation.

As reported previously,9 the milled air-dried stems of Cryptolepis dubia (sample AAA07194, 1935 g) were extracted with MeOH followed by partition with n-hexane and then CH2Cl2. The active CH2Cl2 extract (6.8 g) was subjected to silica gel column chromatography to give 32 pooled fractions (D2F1–D2F32), of which fractions D2F9–D2F11 (650 mg, IC50 around 2 μg/mL ) were used in isolation of the cytotoxic (−)-cryptanoside A (1).9 Following this, fractions 14 and 15 (1.1 g, IC50 2−20 μg/mL) that showed less potent cytotoxicity than fractions D2F9–D2F11 were chromatographed over a silica gel column (4.5 × 40 cm), eluted with a gradient of CH2Cl2-MeOH. Eluates were pooled by TLC analysis to give four major fractions (D2F15F1–D2F15F4), of which fraction D2F15F2 (220 mg) was found to contain steroidal components. This fraction (D2F15F2) was thus subjected to silica gel column chromatography (2.5 × 45 cm) and eluted with a gradient of CH2Cl2-acetone, followed by separation over a Sephadex LH-20 column eluted with CH2Cl2–MeOH (1:1), affording (−)-cryptanoside A (1, 25.0 mg), (−)-17-epi-cryptanoside A (2, 2.0 mg), (+)-2-hydroxyandrosta-4,6–12 diene-3-one-17-carboxylic acid (3, 1.0 mg), and (+)-2,21-dihydroxypregna4,6-diene-3,20-dione (4, 4.0 mg).

(−)-17-epi-Cryptanoside A (2):

Colorless powder; [α]20D −10 (c 0.03, MeOH); [α]20D −23 (c 0.04, CHCl3); UV (MeOH) λmax (log ε) 218 (4.33) nm; ECD (MeOH) λmax (Δε) 203 (−1.44), 226 (−4.53), 244 (−1.21), 291 (+4.58) nm; IR (dried film) νmax 3457, 2924, 1784, 1744, 1709, 1623, 1445, 1380, 1349, 1297, 1143, 993, 899, 733 cm−1; 1H and 13C NMR data, Tables 1 and 2; HRESIMS m/z 563.2825 [M + H]+ (calcd for C30H42O10H+, 563.2851).

(+)-2-Hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3):

Colorless powder; [α]20D +33 (c 0.02, MeOH); [α]20D +25 (c 0.02, CHCl3); UV (MeOH) λmax (log ε) 282.5 (3.69) nm; ECD (MeOH) λmax (Δε) 244 (−0.36), 286 (+0.89) nm; IR (dried film) νmax 3408, 2921, 2851, 1704, 1670, 1615, 1456, 1380, 1219, 1166, 1088, 895, 872, 720 cm−1; 1H and 13C NMR data, Tables 1 and 2; HRESIMS m/z 331.1901 [M + H]+ (calcd for C20H26O4H+, 331.1903).

(+)-2,21-Dihydroxypregna-4,6-diene-3,20-dione (4):

Colorless powder; [α]20D +70 (c 0.03, MeOH); [α]20D +53 (c 0.04, CHCl3); UV (MeOH) λmax (log ε) 282.5 (3.95) nm; ECD (MeOH) λmax (Δε) 242 (−0.46), 292 (+3.02) nm; IR (dried film) νmax 2916, 2849, 1714, 1667, 1616, 1264, 1088, 895, 734, 704 cm−1; 1H and 13C NMR data, Tables 1 and 2; HRESIMS m/z 367.1879 [M + Na]+ (calcd for C21H28O4Na+, 367.1880).

Acetylation of (−)-Cryptanoside A (1).

To a dried 25 mL glass vial equipped with a magnetic stirrer, containing 5.6 mg (0.01 mmol) of (−)-cryptanoside A (1), were added 200 μL of Ac2O and 200 μL of pyridine. The vial was sealed and stirred at 80 °C for 1.5 h and cooled to room temperature. Then, CH2Cl2 (5 mL) was added, and the solution was extracted with distilled H2O. The organic layer was washed with distilled H2O and evaporated at reduced pressure. The residue was purified by silica gel column chromatography, using n-hexane–acetone (5:1 → 1:1), followed by elution over a Sephadex LH-20 column using CH2Cl2−MeOH (1:1) to afford 5.3 mg (0.0081 mmol) of (−)-11,4′-di-O-acetylcryptanoside A (1a) (yield: 80.0%).

(−)-11,4′-Di-O-acetylcryptanoside A (1a):

Colorless powder; [α]20D −15 (c 0.06, MeOH); [α]20D −24 (c 0.08, CHCl3); UV (MeOH) λmax (log ε) 214 (4.21) nm; ECD (MeOH) λmax (Δε) 205 (+3.33), 219 (+1.03), 239 (+0.54), 286 (+1.98) nm; IR (dried film) νmax 3489, 2931, 1780, 1743, 1630, 1442, 1373, 1231, 1128, 1098, 1029, 890, 863, 734 cm−1; 1H and 13C NMR data, Tables 1 and 2; HRESIMS m/z 647.3056 [M + H]+ (calcd for C34H46O12H+, 647.3062).

Cytotoxicity Assay.

Following a previously published protocol,9 the human HT-29 colon, MDA-MB-231 breast, and OVCAR3 and OVCAR5 ovarian cancer, and MDA-MB-435 melanoma cell lines were cultured in RPMI 1640 or DMEM media, and supplemented with fetal bovine serum (FBS) (10%), penicillin (100 units/mL), and streptomycin (100 μg/mL).The cytotoxicity of the compounds obtained was screened, with the vehicle and paclitaxel used as the negative and positive control, respectively. The viability of cells was evaluated using a commercial absorbance assay (CellTiter-Blue Cell Viability Assay, Promega Corp.), with the IC50 values calculated from the vehicle control.

Na+/K+-ATPase Activity Assay.

Na+/K+-ATPase (NKA) activity was assessed using a luminescent ADP detection assay (ADP-Glo Max Assay; Promega) that measures enzymatic activity by quantitating the ADP produced during the enzymatic first half-reaction.9

Molecular Modeling.

Following previous procedures,9 the crystal structure of Sus scrofa NKA (sNKA) was obtained from the Protein Data Bank (PDB, 4RET) and used as the receptor, and the conformations of 1a, 2, and 3 generated by LigPrep. Geometric optimization was performed using the OPLS3 force field with all possible ionization states at pH 7.4 ± 0.1 created by Epik. Molecular docking process against the receptor was conducted with Autodock Vina, which were analyzed by PyMol.

Statistical Analysis.

The in vitro measurements were performed in triplicate and are representative of three independent experiments. The dose response curve was calculated for IC50 determinations using non-linear regression analysis (Table Curve2DV4; AISN Software Inc.).

Supplementary Material

Supporting Information

ACKNOWLEDGMENTS

This investigation was supported by grant P01 CA125066 and grant K12 GM139186 and K99 GM155616, awarded to A. D. Kinghorn by the National Cancer Institute and to E. M. Kaweesa by the National Institute of General Medical Sciences, respectively, NIH, Bethesda. MD, USA. Drs. D. Uchenik, D. Krishnan, and J. C. Gallucci, College of Pharmacy, and Drs. C. Yuan, A. Somogy, and G. Wu, Campus Chemical Instrument Center, The Ohio State University, are thanked for access to the instrumentation used and for helpful information about the conformations and crystal structure of compounds presented.

Footnotes

Notes

The authors declare no competing financial interest.

Supporting Information

The Supporting Information is available free of charge on the ACS Publications website, including IR, mass, and NMR spectra of 24 and 1a, the assigned 1H and 13C NMR spectroscopic data of 1, 1a, and 4, COSY and key HMBC and selected NOESY correlations of 1a and 4, and the dose-response curves for cytotoxicity of 14 and 1a.

Data Availability Statement

The NMR data of 2 and 3 have been deposited in the Natural Products Magnetic Resonance Database (NP-MRD; www.np-mrd.org) and can be found at NP0350563 for (−)-17-epi-cryptanoside A (2) and NP0350564 for (+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3).

REFERENCES

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information

Data Availability Statement

The NMR data of 2 and 3 have been deposited in the Natural Products Magnetic Resonance Database (NP-MRD; www.np-mrd.org) and can be found at NP0350563 for (−)-17-epi-cryptanoside A (2) and NP0350564 for (+)-2-hydroxyandrosta-4,6-diene-3-one-17-carboxylic acid (3).

RESOURCES