Asymmetric synthesis and biological activities of natural product (+)-balasubramide and its derivatives

Abstract The natural product (+)-balasubramide (3j) and its derivatives (3a–3i) were synthesized using a two-step asymmetric synthesis, and the biological activities of 3a–3j were determined in vitro. Methyl (2S,3R)-(+)-3-phenyloxirane-2-carboxylate (1h), the asymmetric synthesis of which was described in a previous paper, was selected as the starting material. Compounds 3a–3j were evaluated for their neuroprotective, antioxidative, and anti-neuroinflammatory effects. (+)-Balasubramide and its derivatives with different electronegative groups in the 6-phenyl ring produced little neuroprotection and antioxidation, but induced potent anti-neuroinflammatory effects in BV-2 microglial cells (with the exception of 3g). Compound 3c, with a trifluoromethyl group in its 6-phenyl ring, was a particularly potent anti-neuroinflammatory agent. These results demonstrated that the electronegativity of the 6-phenyl ring of (+)-balasubramide is an important determinant of its inhibitory effect on neuroinflammation. More electronegative substituents result in more potent anti-neuroinflammatory effects. Moreover, cytotoxicity assays indicated no significant effects of the tested compounds.


Introduction
(+)-Balasubramide is an eight-membered lactam compound with an absolute configuration of 5S,6R (Juárez-Calderón et al. 2013) that is extracted from the leaves of the Sri Lankan plant Clausena indica (Riemer et al. 1997) along with its biosynthetic precursor, (+)-prebalamide ( Figure 1). Studies have shown that five-membered lactam compounds extracted from C. indica have extensive pharmacological uses. For example, clausenamide (Figure 1) produces hepatoprotective (Yang et al. 1987) and neuroprotective effects (Xue et al. 2008), inhibits apoptosis (Yao et al. 2001) and lipid peroxidation (Lin et al. 1992), and acts as an oxygen free radical scavenger (Jiang & Zhang 1998). However, eight-membered lactam compounds are rarely reported to have pharmacological activity. Total synthesis of (+)-balasubramide is difficult because of its eight-membered lactam ring and two chiral centers. Thus, the biological activities and preliminary structure-activity relationships (SaRs) of related compounds have not yet been reported.
at present, although there are three synthetic routes regarding balasubramide (Johansen et al. 2007;Yang et al. 2007;Zheng et al. 2009), (+)-balasubramide, a chiral natural product, only could be obtained by one method (Johansen et al. 2007) which was synthesized by chiral resolution and the overall yield was only 17%.
We investigated an additional method of total synthesis of (+)-balasubramide to allow us to study its bioactivity and to perform a preliminary analysis of the SaRs of (+)-balasubramide and derivative compounds. We previously reported a short asymmetric synthetic route of methyl (2S,3R)-(+)-3-phenyloxirane-2-carboxylate (1h) (Xuan et al. 2013). In this study, we used compound 1h as the starting material to synthesize a series of (+)-balasubramide derivatives in order to explore their biological activities, including neuroprotection, anti-neuroinflammation, cytotoxic effects.
In recent years, as a catalytic method with the same status of enzyme catalysis and metal catalysis, asymmetric catalysis has been an important tool for building chiral molecular scaffold owing to the high efficiency and selectivity Wang et al. 2015;Woźniak et al. 2015;Zhao et al. 2015). our group initially reported a model reaction of cinnamaldehyde with organocatalyst, a diphenyl prolinol TeS ether, to synthesize methyl (2S,3R)-(+)-3-phenyloxirane-2-carboxylate (1h) via a one-pot reaction with 73% yield and 95% enantioselectivity (Xuan et al. 2013). In this report, we used compound 1h to prepare the linear amide 2j by amine-ester interchange with N-methyltryptamine, followed by intramolecular cyclization using ytterbium(III) triflate (Yb(CF 3 So 3 ) 3 ) as a catalyst, resulting in (+)-balasubramide (3j) (Figure 2).
In order to introduce substituents with varying electronegativity into the 6-phenyl ring, cinnamaldehydes with different substitutions were employed to prepare (+)-balasubramide analogues using organocatalyst followed by asymmetric epoxidation, oxidative esterification, an amine-ester interchange reaction, and intramolecular cyclization ( Figure 3).

Chemistry
In the conversion of methyl (2S,3R)-(+)-3-phenyloxirane-2-carboxylate into the linear amide 2j, an amine-ester interchange reaction with N-methyltryptamine was carried out at room temperature in CH 3 oH for 10 h, but resulted in only a 25% yield. Therefore, we investigated reaction conditions using temperatures less than −18 °C and added catalytic basic additives such as NaHCo 3, K 2 Co 3, Na 2 Co 3 , t-BuoK, and CH 3 oNa. When CH 3 oNa was used as the additive, the process resulted in a good yield of compound 2j. Despite our best efforts, the resultant amide 2j could not be separated from at least one other unidentified compound, we speculated that the compound in column chromatography may be decomposed due to its instability, but the mixture could once again be subjected to next step without using purification. The intramolecular cyclization of compound 2j was found to be greatly influenced by the use of Lewis acid activities. When alCl 3 , FeCl 3 , or CuCl were used, compound 3j was not  obtained, but when LaCl 3 , p-TSa, or Yb(CF 3 So 3 ) 3 was used, compound 3j was obtained with varying yields. Yb(CF 3 So 3 ) 3 produced the best yield with >99% enantioselectivity. although the solvent had no effect on enantioselectivity in the last reaction, a polar solvent was found to increase the yield, as evidenced by the improvement produced by the addition of tetrahydrofuran.

In vitro biological evaluation
The (+)-balasubramide (3j) and its derivatives 3a-3i were evaluated for their in vitro biological activities, including neuroprotective, antioxidative and anti-neuroinflammatory effects. our results in Tables S1-3 showed that compounds 3a-3j didn't exhibit the significant neuroprotective effects in primary neurons against glutamate or nutrient deprivation and antioxidative effects in PC12 neuronal cells against H 2 o 2 incubation. all compounds were further subjected to bioassay to test their in vitro antineuroinflammatory effects against LPS-induced pro-inflammatory cytokine TNFα expression in microglial cells. The results are shown in Figure 4 and Table S4. With the exception of  compound 3g (3-Cl-substituted), the target compounds at 10 μM markedly inhibited LPS-induced pro-inflammatory cytokine TNFα release in BV-2 microglial cells. Compound 3c, with a strongly electronegative substituent (4-CF 3 ), at 1 and 10 μM, exhibited significant inhibition of LPS-induced TNFα release in a dose-dependent manner. In addition, compound 3c dose-dependently inhibited LPS-induced TNFα gene expression ( Figure S1). This indicated that a strongly electronegative substituent in the 6-phenyl ring might be required for the anti-neuroinflammatory effects of (+)-balasubramide derivatives. Meanwhile, in the MTT assay for cytotoxicity of these compounds in BV-2 microglial cells ( Figure S2), (+)-balasubramide and its derivatives 3a-3j did not show cytotoxic effects.

Conclusions
In this report, we describe a convenient and efficient method for the synthesis of natural (+)-balasubramide with a 44% overall yield and excellent enantioselectivity (>99%). Substituents with varying electronegativity were introduced into the 6-phenyl ring of (+)-balasubramide and the resulting target compounds 3a-3j were evaluated for neuroprotective, antioxidative, and anti-neuroinflammatory effects. our results indicated that natural (+)-balasubramide and its derivatives had little neuroprotective and antioxidative effects, but significantly inhibited neuroinflammation (with the exception of 3g). These results show that electronegativity in the 6-phenyl ring of (+)-balasubramide and its derivatives is an important determinant of their inhibitory effects on neuroinflammation; substituents with stronger electronegativity produced more potent anti-neuroinflammatory effects. our preliminary SaR study provided information that could facilitate the design of novel anti-neuroinflammatory drug candidates or leads. These molecular structures may be worthy of future study with the aim of developing new anti-neuroinflammatory agents. Further studies to improve anti-neuroinflammatory activity and clarify the molecular mechanism of these compounds are in progress.

Supplementary material
all experimental sections relating to this article are available online, alongside Tables S1-4 and Figures S1-43.

Disclosure statement
No potential conflict of interest were reported by the authors.