Bioorganic & Medicinal Chemistry Letters xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Triterpenes from Alisma orientalis act as androgen receptor agonists, progesterone receptor antagonists, and glucocorticoid receptor antagonists Hsiang-Ru Lin ⇑ Department of Chemistry, College of Science, National Kaohsiung Normal University, No. 62, Shenjhong Rd., Yanchao District, Kaohsiung City 82446, Taiwan, Republic of China
a r t i c l e
i n f o
Article history: Received 28 January 2014 Revised 27 April 2014 Accepted 8 May 2014 Available online xxxx Keywords: Nuclear receptor Natural product Androgen receptor Progesterone receptor Glucocorticoid receptor Alisol M 23-acetate Alisol A 23-acetate
a b s t r a c t Alisma orientalis, a well-known traditional medicine, exerts numerous pharmacological effects including anti-diabetes, anti-hepatitis, and anti-diuretics but its bioactivity is not fully clear. Androgen receptor (AR), progesterone receptor (PR), and glucocorticoid receptor (GR) are three members of nuclear receptor superfamily that has been widely targeted for developing treatments for essential diseases including prostate cancer and breast cancer. In this study, two triterpenes, alisol M 23-acetate and alisol A 23-acetate from Alisma orientalis were determined whether they may act as androgen receptor (AR), progesterone receptor (PR), or glucocorticoid receptor (GR) modulators. Indeed, in the transient transfection reporter assays, alisol M 23-acetate and alisol A 23-acetate transactivated AR in dose-dependent manner, while they transrepressed the transactivation effects exerted by agonist-activated PR and GR. Through molecular modeling docking studies, they were shown to respectively interact with AR, PR, or GR ligand binding pocket fairly well. All these results indicate that alisol M 23-acetate and alisol A 23-acetate from Alisma orientalis might possess therapeutic effects through their modulation of AR, PR, and GR pathways. Ó 2014 Elsevier Ltd. All rights reserved.
Nuclear receptors (NRs) act as transcription factors to modulate transcription actions of target genes involved in maintenance of cellular phenotypes, metabolism, and cell proliferation through homodimerizing or heterodimerizing with other nuclear receptors. Currently, there are more than 30 members in the nuclear receptor superfamily.1 Most of these receptors exert their functions by ligand-activation and they usually contain six functional domains (A–F) including the first transcription activation domain (AF-1, A/ B), DNA binding domain (DBD, C), hinge domain (D), and the second transcription activation domain (AF-2, E/F) also referred to as the ligand binding domain (LBD). Because of their importance in many diseases including breast cancer, prostate cancer, and diabetes diseases, NRs are viewed as potential targets to develop specific therapeutic agents. Androgens including testosterone and dihydrotestosterone (DHT) are steroid hormones that are responsible for the cellular proliferation and differentiation of male sexual organs and secondary sexual characteristics, and they mainly exerted their actions through its respective nuclear receptor, androgen receptor (AR).2 Androgen primarily exerts its action via a genomic mechanism in which androgen passively enters the target cells and binds to AR ⇑ Tel.: +886 7 7172930x7123; fax: +886 7 6051083. E-mail address: [email protected]
in the cytoplasm.3 The androgen/AR complex further translocates into the nucleus, in which the AR complex dimerizes and binds to the promoter region of the androgen-regulated gene to initiate the transcription action and enhance the expression of androgenregulated protein such as PSA, Bcl-2, and maspin.4–6 Generally, androgen plays an essential role in the regulation of men’s health and androgen deficiency results in numerous significant diseases including diabetes and osteoporosis when men enter andropause period.7 Androgen also plays a moderate role in the regulation of physiological functions in women. For example, endometriosis, a benign gynecological disorder, frequently occurs in women and synthetic androgen like danazol (1 in Fig. 1) is one of the endometriosis treatments.8,9 All the evidences mentioned above indicated that the herbal AR agonists should be suitable treatments or supplements for androgen deficiency related diseases and endometriosis. Progesterone plays a critical role in the development, differentiation and function of female reproductive tissues, while it exerted biological effects mainly through agonizing its associated receptor, progesterone receptor (PR), a member of NR family. PR basically behaves as an agonist-activated transcription factor to regulate the expression of essential genes like Bcl-Xl and transforming growth factor (TGF).10,11 However, it is evident that progesterone increases breast cancer risk and PR is involved in the progression of breast cancer.12 Apart from breast cancer, leiomyoma is a benign
http://dx.doi.org/10.1016/j.bmcl.2014.05.039 0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
2
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
O OH
OH HO 11
O
12
23
H H
H
3
N
A
O
16
O
24 25
H
O O
2
1
O
N O
OH
OH
HO
H OH H
H
O
O
4
3 F
F
HO N O
O
O
OH
H H O O
5
6
Figure 1. Chemical structures of danazol (1), alisol M 23-acetate (2), alisol A-23 acetate (3), RU486 (mifepristone, 4), EM5744 (5), and asoprisnil (6).
smooth muscle neoplasm that frequently occurs in uterus. Uterine leiomyoma is worldly prevalent for women and progesterone plays a key role to maintain the growth of uterine leiomyoma.13 Progesterone receptor antagonists have been demonstrated to decrease uterine leiomyoma volume and reduce the bleeding days of tested patients.14 Since progesterone is highly involved in the proliferation of breast cancer and uterine leiomyoma, the herbal antiprogestins would be the effective treatments or supplements for uterine leiomyoma and breast cancer in women. Glucocorticoid receptor (GR) is another member of NR superfamily and the ligand-activated GR homodimerizes to act as an active transcription factor for facilitating the transcription action of regulated genes.15 Glucocorticoid, the natural ligand of GR, has extensive effects on different tissues and the synthetic glucocorticoids have long been used in the treatments of asthma, arthritis, and autoimmune diseases due to their outstanding anti-inflammatory and immunosuppressive effects.16 Although synthetic glucocorticoids have significant anti-inflammatory activity, chronic glucocorticoid therapy has deleterious side effects, including weight gain, osteoporosis, and diabetes mellitus.17 In liver, glucocorticoid receptor antagonists were shown to antagonize insulin’s
action through regulating genes involved in the gluconeogenesis pathway like phosphoenolpyruvate carboxykinase (PEPCK) so the herbal glucocorticoid receptor antagonists might ameliorate insulin resistance and help type II diabetes patients.18 Alisma orientalis (SAM.) JUZEP is widely cultivated in Taiwan, China as well as Japan, and its dried rhizomes are employed as the treatments of diabetes and diuretics. Alisma orientalis is also an important component of a herbal combinatorial drug which is widely utilized to treat hyperglycemia, hyperlipidemia, neuroprotection, and nephritis in Chinese society. The chemical constitutions of Alisma orientalis mainly include sesquiterpenes and protostane type triterpenes like alisol derivatives.19,20 Among them, alisol A and its derivatives have been reported to inhibit Hepatitis B virus (HBV) activity but this activity is present by the secretion of HBV e antigen (HBVeAg).21 Additionally, Alisma orientalis was shown to contain FXR agonists.22 However, Alisma orientalis might exert the pharmacological effects through other molecular mechanisms. In this study, the isolated triterpenes from Alisma orientalis were evaluated for their activity against AR, PR, and GR. The study goal is to determine whether Alisma orientalis might exert therapeutic effect through AR, PR, and GR pathways.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
3
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
6
Relative Luciferase Activity
5
4
3
2
100nΜ AA23a
1μΜ AA23a
10μΜ AA23a
100nΜ AM23a
1μΜ AM23a
10μΜ AM23a
10μΜ Danazol
DMSO
1
Figure 2. Transient transfection reporter assay in HepG2 cells. AR agonistic effect was determined by the MMTV promoter-driven luciferase activity. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of DMSO. The RLA data represented mean ± standard deviation for at least three individual determinations. AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
1.0
Relative Luciferase Activity
0.8
0.6
0.4
0.2
100nΜ AA23a + 10nM Prog
1μΜ AA23a + 10nM Prog
10μΜ AA23a + 10nM Prog
100nΜ AM23a + 10nM Prog
1μΜ AM23a + 10nM Prog
10μΜ AM23a + 10nM Prog
1μΜ RU486 + 10nM Prog
0.0
10nM Prog
The isolation of the protostane-type triterpenes employed the ethyl acetate (EtOAc) soluble fraction of Alisma orientalis that was processed by flash chromatography methods to generate three major constituents, alisol M 23-acetate (2 in Fig. 1), alisol A 23-acetate (3 in Fig. 1), and alisol B.23 Their structures were identified by comparing NMR spectral data with literature values.24–26 Among these isolated triterpenes, alisol B was not evaluated for its activity due to its cytotoxic effect on HepG2 cells, while alisol M 23-acetate and alisol A 23-acetate showed no influence for the growth of HepG2 cells. Primarily, to evaluate the agonistic activity of alisol M 23-acetate and alisol A 23-acetate against AR, the transient transfection reporter assay with a reporter plasmid containing androgen receptor response element (pMMTV-luc) was employed.27 For comparison, the transactivation activity of vehicle (DMSO, RLA = 1) was used as the standard reference and 10 lM danazol (RLA = 3.8) was utilized as the positive control. As shown in Figure 2, alisol M 23-acetate exerted androgenic activity about 4-fold of vehicle activity at 10 lM, while alisol A 23-acetate possessed less androgenic effect. Both compounds exerted weak androgenic effect at 1 lM and completely lost their AR agonism at 100 nM. This result strongly supported that alisol M 23-acetate and alisol A 23-acetate could act as AR agonists as danazol. To determine the agonistic and antagonistic activity of alisol M 23-acetate and alisol A 23-acetate against PR, the transient transfection reporter assay with a reporter plasmid containing progesterone receptor response element (pMMTV-luc) was utilized.27 For comparison, 10 nM progesterone was cotreated with test compounds and its activity was set to 1. RU486 (6% of 10 nM progesterone agonistic effect, 4 in Fig. 1) at 1 lM, was employed as the positive control. Alisol M 23-acetate and alisol A 23-acetate did not exert significant agonistic effect against PR at 20 lM (data not shown). As shown in Figure 3, alisol M 23-acetate and alisol A 23-acetate antagonized PR to 64% and 66% of 10 nM progesterone agonistic effect respectively at 10 lM. Both compounds exerted antagonistic activity of 71% and 82% of 10 nM progesterone agonistic effect at 1 lM. However, alisol M 23-acetate lost its antagonistic activity but alisol A 23-acetate exhibited weak PR
Figure 3. Transient transfection reporter assay in HepG2 cells. PR antagonistic effect was determined by the MMTV promoter-driven luciferase activity in the presence of 10 nM progesterone. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of 10 nM progesterone. The RLA data represented mean ± standard deviation for at least three individual determinations. Prog: progesterone; AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
antagonism at 100 nM. This result showed that both alisol derivatives could serve as PR antagonists. To determine the agonistic and antagonistic activity of alisol M 23-acetate and alisol A 23-acetate against GR, the transient transfection reporter assay with a reporter plasmid containing glucocorticoid receptor response element (pGRE-luc) was utilized.27 For comparison, 5 nM dexamethasone was cotreated with test compounds and its activity was set to 1. RU486 (30% of 5 nM dexamethasone agonistic activity) at 1 lM, was employed as the positive control. Alisol M 23-acetate and alisol A 23-acetate did not exhibit obvious agonistic effect against GR at 20 lM (data not shown). Figure 4 outlined that alisol M 23-acetate and alisol A 23-acetate antagonized GR to 30% and 39% of 5 nM dexamethasone agonistic activity respectively at 10 lM. The antagonistic activity of both compounds dropped to 56% of 5 nM dexamethasone effect at 1 lM. However, alisol M 23-acetate lost its antagonistic activity but alisol A 23-acetate still exerted weak inhibition against GR at 100 nM. This result indicated that both alisol derivatives could act as GR antagonists. Finally, the computer-based modeling program (Discovery Studio 2.1, Accelrys, San Diego, USA) was employed to generate docking models that were further used to discuss how alisol M 23-acetate and alisol A 23-acetate might interact with AR LBD, PR LBD, and GR LBD respectively. The crystal structure of most NR LBD complexed with its modulator contains 12 a-helices to form a ligand binding pocket in which helices 4, 5, 8, and 9 are sandwiched between helices 1, 3, 7, and 10 at the top half of the pocket. Importantly, helices 3, 7, and 11 form a large cavity in the bottom half of the pocket to allow the NR modulator to locate. In the agonistic conformation, once a NR agonist bound to the pocket and located in the cavity, it subsequently induced the movement of helice 12 to the entrance of pocket and exposed a coactivator-binding surface. Traditionally, most AR agonists contain a steroidal skeleton and the addition of a large scaffold on its C12, C13 or C17 will generate the AR antagonists. However, in
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
4
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
1.8 1.6
Relative Luciferase Activity
1.4 1.2 1.0 0.8 0.6 0.4 0.2
100nΜ AA23a + 5nM Dex
1μΜ AA23a + 5nM Dex
10μΜ AA23a + 5nM Dex
100nΜ AM23a + 5nM Dex
1μΜ AM23a + 5nM Dex
10μΜ AM23a + 5nM Dex
1μΜ RU486 + 5nM Dex
5nM Dex
0.0
Figure 4. Transient transfection reporter assay in HepG2 cells. GR antagonistic effect was determined by the GRE promoter-driven luciferase activity in the presence of 5 nM dexamethasone. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of 5 nM dexamethasone. The RLA data represented mean ± standard deviation for at least three individual determinations. Dex: dexamethasone; AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
these years, EM5744 (5 in Fig. 1), an AR agonist containing a large substituent on the C13 of its steroidal core scaffold, revealed a new binding mode of AR agonist in AR LBD.28 Generally, the large side chain on the C12, C13 or C17 of the AR antagonist’s steroidal scaffold leans toward the entrance of AR ligand binding pocket and forces helice 12 to locate at the inactive position. For example, bicalutamide contains a side chain, 4-fluorophenylsulfonyl group, to protrude out of AR ligand binding pocket and prevent helice 12 to locate at the agonistic position. Interestingly, in the EM5744/AR LBD crystal structure, the side chain of EM5744 located in the channel formed by helices 3, 11, and 12. The 3-keto and 17-hydroxyl groups of EM5744 interacted with AR LBD by hydrogen bond interaction as DHT. Importantly, the terminal fluoro substituent of EM5744’s side chain made contact with His 874 of helice 12 and this interaction was not shown in the DHT/ AR LBD structure. In fact, helice 12 in the EM5744/AR LBD structure exhibited some flexibility and moved a little bit away from the AR ligand binding pocket so the AR LBD had the flexibility to accommodate large ligands by changing the position of the side chain of helice 12 and other helices. To determine the binding modes of alisol M 23-acetate and alisol A 23-acetate in AR LBD, the ligandfit docking program and the crystal structure of human AR LBD complexed with EM5744 (Protein Data Bank ID: 2PNU) were utilized. Due to similar binding situations for alisol M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. As shown in Figure 5A and B, the docking model of alisol M 23-acetate indicated that alisol M 23-acetate resided in AR ligand binding pocket in a similar manner as EM5744 by leaning its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5A and B) and locating its 23-acetyl group nearby helice 11(as shown in yellow color in Fig. 5A and B).
Importantly, the side chain of alisol M 23-acetate inserted into the same channel as the side chain of EM5744 fairly well. Alisol M 23-acetate mainly contacted with AR LBD by hydrophobic interaction and its 3-keto moiety did not form any hydrogen bond interaction with AR LBD. Alternatively, the 16-keto group of alisol M 23-acetate served as a hydrogen bond acceptor to make contact with the side chain of Thr 877. This hydrophilic contact stabilized the insertion of the large side chain of alisol M 23-acetate and the overall agonistic alisol M 23-acetate/AR LBD conformation. Furthermore, the lack of hydrogen bond interaction formed by 3-keto moiety also interpreted the similar AR agonism exerted by alisol M 23-acetate and danazol. Moreover, alisol A 23-acetate had similar binding mode as alisol M 23-acetate but it exerted slightly less AR agonism than alisol M 23-acetate. The primary structural difference between both alisol derivatives is the terminal portion of their C17 side chain which was surrounded by hydrophobic residues of AR LBD including Val 716, Leu873, Ile 898, Val 901, and Val 903. Since there is no hydrogen bond interaction between their side chains and AR LBD, the unfavorable insertion of hydrophilic substituents of alisol A 23-acetate into a hydrophobic zone might account for its less AR agonism. In conclusion, this docking result highly supported that both alisol derivatives could serve as AR agonists. Most NR antagonists have different structural requirement due to their distinct interaction with their respective receptor ligand binding pocket. Once a NR antagonist bound to the pocket, it subsequently induced the movement of helice 12 to pack antiparallel to helice 11 and further prevented the formation of the coactivator-binding surface or facilitated corepressor-binding. Some novel PR antagonists like asoprisnil (6 in Fig. 1), contain a steroidal-like skeleton as the core scaffold with 3-keto group and an antiprogestin side chain located at 11b position of the steroidal scaffold.29 Asoprisnil located in PR LBD, while its 3-keto moiety oriented toward helice 3 and its C11 side chain located in the channel formed by helices 3 and 11. Primarily asoprisnil contacted with PR LBD by hydrophobic interaction and made two hydrogen bond interactions with PR LBD including (1) its 3-keto moiety acted as the hydrogen bond acceptor to form the hydrogen bond interaction with Gln 725 and Arg 766; (2) its terminal oxime group on C11 side chain acted as hydrogen bond acceptors to make contact with two water molecules that also interacted with Glu 723 and Asn 719 to form hydrogen bond networks. Essentially, asoprisnil as other PR antagonists antagonized PR by its antagonistic side chain to protrude out of PR ligand binding pocket and prevent helice 12 from locating at the active position. To determine the binding modes of alisol M 23-acetate and alisol A 23-acetate in PR LBD, the ligandfit docking program and the crystal structure of human PR LBD/ asoprisnil (Protein Data Bank ID: 2OVH) were utilized. Due to similar binding situations for alisol M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. Figure 5C and D outlined the binding mode of alisol M 23-acetate in PR LBD, in which alisol M 23-acetate leaned its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5C and D) and located its 23-acetyl group nearby helice 11 (as shown in yellow color in Fig. 5C and D). Importantly, unlike asoprisnil, alisol M 23-acetate only made one hydrogen bond interaction by its 3-keto group serving as a hydrogen bond acceptor to contact with the terminal amino group of Gln 725. This binding difference might be resulted from alisol M 23-acetate’s bulky C4 dimethyl substituents that were closed to helice 5 and prevented the interaction between alisol M 23-acetate’s 3-keto group and Arg 766. Additionally, although part of helice 12 of hPR LBD/asoprisnil structure was missing, it cannot be excluded that the side chain of alisol M 23-acetate might contact with helice 12 of PR LBD. Especially, the structure–activity relationship study for 19-nortestosterone and RU486 derivatives mentioned that the addition of five- or
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
5
Figure 5. A: The binding model of alisol M-23 acetate in AR ligand binding pocket. B: The hydrophilic interactive mode of alisol M-23 acetate in AR ligand binding pocket. C: The binding model of alisol M-23 acetate in PR ligand binding pocket. D: The hydrophilic interactive mode of alisol M-23 acetate in PR ligand binding pocket. E: The binding model of alisol M-23 acetate in GR ligand binding pocket. F: The hydrophilic interactive mode of alisol M-23 acetate in GR ligand binding pocket. Oxygen atom in red color and hydrogen atom in white color. Green line indicates the hydrogen bond interaction.
six-membered substituent on the C17 of steroidal progestins and antiprogestins could enhance the binding affinity and progestin/ antiprogestin activity, so the complete binding mode of alisol M 23-acetate in PR LBD might still need the generation of hPR LBD/ alisol M 23-acetate crystal structure. RU486, a novel GR antagonist, contains a steroidal-like core scaffold with a 3-keto group and a side chain located at 11b position of the steroidal scaffold. RU 486 resided in GR LBD, while its 3-keto moiety oriented toward helice 3 and its C11 side chain located in the channel formed by helices 3 and 11.30 Primarily
RU486 contacted with GR LBD by hydrophobic interaction and made the hydrogen bond interactions with GR LBD by its 3-keto group acting as the hydrogen bond acceptor to interact with Gln 570 and Arg 611. Moreover, the C11 side chain of RU486 leaned closed to helice 3 and protruded out of GR ligand pocket but did not make any significant hydrophilic contact with helice 12 of GR LBD. To further determine the binding modes of alisol M 23-acetate and alisol A 23-acetate, the crystal structure of human GR LBD/RU486 (Protein Data Bank ID: 3H52) was utilized to perform the docking study. Due to similar binding situations for alisol
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
6
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. As shown in Figure 5E and F, the binding mode of alisol M 23-acetate in GR LBD outlined that alisol M 23-acetate located its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5E and F) and its 23-acetyl group nearby helice 11 (as shown in yellow color in Fig. 5E and F). Importantly, alisol M 23-acetate mainly contacted with GR LBD by hydrophobic interaction and made two hydrogen bond interactions including (1) its 3keto group served as a hydrogen bond acceptor to contact with the terminal amino group of Gln 570; (2) its 11-hydroxyl group acted as a hydrogen bond donor to contact with the backbone of Leu 563. Furthermore, the 23-acetyl side chain of alisol M 23-acetate located in the center of the channel formed by helices 3 and 11 to cause helice 12 of GR LBD shifting to the antagonistic position. The results from the docking studies showed that alisol M 23-acetate and alisol A 23-acetate could act as GR antagonists. AR, PR, and GR are three members of NRs to be viewed as the target proteins for developing multi-disease treatments by pharmaceutical industry. In this study, it is the first report to demonstrate that alisol M 23-acetate and alisol A 23-acetate served as AR agonists, PR antagonists, and GR antagonists so Alisma orientalis might exert therapeutic effects through AR, PR, and GR pathways. Acknowledgments I thank Dr. Tsai S.Y. at Baylor College of Medicine for kindly offering hPR expression vector and Dr. Gelmann E.P. for generously offering pMMTV-luc vector. References and notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Olefsky, J. M. J. Biol. Chem. 2001, 276, 36863. Culig, Z.; Klocker, H.; Bartsch, G.; Hobisch, A. Endocr. Relat. Cancer 2002, 9, 155. Gao, W.; Bohl, C. E.; Dalton, J. T. Chem. Rev. 2005, 105, 3352. Luke, M. C.; Coffey, D. S. J. Androl. 1994, 15, 41. Huang, H.; Zegarra-Moro, O. L.; Benson, D.; Tindall, D. J. Oncogene 2004, 23, 2161. Zhang, M.; Magit, D.; Sager, R. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 5673. Theodoraki, A.; Bouloux, P. M. Menopause Int. 2009, 15, 87. Fedele, L.; Berlanda, N. Expert Opin. Emerg. Drugs 2004, 9, 167. Vercellini, P.; Somigliana, E.; Viganò, P.; Abbiati, A.; Barbara, G.; Crosignani, P. G. Drugs 2009, 69, 649. Moore, M. R.; Conover, J. L.; Franks, K. M. Biochem. Biophys. Res. Commun. 2000, 277, 650. Gong, Y.; Anzai, Y.; Murphy, L. C.; Ballejo, G.; Holinka, C. F.; Gurpide, E.; Murphy, L. J. Cancer Res. 1991, 51, 5476. Moore, M. R. Curr. Cancer Drug Targets 2004, 4, 183. Kim, J. J.; Kurita, T.; Bulun, S. E. Endocr. Rev. 2013, 34, 130. Engman, M.; Granberg, S.; Williams, A. R.; Meng, C. X.; Lalitkumar, P. G.; Gemzell-Danielsson, K. Hum. Reprod. 2009, 24, 1870. Oakley, R. H.; Cidlowski, J. A. J. Allergy Clin. Immunol. 2013, 132, 1033. Quax, R. A.; Manenschijn, L.; Koper, J. W.; Hazes, J. M.; Lamberts, S. W.; van Rossum, E. F.; Feelders, R. A. Nat. Rev. Endocrinol. 2013, 9, 670. Buehring, B.; Viswanathan, R.; Binkley, N.; Busse, W. J. Allergy Clin. Immunol. 2013, 132, 1019. Taylor, A. I.; Frizzell, N.; McKillop, A. M.; Flatt, P. R.; Gault, V. A. Horm. Metab. Res. 2009, 41, 899. Peng, G. P.; Tian, G.; Huang, X. F.; Lou, F. C. Phytochemistry 2003, 63, 877. Hu, X. Y.; Guo, Y. Q.; Gao, W. Y.; Zhang, T. J.; Chen, H. X. J. Asian Nat. Prod. Res. 2008, 10, 481. Jiang, Z. Y.; Zhang, X. M.; Zhang, F. X.; Liu, N.; Zhao, F.; Zhou, J.; Chen, J. J. Planta Med. 2006, 72, 951. Lin, H. R. Bioorg. Med. Chem. Lett. 2012, 22, 4787. The dried and powdered Alismatis rhizoma (3 kg) was extracted with n-hexane, ethyl acetate and ethanol separately at room temperature. The extracts were concentrated in vacuo to yield crude oil. The combined ethyl acetate fraction (10 g) was chromatographed over SiO2 column (70–230 mesh) using gradient n-hexane-ethyl acetate as eluent to separate ten sub-fractions according to the polarity. The seventh sub-fraction (n-hexane/ethyl acetate 3:7) was rechromatographed by SiO2 column (230–400 mesh) and followed by preparative TLC to afford colorless oil, alisol M 23-acetate (48 mg). The ninth sub-fraction (n-hexane/ethyl acetate 1:9) was chromatographed by SiO2 column (230–400 mesh) and followed by preparative TLC to afford colorless oil, alisol A 23-acetate (18 mg). In NMR, the 1H NMR (300 MHz, CDCl3) spectra
24. 25. 26. 27.
data of alisol M 23-acetate and alisol A 23-acetate is in agreement with the reported literature value that is listed in the respective bracket. alisol M 23-acetate: 2.22(H1, 2.24), 2.31/2.62(H2, 2.38/2.65), 2.03(H5, 2.05), 1.99(H9, 2.08), 3.83(H11, 3.86), 4.48(H12, 4.53), 1.75/2.41(H15, 1.8/2.47), 1.42(H18, 1.48), 1.0(H19, 1.1), 1.12(H21, 1.16), 4.53(H23, 4.57), 1.16(H26, 1.17), 1.3(H27, 1.31), 1.05(H28, 1.09), 1.03(H29, 1.08), 0.91(H30, 0.88), 2.14(23-OAc, 2.16); alisol A 23-acetate: 2.2(H1, 2.22), 2.3/2.61(H2, 2.33/2.68), 2.04(H5, 2.06), 1.24/ 1.42(H6, 1.28/1.45), 1.19(H7, 1.23), 1.71(H9, 1.72), 3.82(H11, 3.83), 1.91/ 2.58(H12, 1.95/2.6), 1.32/1.83(H15, 1.33/1.87), 2.17(H16, 2.18), 1.13(H18, 1.14), 1.04(H19, 1.05), 2.58(H20, 2.59), 1.01(H21, 0.98), 1.66/1.75(H22, 1.67/ 1.77), 4.88(H23, 4.91), 3.23(H24, 3.23), 1.16(H26, 1.19), 1.2(H27, 1.2), 1.04(H28, 1.05), 1.03(H29, 1.06), 0.95(H30, 0.97), 2.11(23-OAc, 2.05); the 13C NMR (75 MHz, CDCl3) spectra data of alisol M 23-acetate and alisol A 23acetate is in agreement with the reported literature value that is listed in the respective bracket. alisol M 23-acetate: 30.8(C1, 30.8), 34.8(C2, 33.7), 219.5(C3, 219.6), 48.4(C4, 46.9), 48.8(C5, 48.8), 19.6(C6, 20), 35(C7, 34.7), 40.03(C8, 40.2), 45.6(C9, 44.4), 36.8(C10, 36.7), 69.7(C11, 70.7), 64.86(C12, 66.4), 177(C13, 175.8), 49.7(C14, 49.2), 46.9(C15, 46.7), 208(C16, 208.6), 138.4(C17, 140.1), 24.6(C18, 25.2), 25.4(C19, 25.5), 27(C20, 27.4), 19.9(C21, 20.2), 35.5(C22, 35.9), 71.9(C23, 73.8), 60.4(C24, 64.3), 58.6(C25, 59), 19.2(C26, 19.1), 24.59(C27, 24.6), 29.4(C28, 29.6), 20(C29, 20.1), 23(C30, 22.9), 170.3(Ac1, 173.1), 21(Ac2, 21.4); alisol A 23-acetate: 30.9(C1, 30.9), 34.12(C2, 34.17), 220.2(C3, 220.3), 46.9(C4, 46.9), 48.4(C5, 48.4), 20(C6, 20), 33.7(C7, 33.6), 40.7(C8, 40.7), 49.9(C9, 49.7), 36.9(C10, 36.9), 70.2(C11, 70.1), 34.5(C12, 34.5), 137.8(C13, 137.3), 57(C14, 56.9), 30.6(C15, 30.6), 29.5(C16, 29.3), 134.8(C17, 135), 23.1(C18, 22.92), 25.6(C19, 25.6), 28.4(C20, 28.7), 20.1(C21, 20.09), 36.6(C22, 37.9), 71.9(C23, 71.7), 78.4(C24, 78.2), 74.2(C25, 72.5), 25.6(C26, 25.4), 25.5(C27, 26.55), 29.6(C28, 29.6), 20.1(C29, 20), 23.7(C30, 23.8),170.9(Ac1, 170.9), 21.4(Ac2, 21.4); The EI-MS data: alisol M 23-acetate: EI-MS m/z = 544 [M+]; alisol A 23-acetate: EI-MS m/z = 532 [M+]. Makabel, B.; Zhao, Y.; Wang, B.; Bai, Y.; Zhang, Q.; Wu, L.; Yang, L. V. Chem. Pharm. Bull. 2008, 56, 41. Yoshikawa, M.; Tomohiro, N.; Murakami, T.; Ikebata, A.; Matsuda, H.; Matsuda, H.; Kubo, M. Chem. Pharm. Bull. 1999, 47, 524. Lee, S.; Kho, Y.; Min, B.; Kim, J.; Na, M.; Kang, S.; Maeng, H.; Bae, K. Arch. Pharm. Res. 2001, 24, 524. Cell culture and transient transfection reporter assays: The HepG2 cells routinely were cultured as monolayer in Dulbecco’s modified minimal essential medium (DMEM) supplemented with 10% fetal bovine serum, Penicillin (100 unit/ml)/Streptomycin (100 lg/ml), and incubated at 37 °C in a humidified atmosphere of 5% CO2/air. For the transient transfection reporter assays, the HepG2 cells were plated in triplicate in 48-well plates at a density of 1 105 cells/well in the phenol red-free DMEM supplemented with 10% charcoal-stripped fetal bovine serum, Penicillin (100 unit/ml)/ Streptomycin (100 lg/ml). Tweenty four hours later, the cells were transfected with three plasmids by using the Superfect transfection kit. For the detection of agonistic activity against AR, cells were transfected with 2 lg wild type AR expression plasmid (pCMV-AR), 6 lg respective luciferase reporter plasmid containing AR response element, pMMTV-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b, Clontech, Palo Alto, CA). For the detection of agonistic/antagonistic activity against GR, cells were transfected with 2 lg wild type GR expression plasmid (pCMV-GR), 6 lg respective luciferase reporter plasmid containing GR response element, pGRE-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b). For the detection of agonistic/antagonistic activity against PR, cells were transfected with 2 lg wild type PR expression plasmid (pCMV-PR), 6 lg respective luciferase reporter plasmid containing PR response element, pMMTV-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b). After transfetced, the cells were treated with various concentrations of test molecules and one positive control (vehicle, DMSO) in phenol red-free culture medium. To determine the GR antagonistic activity, the test compound was cotreated with 5 nM dexamethasone, while to determine the PR antagonistic activity, the test compound was cotreated with 10 nM progesterone. After incubated for further 24 h, the cells were washed with PBS and lysed with lysis buffer. The lysate was used to determine the transactivation activity of AR, GR, or PR respectively and the b-galactosidase activity for the normalization of transfection efficiency. For the luciferase activity assay, 20 ll of lysate and 100 ll luciferase assay buffer were added into a well of 96-well plate. The luminescence was detected by using a luminescence microplate reader. For the b-galactosidase activity assay, 20 ll of lysate and 180 ll b-galactosidase assay buffer were added into a well of 96well plate. The b-galactosidase activity was measured as luminescence strength. The normalized luciferase activity was calculated by the equation, normalized luciferase activity = luciferase activity/b-galactosidase activity. For AR agonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using vehicle (DMSO) as the standard that was set to 1. For GR antagonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using the normalized luciferase activity exerted by 5 nM dexamethasone as the standard that was set to 1. For PR antagonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using the normalized luciferase activity exerted by 10 nM progesterone as the
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx standard that was set to 1. Experiments were at least triplicated for each test compound. 28. Cantin, L.; Faucher, F.; Couture, J. F.; de Jésus-Tran, K. P.; Legrand, P.; Ciobanu, L. C.; Fréchette, Y.; Labrecque, R.; Singh, S. M.; Labrie, F.; Breton, R. J. Biol. Chem. 2007, 282, 30910.
7
29. Lusher, S. J.; Raaijmakers, H. C.; Vu-Pham, D.; Kazemier, B.; Bosch, R.; McGuire, R.; Azevedo, R.; Hamersma, H.; Dechering, K.; Oubrie, A.; van Duin, M.; de Vlieg, J. J. Biol. Chem. 2012, 287, 20333. 30. Schoch, G. A.; D’Arcy, B.; Stihle, M.; Burger, D.; Bär, D.; Benz, J.; Thoma, R.; Ruf, A. J. Mol. Biol. 2010, 395, 568.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Triterpenes from Alisma orientalis act as androgen receptor agonists, progesterone receptor antagonists, and glucocorticoid receptor antagonists Hsiang-Ru Lin ⇑ Department of Chemistry, College of Science, National Kaohsiung Normal University, No. 62, Shenjhong Rd., Yanchao District, Kaohsiung City 82446, Taiwan, Republic of China
a r t i c l e
i n f o
Article history: Received 28 January 2014 Revised 27 April 2014 Accepted 8 May 2014 Available online xxxx Keywords: Nuclear receptor Natural product Androgen receptor Progesterone receptor Glucocorticoid receptor Alisol M 23-acetate Alisol A 23-acetate
a b s t r a c t Alisma orientalis, a well-known traditional medicine, exerts numerous pharmacological effects including anti-diabetes, anti-hepatitis, and anti-diuretics but its bioactivity is not fully clear. Androgen receptor (AR), progesterone receptor (PR), and glucocorticoid receptor (GR) are three members of nuclear receptor superfamily that has been widely targeted for developing treatments for essential diseases including prostate cancer and breast cancer. In this study, two triterpenes, alisol M 23-acetate and alisol A 23-acetate from Alisma orientalis were determined whether they may act as androgen receptor (AR), progesterone receptor (PR), or glucocorticoid receptor (GR) modulators. Indeed, in the transient transfection reporter assays, alisol M 23-acetate and alisol A 23-acetate transactivated AR in dose-dependent manner, while they transrepressed the transactivation effects exerted by agonist-activated PR and GR. Through molecular modeling docking studies, they were shown to respectively interact with AR, PR, or GR ligand binding pocket fairly well. All these results indicate that alisol M 23-acetate and alisol A 23-acetate from Alisma orientalis might possess therapeutic effects through their modulation of AR, PR, and GR pathways. Ó 2014 Elsevier Ltd. All rights reserved.
Nuclear receptors (NRs) act as transcription factors to modulate transcription actions of target genes involved in maintenance of cellular phenotypes, metabolism, and cell proliferation through homodimerizing or heterodimerizing with other nuclear receptors. Currently, there are more than 30 members in the nuclear receptor superfamily.1 Most of these receptors exert their functions by ligand-activation and they usually contain six functional domains (A–F) including the first transcription activation domain (AF-1, A/ B), DNA binding domain (DBD, C), hinge domain (D), and the second transcription activation domain (AF-2, E/F) also referred to as the ligand binding domain (LBD). Because of their importance in many diseases including breast cancer, prostate cancer, and diabetes diseases, NRs are viewed as potential targets to develop specific therapeutic agents. Androgens including testosterone and dihydrotestosterone (DHT) are steroid hormones that are responsible for the cellular proliferation and differentiation of male sexual organs and secondary sexual characteristics, and they mainly exerted their actions through its respective nuclear receptor, androgen receptor (AR).2 Androgen primarily exerts its action via a genomic mechanism in which androgen passively enters the target cells and binds to AR ⇑ Tel.: +886 7 7172930x7123; fax: +886 7 6051083. E-mail address: [email protected]
in the cytoplasm.3 The androgen/AR complex further translocates into the nucleus, in which the AR complex dimerizes and binds to the promoter region of the androgen-regulated gene to initiate the transcription action and enhance the expression of androgenregulated protein such as PSA, Bcl-2, and maspin.4–6 Generally, androgen plays an essential role in the regulation of men’s health and androgen deficiency results in numerous significant diseases including diabetes and osteoporosis when men enter andropause period.7 Androgen also plays a moderate role in the regulation of physiological functions in women. For example, endometriosis, a benign gynecological disorder, frequently occurs in women and synthetic androgen like danazol (1 in Fig. 1) is one of the endometriosis treatments.8,9 All the evidences mentioned above indicated that the herbal AR agonists should be suitable treatments or supplements for androgen deficiency related diseases and endometriosis. Progesterone plays a critical role in the development, differentiation and function of female reproductive tissues, while it exerted biological effects mainly through agonizing its associated receptor, progesterone receptor (PR), a member of NR family. PR basically behaves as an agonist-activated transcription factor to regulate the expression of essential genes like Bcl-Xl and transforming growth factor (TGF).10,11 However, it is evident that progesterone increases breast cancer risk and PR is involved in the progression of breast cancer.12 Apart from breast cancer, leiomyoma is a benign
http://dx.doi.org/10.1016/j.bmcl.2014.05.039 0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
2
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
O OH
OH HO 11
O
12
23
H H
H
3
N
A
O
16
O
24 25
H
O O
2
1
O
N O
OH
OH
HO
H OH H
H
O
O
4
3 F
F
HO N O
O
O
OH
H H O O
5
6
Figure 1. Chemical structures of danazol (1), alisol M 23-acetate (2), alisol A-23 acetate (3), RU486 (mifepristone, 4), EM5744 (5), and asoprisnil (6).
smooth muscle neoplasm that frequently occurs in uterus. Uterine leiomyoma is worldly prevalent for women and progesterone plays a key role to maintain the growth of uterine leiomyoma.13 Progesterone receptor antagonists have been demonstrated to decrease uterine leiomyoma volume and reduce the bleeding days of tested patients.14 Since progesterone is highly involved in the proliferation of breast cancer and uterine leiomyoma, the herbal antiprogestins would be the effective treatments or supplements for uterine leiomyoma and breast cancer in women. Glucocorticoid receptor (GR) is another member of NR superfamily and the ligand-activated GR homodimerizes to act as an active transcription factor for facilitating the transcription action of regulated genes.15 Glucocorticoid, the natural ligand of GR, has extensive effects on different tissues and the synthetic glucocorticoids have long been used in the treatments of asthma, arthritis, and autoimmune diseases due to their outstanding anti-inflammatory and immunosuppressive effects.16 Although synthetic glucocorticoids have significant anti-inflammatory activity, chronic glucocorticoid therapy has deleterious side effects, including weight gain, osteoporosis, and diabetes mellitus.17 In liver, glucocorticoid receptor antagonists were shown to antagonize insulin’s
action through regulating genes involved in the gluconeogenesis pathway like phosphoenolpyruvate carboxykinase (PEPCK) so the herbal glucocorticoid receptor antagonists might ameliorate insulin resistance and help type II diabetes patients.18 Alisma orientalis (SAM.) JUZEP is widely cultivated in Taiwan, China as well as Japan, and its dried rhizomes are employed as the treatments of diabetes and diuretics. Alisma orientalis is also an important component of a herbal combinatorial drug which is widely utilized to treat hyperglycemia, hyperlipidemia, neuroprotection, and nephritis in Chinese society. The chemical constitutions of Alisma orientalis mainly include sesquiterpenes and protostane type triterpenes like alisol derivatives.19,20 Among them, alisol A and its derivatives have been reported to inhibit Hepatitis B virus (HBV) activity but this activity is present by the secretion of HBV e antigen (HBVeAg).21 Additionally, Alisma orientalis was shown to contain FXR agonists.22 However, Alisma orientalis might exert the pharmacological effects through other molecular mechanisms. In this study, the isolated triterpenes from Alisma orientalis were evaluated for their activity against AR, PR, and GR. The study goal is to determine whether Alisma orientalis might exert therapeutic effect through AR, PR, and GR pathways.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
3
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
6
Relative Luciferase Activity
5
4
3
2
100nΜ AA23a
1μΜ AA23a
10μΜ AA23a
100nΜ AM23a
1μΜ AM23a
10μΜ AM23a
10μΜ Danazol
DMSO
1
Figure 2. Transient transfection reporter assay in HepG2 cells. AR agonistic effect was determined by the MMTV promoter-driven luciferase activity. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of DMSO. The RLA data represented mean ± standard deviation for at least three individual determinations. AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
1.0
Relative Luciferase Activity
0.8
0.6
0.4
0.2
100nΜ AA23a + 10nM Prog
1μΜ AA23a + 10nM Prog
10μΜ AA23a + 10nM Prog
100nΜ AM23a + 10nM Prog
1μΜ AM23a + 10nM Prog
10μΜ AM23a + 10nM Prog
1μΜ RU486 + 10nM Prog
0.0
10nM Prog
The isolation of the protostane-type triterpenes employed the ethyl acetate (EtOAc) soluble fraction of Alisma orientalis that was processed by flash chromatography methods to generate three major constituents, alisol M 23-acetate (2 in Fig. 1), alisol A 23-acetate (3 in Fig. 1), and alisol B.23 Their structures were identified by comparing NMR spectral data with literature values.24–26 Among these isolated triterpenes, alisol B was not evaluated for its activity due to its cytotoxic effect on HepG2 cells, while alisol M 23-acetate and alisol A 23-acetate showed no influence for the growth of HepG2 cells. Primarily, to evaluate the agonistic activity of alisol M 23-acetate and alisol A 23-acetate against AR, the transient transfection reporter assay with a reporter plasmid containing androgen receptor response element (pMMTV-luc) was employed.27 For comparison, the transactivation activity of vehicle (DMSO, RLA = 1) was used as the standard reference and 10 lM danazol (RLA = 3.8) was utilized as the positive control. As shown in Figure 2, alisol M 23-acetate exerted androgenic activity about 4-fold of vehicle activity at 10 lM, while alisol A 23-acetate possessed less androgenic effect. Both compounds exerted weak androgenic effect at 1 lM and completely lost their AR agonism at 100 nM. This result strongly supported that alisol M 23-acetate and alisol A 23-acetate could act as AR agonists as danazol. To determine the agonistic and antagonistic activity of alisol M 23-acetate and alisol A 23-acetate against PR, the transient transfection reporter assay with a reporter plasmid containing progesterone receptor response element (pMMTV-luc) was utilized.27 For comparison, 10 nM progesterone was cotreated with test compounds and its activity was set to 1. RU486 (6% of 10 nM progesterone agonistic effect, 4 in Fig. 1) at 1 lM, was employed as the positive control. Alisol M 23-acetate and alisol A 23-acetate did not exert significant agonistic effect against PR at 20 lM (data not shown). As shown in Figure 3, alisol M 23-acetate and alisol A 23-acetate antagonized PR to 64% and 66% of 10 nM progesterone agonistic effect respectively at 10 lM. Both compounds exerted antagonistic activity of 71% and 82% of 10 nM progesterone agonistic effect at 1 lM. However, alisol M 23-acetate lost its antagonistic activity but alisol A 23-acetate exhibited weak PR
Figure 3. Transient transfection reporter assay in HepG2 cells. PR antagonistic effect was determined by the MMTV promoter-driven luciferase activity in the presence of 10 nM progesterone. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of 10 nM progesterone. The RLA data represented mean ± standard deviation for at least three individual determinations. Prog: progesterone; AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
antagonism at 100 nM. This result showed that both alisol derivatives could serve as PR antagonists. To determine the agonistic and antagonistic activity of alisol M 23-acetate and alisol A 23-acetate against GR, the transient transfection reporter assay with a reporter plasmid containing glucocorticoid receptor response element (pGRE-luc) was utilized.27 For comparison, 5 nM dexamethasone was cotreated with test compounds and its activity was set to 1. RU486 (30% of 5 nM dexamethasone agonistic activity) at 1 lM, was employed as the positive control. Alisol M 23-acetate and alisol A 23-acetate did not exhibit obvious agonistic effect against GR at 20 lM (data not shown). Figure 4 outlined that alisol M 23-acetate and alisol A 23-acetate antagonized GR to 30% and 39% of 5 nM dexamethasone agonistic activity respectively at 10 lM. The antagonistic activity of both compounds dropped to 56% of 5 nM dexamethasone effect at 1 lM. However, alisol M 23-acetate lost its antagonistic activity but alisol A 23-acetate still exerted weak inhibition against GR at 100 nM. This result indicated that both alisol derivatives could act as GR antagonists. Finally, the computer-based modeling program (Discovery Studio 2.1, Accelrys, San Diego, USA) was employed to generate docking models that were further used to discuss how alisol M 23-acetate and alisol A 23-acetate might interact with AR LBD, PR LBD, and GR LBD respectively. The crystal structure of most NR LBD complexed with its modulator contains 12 a-helices to form a ligand binding pocket in which helices 4, 5, 8, and 9 are sandwiched between helices 1, 3, 7, and 10 at the top half of the pocket. Importantly, helices 3, 7, and 11 form a large cavity in the bottom half of the pocket to allow the NR modulator to locate. In the agonistic conformation, once a NR agonist bound to the pocket and located in the cavity, it subsequently induced the movement of helice 12 to the entrance of pocket and exposed a coactivator-binding surface. Traditionally, most AR agonists contain a steroidal skeleton and the addition of a large scaffold on its C12, C13 or C17 will generate the AR antagonists. However, in
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
4
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
1.8 1.6
Relative Luciferase Activity
1.4 1.2 1.0 0.8 0.6 0.4 0.2
100nΜ AA23a + 5nM Dex
1μΜ AA23a + 5nM Dex
10μΜ AA23a + 5nM Dex
100nΜ AM23a + 5nM Dex
1μΜ AM23a + 5nM Dex
10μΜ AM23a + 5nM Dex
1μΜ RU486 + 5nM Dex
5nM Dex
0.0
Figure 4. Transient transfection reporter assay in HepG2 cells. GR antagonistic effect was determined by the GRE promoter-driven luciferase activity in the presence of 5 nM dexamethasone. The relative luciferase activity (RLA) was quantified as normalized luciferase activity of test compound/normalized luciferase activity of 5 nM dexamethasone. The RLA data represented mean ± standard deviation for at least three individual determinations. Dex: dexamethasone; AM23a: alisol M 23-acetate; AA23a: alisol A-23 acetate.
these years, EM5744 (5 in Fig. 1), an AR agonist containing a large substituent on the C13 of its steroidal core scaffold, revealed a new binding mode of AR agonist in AR LBD.28 Generally, the large side chain on the C12, C13 or C17 of the AR antagonist’s steroidal scaffold leans toward the entrance of AR ligand binding pocket and forces helice 12 to locate at the inactive position. For example, bicalutamide contains a side chain, 4-fluorophenylsulfonyl group, to protrude out of AR ligand binding pocket and prevent helice 12 to locate at the agonistic position. Interestingly, in the EM5744/AR LBD crystal structure, the side chain of EM5744 located in the channel formed by helices 3, 11, and 12. The 3-keto and 17-hydroxyl groups of EM5744 interacted with AR LBD by hydrogen bond interaction as DHT. Importantly, the terminal fluoro substituent of EM5744’s side chain made contact with His 874 of helice 12 and this interaction was not shown in the DHT/ AR LBD structure. In fact, helice 12 in the EM5744/AR LBD structure exhibited some flexibility and moved a little bit away from the AR ligand binding pocket so the AR LBD had the flexibility to accommodate large ligands by changing the position of the side chain of helice 12 and other helices. To determine the binding modes of alisol M 23-acetate and alisol A 23-acetate in AR LBD, the ligandfit docking program and the crystal structure of human AR LBD complexed with EM5744 (Protein Data Bank ID: 2PNU) were utilized. Due to similar binding situations for alisol M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. As shown in Figure 5A and B, the docking model of alisol M 23-acetate indicated that alisol M 23-acetate resided in AR ligand binding pocket in a similar manner as EM5744 by leaning its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5A and B) and locating its 23-acetyl group nearby helice 11(as shown in yellow color in Fig. 5A and B).
Importantly, the side chain of alisol M 23-acetate inserted into the same channel as the side chain of EM5744 fairly well. Alisol M 23-acetate mainly contacted with AR LBD by hydrophobic interaction and its 3-keto moiety did not form any hydrogen bond interaction with AR LBD. Alternatively, the 16-keto group of alisol M 23-acetate served as a hydrogen bond acceptor to make contact with the side chain of Thr 877. This hydrophilic contact stabilized the insertion of the large side chain of alisol M 23-acetate and the overall agonistic alisol M 23-acetate/AR LBD conformation. Furthermore, the lack of hydrogen bond interaction formed by 3-keto moiety also interpreted the similar AR agonism exerted by alisol M 23-acetate and danazol. Moreover, alisol A 23-acetate had similar binding mode as alisol M 23-acetate but it exerted slightly less AR agonism than alisol M 23-acetate. The primary structural difference between both alisol derivatives is the terminal portion of their C17 side chain which was surrounded by hydrophobic residues of AR LBD including Val 716, Leu873, Ile 898, Val 901, and Val 903. Since there is no hydrogen bond interaction between their side chains and AR LBD, the unfavorable insertion of hydrophilic substituents of alisol A 23-acetate into a hydrophobic zone might account for its less AR agonism. In conclusion, this docking result highly supported that both alisol derivatives could serve as AR agonists. Most NR antagonists have different structural requirement due to their distinct interaction with their respective receptor ligand binding pocket. Once a NR antagonist bound to the pocket, it subsequently induced the movement of helice 12 to pack antiparallel to helice 11 and further prevented the formation of the coactivator-binding surface or facilitated corepressor-binding. Some novel PR antagonists like asoprisnil (6 in Fig. 1), contain a steroidal-like skeleton as the core scaffold with 3-keto group and an antiprogestin side chain located at 11b position of the steroidal scaffold.29 Asoprisnil located in PR LBD, while its 3-keto moiety oriented toward helice 3 and its C11 side chain located in the channel formed by helices 3 and 11. Primarily asoprisnil contacted with PR LBD by hydrophobic interaction and made two hydrogen bond interactions with PR LBD including (1) its 3-keto moiety acted as the hydrogen bond acceptor to form the hydrogen bond interaction with Gln 725 and Arg 766; (2) its terminal oxime group on C11 side chain acted as hydrogen bond acceptors to make contact with two water molecules that also interacted with Glu 723 and Asn 719 to form hydrogen bond networks. Essentially, asoprisnil as other PR antagonists antagonized PR by its antagonistic side chain to protrude out of PR ligand binding pocket and prevent helice 12 from locating at the active position. To determine the binding modes of alisol M 23-acetate and alisol A 23-acetate in PR LBD, the ligandfit docking program and the crystal structure of human PR LBD/ asoprisnil (Protein Data Bank ID: 2OVH) were utilized. Due to similar binding situations for alisol M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. Figure 5C and D outlined the binding mode of alisol M 23-acetate in PR LBD, in which alisol M 23-acetate leaned its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5C and D) and located its 23-acetyl group nearby helice 11 (as shown in yellow color in Fig. 5C and D). Importantly, unlike asoprisnil, alisol M 23-acetate only made one hydrogen bond interaction by its 3-keto group serving as a hydrogen bond acceptor to contact with the terminal amino group of Gln 725. This binding difference might be resulted from alisol M 23-acetate’s bulky C4 dimethyl substituents that were closed to helice 5 and prevented the interaction between alisol M 23-acetate’s 3-keto group and Arg 766. Additionally, although part of helice 12 of hPR LBD/asoprisnil structure was missing, it cannot be excluded that the side chain of alisol M 23-acetate might contact with helice 12 of PR LBD. Especially, the structure–activity relationship study for 19-nortestosterone and RU486 derivatives mentioned that the addition of five- or
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
5
Figure 5. A: The binding model of alisol M-23 acetate in AR ligand binding pocket. B: The hydrophilic interactive mode of alisol M-23 acetate in AR ligand binding pocket. C: The binding model of alisol M-23 acetate in PR ligand binding pocket. D: The hydrophilic interactive mode of alisol M-23 acetate in PR ligand binding pocket. E: The binding model of alisol M-23 acetate in GR ligand binding pocket. F: The hydrophilic interactive mode of alisol M-23 acetate in GR ligand binding pocket. Oxygen atom in red color and hydrogen atom in white color. Green line indicates the hydrogen bond interaction.
six-membered substituent on the C17 of steroidal progestins and antiprogestins could enhance the binding affinity and progestin/ antiprogestin activity, so the complete binding mode of alisol M 23-acetate in PR LBD might still need the generation of hPR LBD/ alisol M 23-acetate crystal structure. RU486, a novel GR antagonist, contains a steroidal-like core scaffold with a 3-keto group and a side chain located at 11b position of the steroidal scaffold. RU 486 resided in GR LBD, while its 3-keto moiety oriented toward helice 3 and its C11 side chain located in the channel formed by helices 3 and 11.30 Primarily
RU486 contacted with GR LBD by hydrophobic interaction and made the hydrogen bond interactions with GR LBD by its 3-keto group acting as the hydrogen bond acceptor to interact with Gln 570 and Arg 611. Moreover, the C11 side chain of RU486 leaned closed to helice 3 and protruded out of GR ligand pocket but did not make any significant hydrophilic contact with helice 12 of GR LBD. To further determine the binding modes of alisol M 23-acetate and alisol A 23-acetate, the crystal structure of human GR LBD/RU486 (Protein Data Bank ID: 3H52) was utilized to perform the docking study. Due to similar binding situations for alisol
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
6
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
M 23-acetate and alisol A 23-acetate, only the binding mode of alisol M 23-acetate was shown here. As shown in Figure 5E and F, the binding mode of alisol M 23-acetate in GR LBD outlined that alisol M 23-acetate located its 3-keto moiety toward helice 3 (as shown in blue color in Fig. 5E and F) and its 23-acetyl group nearby helice 11 (as shown in yellow color in Fig. 5E and F). Importantly, alisol M 23-acetate mainly contacted with GR LBD by hydrophobic interaction and made two hydrogen bond interactions including (1) its 3keto group served as a hydrogen bond acceptor to contact with the terminal amino group of Gln 570; (2) its 11-hydroxyl group acted as a hydrogen bond donor to contact with the backbone of Leu 563. Furthermore, the 23-acetyl side chain of alisol M 23-acetate located in the center of the channel formed by helices 3 and 11 to cause helice 12 of GR LBD shifting to the antagonistic position. The results from the docking studies showed that alisol M 23-acetate and alisol A 23-acetate could act as GR antagonists. AR, PR, and GR are three members of NRs to be viewed as the target proteins for developing multi-disease treatments by pharmaceutical industry. In this study, it is the first report to demonstrate that alisol M 23-acetate and alisol A 23-acetate served as AR agonists, PR antagonists, and GR antagonists so Alisma orientalis might exert therapeutic effects through AR, PR, and GR pathways. Acknowledgments I thank Dr. Tsai S.Y. at Baylor College of Medicine for kindly offering hPR expression vector and Dr. Gelmann E.P. for generously offering pMMTV-luc vector. References and notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Olefsky, J. M. J. Biol. Chem. 2001, 276, 36863. Culig, Z.; Klocker, H.; Bartsch, G.; Hobisch, A. Endocr. Relat. Cancer 2002, 9, 155. Gao, W.; Bohl, C. E.; Dalton, J. T. Chem. Rev. 2005, 105, 3352. Luke, M. C.; Coffey, D. S. J. Androl. 1994, 15, 41. Huang, H.; Zegarra-Moro, O. L.; Benson, D.; Tindall, D. J. Oncogene 2004, 23, 2161. Zhang, M.; Magit, D.; Sager, R. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 5673. Theodoraki, A.; Bouloux, P. M. Menopause Int. 2009, 15, 87. Fedele, L.; Berlanda, N. Expert Opin. Emerg. Drugs 2004, 9, 167. Vercellini, P.; Somigliana, E.; Viganò, P.; Abbiati, A.; Barbara, G.; Crosignani, P. G. Drugs 2009, 69, 649. Moore, M. R.; Conover, J. L.; Franks, K. M. Biochem. Biophys. Res. Commun. 2000, 277, 650. Gong, Y.; Anzai, Y.; Murphy, L. C.; Ballejo, G.; Holinka, C. F.; Gurpide, E.; Murphy, L. J. Cancer Res. 1991, 51, 5476. Moore, M. R. Curr. Cancer Drug Targets 2004, 4, 183. Kim, J. J.; Kurita, T.; Bulun, S. E. Endocr. Rev. 2013, 34, 130. Engman, M.; Granberg, S.; Williams, A. R.; Meng, C. X.; Lalitkumar, P. G.; Gemzell-Danielsson, K. Hum. Reprod. 2009, 24, 1870. Oakley, R. H.; Cidlowski, J. A. J. Allergy Clin. Immunol. 2013, 132, 1033. Quax, R. A.; Manenschijn, L.; Koper, J. W.; Hazes, J. M.; Lamberts, S. W.; van Rossum, E. F.; Feelders, R. A. Nat. Rev. Endocrinol. 2013, 9, 670. Buehring, B.; Viswanathan, R.; Binkley, N.; Busse, W. J. Allergy Clin. Immunol. 2013, 132, 1019. Taylor, A. I.; Frizzell, N.; McKillop, A. M.; Flatt, P. R.; Gault, V. A. Horm. Metab. Res. 2009, 41, 899. Peng, G. P.; Tian, G.; Huang, X. F.; Lou, F. C. Phytochemistry 2003, 63, 877. Hu, X. Y.; Guo, Y. Q.; Gao, W. Y.; Zhang, T. J.; Chen, H. X. J. Asian Nat. Prod. Res. 2008, 10, 481. Jiang, Z. Y.; Zhang, X. M.; Zhang, F. X.; Liu, N.; Zhao, F.; Zhou, J.; Chen, J. J. Planta Med. 2006, 72, 951. Lin, H. R. Bioorg. Med. Chem. Lett. 2012, 22, 4787. The dried and powdered Alismatis rhizoma (3 kg) was extracted with n-hexane, ethyl acetate and ethanol separately at room temperature. The extracts were concentrated in vacuo to yield crude oil. The combined ethyl acetate fraction (10 g) was chromatographed over SiO2 column (70–230 mesh) using gradient n-hexane-ethyl acetate as eluent to separate ten sub-fractions according to the polarity. The seventh sub-fraction (n-hexane/ethyl acetate 3:7) was rechromatographed by SiO2 column (230–400 mesh) and followed by preparative TLC to afford colorless oil, alisol M 23-acetate (48 mg). The ninth sub-fraction (n-hexane/ethyl acetate 1:9) was chromatographed by SiO2 column (230–400 mesh) and followed by preparative TLC to afford colorless oil, alisol A 23-acetate (18 mg). In NMR, the 1H NMR (300 MHz, CDCl3) spectra
24. 25. 26. 27.
data of alisol M 23-acetate and alisol A 23-acetate is in agreement with the reported literature value that is listed in the respective bracket. alisol M 23-acetate: 2.22(H1, 2.24), 2.31/2.62(H2, 2.38/2.65), 2.03(H5, 2.05), 1.99(H9, 2.08), 3.83(H11, 3.86), 4.48(H12, 4.53), 1.75/2.41(H15, 1.8/2.47), 1.42(H18, 1.48), 1.0(H19, 1.1), 1.12(H21, 1.16), 4.53(H23, 4.57), 1.16(H26, 1.17), 1.3(H27, 1.31), 1.05(H28, 1.09), 1.03(H29, 1.08), 0.91(H30, 0.88), 2.14(23-OAc, 2.16); alisol A 23-acetate: 2.2(H1, 2.22), 2.3/2.61(H2, 2.33/2.68), 2.04(H5, 2.06), 1.24/ 1.42(H6, 1.28/1.45), 1.19(H7, 1.23), 1.71(H9, 1.72), 3.82(H11, 3.83), 1.91/ 2.58(H12, 1.95/2.6), 1.32/1.83(H15, 1.33/1.87), 2.17(H16, 2.18), 1.13(H18, 1.14), 1.04(H19, 1.05), 2.58(H20, 2.59), 1.01(H21, 0.98), 1.66/1.75(H22, 1.67/ 1.77), 4.88(H23, 4.91), 3.23(H24, 3.23), 1.16(H26, 1.19), 1.2(H27, 1.2), 1.04(H28, 1.05), 1.03(H29, 1.06), 0.95(H30, 0.97), 2.11(23-OAc, 2.05); the 13C NMR (75 MHz, CDCl3) spectra data of alisol M 23-acetate and alisol A 23acetate is in agreement with the reported literature value that is listed in the respective bracket. alisol M 23-acetate: 30.8(C1, 30.8), 34.8(C2, 33.7), 219.5(C3, 219.6), 48.4(C4, 46.9), 48.8(C5, 48.8), 19.6(C6, 20), 35(C7, 34.7), 40.03(C8, 40.2), 45.6(C9, 44.4), 36.8(C10, 36.7), 69.7(C11, 70.7), 64.86(C12, 66.4), 177(C13, 175.8), 49.7(C14, 49.2), 46.9(C15, 46.7), 208(C16, 208.6), 138.4(C17, 140.1), 24.6(C18, 25.2), 25.4(C19, 25.5), 27(C20, 27.4), 19.9(C21, 20.2), 35.5(C22, 35.9), 71.9(C23, 73.8), 60.4(C24, 64.3), 58.6(C25, 59), 19.2(C26, 19.1), 24.59(C27, 24.6), 29.4(C28, 29.6), 20(C29, 20.1), 23(C30, 22.9), 170.3(Ac1, 173.1), 21(Ac2, 21.4); alisol A 23-acetate: 30.9(C1, 30.9), 34.12(C2, 34.17), 220.2(C3, 220.3), 46.9(C4, 46.9), 48.4(C5, 48.4), 20(C6, 20), 33.7(C7, 33.6), 40.7(C8, 40.7), 49.9(C9, 49.7), 36.9(C10, 36.9), 70.2(C11, 70.1), 34.5(C12, 34.5), 137.8(C13, 137.3), 57(C14, 56.9), 30.6(C15, 30.6), 29.5(C16, 29.3), 134.8(C17, 135), 23.1(C18, 22.92), 25.6(C19, 25.6), 28.4(C20, 28.7), 20.1(C21, 20.09), 36.6(C22, 37.9), 71.9(C23, 71.7), 78.4(C24, 78.2), 74.2(C25, 72.5), 25.6(C26, 25.4), 25.5(C27, 26.55), 29.6(C28, 29.6), 20.1(C29, 20), 23.7(C30, 23.8),170.9(Ac1, 170.9), 21.4(Ac2, 21.4); The EI-MS data: alisol M 23-acetate: EI-MS m/z = 544 [M+]; alisol A 23-acetate: EI-MS m/z = 532 [M+]. Makabel, B.; Zhao, Y.; Wang, B.; Bai, Y.; Zhang, Q.; Wu, L.; Yang, L. V. Chem. Pharm. Bull. 2008, 56, 41. Yoshikawa, M.; Tomohiro, N.; Murakami, T.; Ikebata, A.; Matsuda, H.; Matsuda, H.; Kubo, M. Chem. Pharm. Bull. 1999, 47, 524. Lee, S.; Kho, Y.; Min, B.; Kim, J.; Na, M.; Kang, S.; Maeng, H.; Bae, K. Arch. Pharm. Res. 2001, 24, 524. Cell culture and transient transfection reporter assays: The HepG2 cells routinely were cultured as monolayer in Dulbecco’s modified minimal essential medium (DMEM) supplemented with 10% fetal bovine serum, Penicillin (100 unit/ml)/Streptomycin (100 lg/ml), and incubated at 37 °C in a humidified atmosphere of 5% CO2/air. For the transient transfection reporter assays, the HepG2 cells were plated in triplicate in 48-well plates at a density of 1 105 cells/well in the phenol red-free DMEM supplemented with 10% charcoal-stripped fetal bovine serum, Penicillin (100 unit/ml)/ Streptomycin (100 lg/ml). Tweenty four hours later, the cells were transfected with three plasmids by using the Superfect transfection kit. For the detection of agonistic activity against AR, cells were transfected with 2 lg wild type AR expression plasmid (pCMV-AR), 6 lg respective luciferase reporter plasmid containing AR response element, pMMTV-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b, Clontech, Palo Alto, CA). For the detection of agonistic/antagonistic activity against GR, cells were transfected with 2 lg wild type GR expression plasmid (pCMV-GR), 6 lg respective luciferase reporter plasmid containing GR response element, pGRE-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b). For the detection of agonistic/antagonistic activity against PR, cells were transfected with 2 lg wild type PR expression plasmid (pCMV-PR), 6 lg respective luciferase reporter plasmid containing PR response element, pMMTV-luc, and 1 lg normalization control, b-galactosidase reporter plasmid (pCMV-b). After transfetced, the cells were treated with various concentrations of test molecules and one positive control (vehicle, DMSO) in phenol red-free culture medium. To determine the GR antagonistic activity, the test compound was cotreated with 5 nM dexamethasone, while to determine the PR antagonistic activity, the test compound was cotreated with 10 nM progesterone. After incubated for further 24 h, the cells were washed with PBS and lysed with lysis buffer. The lysate was used to determine the transactivation activity of AR, GR, or PR respectively and the b-galactosidase activity for the normalization of transfection efficiency. For the luciferase activity assay, 20 ll of lysate and 100 ll luciferase assay buffer were added into a well of 96-well plate. The luminescence was detected by using a luminescence microplate reader. For the b-galactosidase activity assay, 20 ll of lysate and 180 ll b-galactosidase assay buffer were added into a well of 96well plate. The b-galactosidase activity was measured as luminescence strength. The normalized luciferase activity was calculated by the equation, normalized luciferase activity = luciferase activity/b-galactosidase activity. For AR agonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using vehicle (DMSO) as the standard that was set to 1. For GR antagonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using the normalized luciferase activity exerted by 5 nM dexamethasone as the standard that was set to 1. For PR antagonistic effects of test molecules, the normalized luciferase activity value was further converted to relative luciferase activity (RLA) by using the normalized luciferase activity exerted by 10 nM progesterone as the
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
H.-R. Lin / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx standard that was set to 1. Experiments were at least triplicated for each test compound. 28. Cantin, L.; Faucher, F.; Couture, J. F.; de Jésus-Tran, K. P.; Legrand, P.; Ciobanu, L. C.; Fréchette, Y.; Labrecque, R.; Singh, S. M.; Labrie, F.; Breton, R. J. Biol. Chem. 2007, 282, 30910.
7
29. Lusher, S. J.; Raaijmakers, H. C.; Vu-Pham, D.; Kazemier, B.; Bosch, R.; McGuire, R.; Azevedo, R.; Hamersma, H.; Dechering, K.; Oubrie, A.; van Duin, M.; de Vlieg, J. J. Biol. Chem. 2012, 287, 20333. 30. Schoch, G. A.; D’Arcy, B.; Stihle, M.; Burger, D.; Bär, D.; Benz, J.; Thoma, R.; Ruf, A. J. Mol. Biol. 2010, 395, 568.
Please cite this article in press as: Lin, H.-R. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.05.039
