| Osteoporosis is a metabolic bone disease characterized by low bone mass and structural deterioration of bone tissue associated with bone fragility and an increased vulnerability to low-trauma fractures. Bone is a living and dynamic tissue that undergoes constant remodeling process controlled by two coupled processes begins with bone resorption by osteoclasts and follows by bone formation by osteoblasts. Under normal circumstances, bone resorption rate and bone formation rate are in a dynamic equilibrium. An imbalance that bone resorption exceeds bone formation in the bone remodeling process results in osteoporosis. Inhibition of osteoclast functions and enhancement of osteoblast activities are key therapeutical tools for osteoporosis. Osteoporosis not only represents a considerable medical and socioeconomic burden for modern societies, but also affects people’s life and health. Osteoporosis drugs available among the current market are mainly divided into two categories:anti-resorptive agents and bone anabolic agents. Anti-resorptive drugs, such as bisphosphonates, selective estrogen receptor modulators, and active vitamin D and its analogues, have been the backbone of osteoporosis management. Unfortunately, PTH1-34, the only clinically available bone anabolic drug for the treatment of osteoporosis, has dosing period limitation and great side effects. Hence, development of anabolic agents for improving bone formation and bone repair has considerable social and economic impact.A growing number of reports indicated that peripheral serotonin biosynthesized by EC cells of the gut (Gut-derived serotonin, GDS) enters the circulation, and acts as a strong inhibitor of osteoblast proliferation and bone formation. Thus, inhibition of GDS biosynthesis may be helpful to improve the bone formation rate. Tryptophan hydroxylase 1 (TPH-1) is the principal enzyme in the biosynthesis of serotonin (converts 1-tryptophan into 5-hydroxy-tryptophan, the precursor to serotonin) and a primary rate-limiting factor in serotonergic functioning. Therefore, TPH-1 can be recognized as a new target of anti-osteoporosis drug discovery and suppression of GDS biosynthesis via TPH-1 inhibitor provides a novel tool to design anabolic agents for osteoporosis treatment.In this paper, we virtually designed a series of ursolic acid (UA) derivatives based on structure-based drug design (SBDD) and predicted the orientation of essential substituents on UA at the TPH-1 binding site. Guided by the docking results, a series of derivatives were synthesized and their activities were assayed in vitro and in vivo.1. Design of UA DerivativesUnder the guidance of computer-aided design, we virtually designed series of UA derivatives according to the binding free energy (ΔE) and the inhibition constant (Ki) obtained by the docking technology. Guided by the docking results, we were able to synthesize series of novel TPH-1 inhibitors based on UA.2. Synthesis of UA DerivativesFirstly, treating UA with corresponding acids by the smaller steric condensing agent EDCI in the presence of DMAP afforded the target products 1a-d according to the docking results. However, reactions of acids with electron-withdrawing groups under the condition gave the undesired products that two UA core were directly coupled (1e-h). Acetylation of the hydroxy of 1 with acetic anhydride in reflux acetic acid in the presence of pyridine afforded 2.For the oxidation of UA (1) to prepare 3-keto UA (2), the reagent H2O2/Na2WO4 with the phosphate buffer solution (NaH2PO4/Na2HPO4) was used and the key intermediate 3 was obtained in an excellent yield and good selectivity without much environmental pollution. Indole was introduced into the ring A of UA by Fischer reaction and three different reaction conditions were explored. Although the indole derivatives 9-10 were obtained by three different methods in the similar yields, when the corresponding hydrazine hydrochloride was used in acetic acid the reaction went smoothly and almost without byproducts. Compound 3 was treated with excess amount of sodium methoxide followed by addition of ethyl formate without protection of carboxyl group (C-28) under atmospheric condition to give 4 in high yield. Treatment of 4 with corresponding amines in EtOH at 80℃ in the presence of Na gave the desired aldehyde amino condensation products 5a-d. Next, we turned our attention to the preparation of pryrimidine derivatives by treating 4 with formamidine acetate or benzamidine hydrochloride in the presence of sodium in EtOH, however, an undesired aldehyde alcohol condensation derivative 6 was obtained in high yield.Heterocyclization of 4 with hydroxylamine hydrochloride in refflux ethanol gave the isoxazole derivative 11. Treatment of 4 with hydrazine di-hydrochloride, phenyl-hydrazine hydrochloride, and 4-fluorophenyl-hydrazine hydrochloride in EtOH under reflux afforded pyrazole derivatives (13-15), respectively. Cleavage of 11 with NaOH in a co-solvent system of CH30H/Et2O followed by treating with DDQ in ethanol afforded 12. Considering introduction of a hydrogen bond donating group might be helpful to increase the activity, the conversion of cyano group to carboxyl group was performed. Compound 12 was treated with NaOH in reflux CH3OH overnight followed by the acidification of hydrochloric acid afforded 26. It is noteworthy that this reaction is completed only when the concentration of NaOH in methanol is higher than 1 N, otherwise, the reaction did not happen.Quinoxalines could be introduced into the ring A of UA by many different methods, but the yields were low and the byproducts were inevitable. Therefore, the ketone 3 was firstly oxidized into di-ketone 16, and then di-ketone 16 was treated with o-phenylenediamines in EtOH afforded the quinoxaline derivatives (17-18). As SeO2 showed a great toxicity in the oxidation reaction process, therefore, other synthetic protocol was tried. A t-BuOK in t-BuOH green oxidation reagent was used and the experimental conditions were optimized to give 16. Then,16 was reacted with corresponding 1,2-diamines in refluxing ethanol gave the target compounds 17 and 18 in excellent yields.The introduction of pyrimidine ring into UA is still a major challenge. Firstly, the enamine group was successfully introduced into C-2 of 3. Then condensation of 7 with formamidine acetate in the presence of sodium methylate afforded 19. C-28 ester was removed by anhydrous LiI in the presence of n-octylamine in DMF at 165℃ gave the target compound 20.The amino acid esters, and heterocyclic moieties were further introduced into C-28 position of UA derivatives as designed. Firstly, C-28 chlorides of UA analogues were prepared by oxalyl chloride in anhydrous CH2Cl2, and then treatment with corresponding amino acid esters and heterocyclic moieties in the presence of triethylamine afforded corresponding amides.Introduction of amide bonds at C-28 of 11 and 12 may further increase their potency according to the docking results. Therefore, a series of typical amines were introduced into C-28 of 11 and 12 following the same procedure as above (11d-n,12d-n).3. Activities in Vitro1. Inhibitory activity on serotonin biosynthesisA HPLC method was initially used to evaluate the inhibitory activity of the synthesized compounds on serotonin biosynthesis in RBL2H3 cells (TPH-1 expressing cells). Together with the cytotoxicity data, a structure activity relationship (SAR) for the test derivatives could be summarized as follows:(1) Esterification of C-3 hydroxyl group with carboxylic acids did not show significant inhibitory effect on serotonin biosynthesis compared with UA even though two UA core were coupled together (la-h);(2) Acetylation or oxidation of C-3 hydroxyl group of UA (2,3) did not give positive impact on serotonin biosynthesis compared with UA. Oxidation or acetylation of C-3 hydroxyl group in addition to modification of C-28 carboxyl group with amino acids had significant improvement on the inhibitory activity (2a-b,3a-c) with low cytotoxicity. While, introducing phenylalanine moiety to 2 lowered the activity (2c);(3) Introducing hydrogen bond donating groups or hydrogen bond acceptors at C-2 position based on ketone 3 resulted in low inhibitory activity (5a-d,6,8,26);(4) Introduction of isoxazole, pyrazole, phenyl pyrazolyl, and pyrazol-fluorophenyl into UA did not show significant enhacement on inhibition of serotonin biosynthesis compared with UA (11, 13-18). However, pyrazole-extended derivatives gained more potent inhibitory activity by insertion of amino acids at C-28 carboxyl (13a-c,15a-c) without significantly cytotoxicity. Surprisingly, the inhibitory activity was not improved in case of cyclization of ketone 4 with other phenylhydrazine or hydroxylamine hydrochloride and introducing amino acids at C-28 (lla-c, 14a-c) compared with (13a-c,15a-c). The same trend was observed in the cleaving the isoxazole ring derivatives (12,12a-c);(5) Indolization of UA without modification with amino acids at C-28 carboxyl did not show significant effect compared with UA (9-10). Indolization of ketone 2 with substitution of the C-28 carboxyl group by amino acids resulted in a reduction of the activity (9a-b, lOa-c). Interestingly, introduction of phenylalanine methyl ester into C-28 carboxyl of 12 significantly increased its activity (9c);(6) Introduction of alkyl amine, benzylamine or aromatic amines sunstituted with halogen, methoxy or hydroxy at C-28 carboxyl of 11 has no significant impact on the inhibitory activity (12d-e,121,12n). Furthermore, conjugates of piperidine or morpholine moieties at 11 provided effective improvement of the activity (11i-k, 11m) which was similar with 12 (12i-k,12m), while, together with a high cytotoxicity upon RBL2H3 cells (lli-k, 11m,12i-k,12m). When the aromatic amines with strong electron-withdrawing group and strong electron-donating groups were introduced to 11 and 12, their inhibitory activity was substantial improved (11g-h,12g-h). However, compound llg displayed potent activity inhibiting serotonin biosynthesis (79.3%) with slight cytotoxic effect at 10μM;(7) Introduction of quinoxaline ring into UA could slightly enhance the activity (17-18). Interestingly, conjugates of glycine ethyl ester and quinoxaline derivatives could provide effective improvement of the activity without significant cytotoxicity (18a) upon RBL2H3 cells. In most cases, the quinoxaline, six membered heterocyclic ring derivatives that introduced amino acids at C-28 carboxyl displayed no effects on inhibitory activity (17a-c,18b-c,21). The same phenomenon was also observed in pyrimidine derivatives (20,20a-c);(8) The quinoxaline derivatives introduced with piperidine moieties displayed an increase activity but with higher cytotoxicity (22-25).Thus, compounds 11g and 18a which showed high activity and no significant cytotoxicity were choosed as TPH-1 inhibitors for further experiments.2. Inhibitions on TPH-1 expressionFurther study indicated that compounds llg and 18a at all concentrations significantly suppressed TPH-1 protein expression in a dose dependent manner. Besides, compounds llg and 18a at 10 μM significantly down-regulated TPH-1 mRNA expression in the RBL2H3 cells in a time dependent manner, which was more potent than the effects on TPH-2 and its mRNA expression. The results indicated that compounds llg andl8a exerted a selective inhibition on TPH-1 to a certain extent.The interactions between 11g,18a and TPH-1 were analyzed by surface plasmon resonance (SPR) and the results showed that compounds llg and 18a have the capability of specifically recognizing the immobilized TPH-1.4. In Vivo Activity of llg and 18aAdministration of llg and 18a at doses of 10 and 20 mg/kg per day for 30 days resulted in a significantly decrease of serum serotonin concentration in a dose-dependent manner compared with the control group in ovariectomized (OVX) rats. The results also demonstrated that llg and 18a did not cause marked changes of serotonin levels in brain at all doses, which was crucial, as brain- and gut-derived serotonin exerts opposite influences on bone formation. Histomorphometric analyses of lumbar vertebrae (L2-L4) with microcomputed tomography (μCT) indicated that llg and 18a at dosages of 10 and 20 mg/kg markedly increased the bone quality related parameters including BMD (bone mineral density), BV/TV (bone volume over total volume), Tb.N (trabecular number), Tb.Th (trabecular thickness), and decreased Tb.SP (trabecular space). The results indicated that llg and 18a could effectively prevent OVX-induced bone loss in vivo.Besides, OVX induced bone loss is accompanied with higher bone turnover rates resulted in increased bone formation and resorption. ELISA analysis revealed that llg and 18a could increase the P1NP (N-terminal propeptide of procollagen type 1) level in serum, but did not affect CTX-1 (carboxy terminal telopeptide of collagen type I) level in serum. Overall, our results demonstrated that llg and 18a rescued OVX induced bone loss in rats through a bone anabolic fashion.From the perspective of drug safety we evaluated the side effects of llg and 18a on special organs such as uterus. The uteri of OVX rats with 30-day administration of llg and 18a at all doses exhibited no changes compared with those of untreated OVX rats, indicating llg and 18a have no estrogenic side effects. Besides, morphological and pathological changes were not observed in the liver tissue.Overall, 11g and 18a could prevent OVX induced bone loss in vivo by inhibiting serotonin biosynthesis, and may serve as a worthy leader for bone anabolic drug discovery. |
PDF Full Text Request