Synthesis and Cytotoxic Activity of Novel C-23-Modified Asiatic Acid Derivatives

We selectively oxidized the C-23 hydroxyl group in an asiatic acid (AA) derivative and then, for the first time with AA, modification of the C-23 carboxyl group was conducted to synthesize a series of new AA derivatives. The evaluation of their cytotoxic activities against two human cancer cell lines (SKOV-3 and HCT116) using the MTT assay in vitro revealed a distinctive structure activity relationship (SAR) associated with the intramolecular hydrogen bonding of the amide moiety at C-23. According to the established SAR, the cytotoxic activities of four promising compounds were then evaluated against MCF-7, A549, A2780, HepG2 and HL-60 cancer cell lines. Compound 10 had the best cytotoxic activity among all tested derivatives in the HL-60 cell line, giving IC50 = 0.47 μM, while showing no cytotoxic effect against human normal cells (HUVEC).


Introduction
Natural products are a major source of lead compounds for new therapeutic agents [1,2]. In recent years, triterpenoids, as one of the largest classes of natural products, have been frequently reported for their remarkable bioactivities including anticancer effects; in particular, triterpenoid acids, such as oleanolic acid (OA) [3], asiatic acid (AA) (Figure 1) [4,5] and maslinic acid (MA) [6,7], have received increased attention.

Biology
To establish a SAR of the C-3 (R 2 ) and C-23 (R 1 ) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC 50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug.

Biology
To establish a SAR of the C-3 (R2) and C-23 (R1) of the AA derivatives, new compounds were subjected to MTT assays to evaluate their in-vitro antiproliferative activities against ovarian (SKOV-3) and colon (HCT-116) cancer cell lines, the IC50 values (the concentration that inhibits 50% of cell growth) of which are summarized in Table 1. Doxorubicin was used as the positive drug. OAc OAc HN(CH 2 ) 3 COOCH 2 CH 3 >100 >100 14j OAc Molecules 2020, 25 For compounds 13f-i, with the introduction of various amino acid esters at C-23, their cytotoxic activities against the HCT116 cell line decreased progressively along with that of the larger group at C-23. In addition, the cytotoxic activities of compound 13f against the SKOV-3 cell line improved considerably, and yet compound 13h was not active. These combined results suggested that the smaller amide moiety at C-23 led to increased cytotoxic activity.
For the modification at C-3, compounds 9 and 10 were 8-and 23-fold more active, respectively, For compounds 13f-i, with the introduction of various amino acid esters at C-23, their cytotoxic activities against the HCT116 cell line decreased progressively along with that of the larger group at C-23. In addition, the cytotoxic activities of compound 13f against the SKOV-3 cell line improved considerably, and yet compound 13h was not active. These combined results suggested that the smaller amide moiety at C-23 led to increased cytotoxic activity.
For the modification at C-3, compounds 9 and 10 were 8-and 23-fold more active, respectively, than AA against HCT116 cell line, and compound 9 was 3-fold more active than AA against SKOV-3 cell line, while the 3β-hydroxy oxidation product, compound 11, lost activity towards both cell lines. In addition, compounds 6 and 12 were both much more active than compound 5 against the HCT116 cell line. These results indicated that the carbonyl moiety at C-3 decreased the cytotoxicity.
We studied the intramolecular hydrogen bonds of the 23-amide AA derivatives via exhaustive NMR characterization ( Table 2). The assignments of C-2, C-3, C-23 and NH-23 of compounds 13b and 14b were unambiguously achieved by a combination of 1 H, 13   According to Table 2, after the acetylation of 3β-hydroxy, the chemical shift at C-2, C-3, C-23 and NH-23 had changed because the loss of 3β-OH donor influenced or deprived the intramolecular hydrogen bonds. Thus, we could deduce the hydrogen bond from the change of the chemical shift (Table 2). Specifically, we could deduce the hydrogen bond connecting 3β-OH donor with 2α-OCH 3 acceptor from the change of the chemical shift at C-2 and C-3. We could deduce the hydrogen bond connecting 3β-OH donor with 23-carbonyl acceptor from the change of the chemical shift at C-23. Additionally, we could deduce the hydrogen bond connecting NH-23 donor with 3β-OH acceptor from the change of the chemical shift at NH-23. Further, the chemical shifts of two amide hydrogen atoms at NH 2 -23 in compound 14a were apparently different, which revealed the hydrogen bond connecting NH-23 donor with 3β-acetoxy acceptor. By analogy, it was possible to form the hydrogen bonds in 3β-acetoxy derivatives 14. To sum up, the hydrogen bonding modes A and B (Figure 2) for 3β-hydroxy derivatives 13 were supposed to exist in the solvent at the same time, and mode C for 3β-acetoxy derivatives 14 may exist.  Table 2, after the acetylation of 3β-hydroxy, the chemical shift at C-2, C-3, C-23 and NH-23 had changed because the loss of 3β-OH donor influenced or deprived the intramolecular hydrogen bonds. Thus, we could deduce the hydrogen bond from the change of the chemical shift (Table 2). Specifically, we could deduce the hydrogen bond connecting 3β-OH donor with 2α-OCH3 acceptor from the change of the chemical shift at C-2 and C-3. We could deduce the hydrogen bond connecting 3β-OH donor with 23-carbonyl acceptor from the change of the chemical shift at C-23. Additionally, we could deduce the hydrogen bond connecting NH-23 donor with 3β-OH acceptor from the change of the chemical shift at NH-23. Further, the chemical shifts of two amide hydrogen atoms at NH2-23 in compound 14a were apparently different, which revealed the hydrogen bond connecting NH-23 donor with 3β-acetoxy acceptor. By analogy, it was possible to form the hydrogen bonds in 3β-acetoxy derivatives 14. To sum up, the hydrogen bonding modes A and B (Figure 2) for 3β-hydroxy derivatives 13 were supposed to exist in the solvent at the same time, and mode C for 3β-acetoxy derivatives 14 may exist. Surprisingly, for compounds 13a-i and 14a-i, we found that the 23-amide AA derivatives which showed rather good cytotoxic activities, could fix the amide hydrogen atom in the AA skeleton through intramolecular hydrogen bonding, no matter what the mode (A, B or C). Comparatively, the 23-amide AA derivatives which were not active, were unable to fix the amide hydrogen atom in the AA skeleton. Specifically, for compounds 13, according to Table 2, the amide hydrogen atom at C-23 of 13b-h could be fixed in the AA skeleton through mode A or B, the amide hydrogen atom of 13i could be fixed through mode B, and the amide hydrogen atom of 13a could be fixed through mode A. After the acetylation of 3β-hydroxy, the cytotoxic activities of compounds 14 against HCT116 cell line decreased progressively along with that of the larger group at C-23, and 14c-i even lost its cytotoxic activities, probably because the steric hindrance of the amide moiety at C-23 blocked the intramolecular hydrogen bond in mode C. Therefore, compounds 13a-g, i and 14a which fixed the amide hydrogen atom in the AA skeleton, exhibited increased cytotoxic activity against HCT116 cell line. Similarly, compounds 13e-f and 14a showed rather good cytotoxic activities against the SKOV-3 cell line. Comparatively, compounds 14c-i, which were unable to fix the amide hydrogen atom in the AA skeleton, were not active in either cell line. Additionally, compounds 9, 10, 13j and 14j without the amide hydrogen atom at C-23 exhibited a considerable improvement in the cytotoxic activities against the HCT116 cell line. To summarize, the cytotoxicity results suggested that the freely rotating amide hydrogen atom at C-23 led to the loss of cytotoxic activity.
Hence, the cytotoxic activities of compounds 10, 13k, 14j and 14k without any amide hydrogen atom at C-23 were evaluated in a panel of additional five cancer cell lines (MCF-7, A549, A2780, Surprisingly, for compounds 13a-i and 14a-i, we found that the 23-amide AA derivatives which showed rather good cytotoxic activities, could fix the amide hydrogen atom in the AA skeleton through intramolecular hydrogen bonding, no matter what the mode (A, B or C). Comparatively, the 23-amide AA derivatives which were not active, were unable to fix the amide hydrogen atom in the AA skeleton. Specifically, for compounds 13, according to Table 2, the amide hydrogen atom at C-23 of 13b-h could be fixed in the AA skeleton through mode A or B, the amide hydrogen atom of 13i could be fixed through mode B, and the amide hydrogen atom of 13a could be fixed through mode A. After the acetylation of 3β-hydroxy, the cytotoxic activities of compounds 14 against HCT116 cell line decreased progressively along with that of the larger group at C-23, and 14c-i even lost its cytotoxic activities, probably because the steric hindrance of the amide moiety at C-23 blocked the intramolecular hydrogen bond in mode C. Therefore, compounds 13a-g, i and 14a which fixed the amide hydrogen atom in the AA skeleton, exhibited increased cytotoxic activity against HCT116 cell line. Similarly, compounds 13e-f and 14a showed rather good cytotoxic activities against the SKOV-3 cell line. Comparatively, compounds 14c-i, which were unable to fix the amide hydrogen atom in the AA skeleton, were not active in either cell line. Additionally, compounds 9, 10, 13j and 14j without the amide hydrogen atom at C-23 exhibited a considerable improvement in the cytotoxic activities against the HCT116 cell line. To summarize, the cytotoxicity results suggested that the freely rotating amide hydrogen atom at C-23 led to the loss of cytotoxic activity.
Hence, the cytotoxic activities of compounds 10, 13k, 14j and 14k without any amide hydrogen atom at C-23 were evaluated in a panel of additional five cancer cell lines (MCF-7, A549, A2780, HepG2 and HL-60) using the MTT assay (Table 3). 3β-acetoxy derivative 14k was more active than 3β-hydroxy derivative 13k against A549, A2780, HL-60 and MCF-7 cell lines, and was commensurately active with 14k against HepG2 cell lines. Combined with the results against the HCT116 cell line shown in Table 1, 3β-acetoxy derivatives 10 and 14j exhibited increased activity compared to the corresponding 3β-hydroxy derivatives 9 and 13j. These combined results suggested that 3β-acetoxy moiety improved the cytotoxic activity, in the case of a lack of amide hydrogen atoms at C-23. Additionally, the cytotoxic activities of compounds 10, 14j and 14k were evaluated against one human normal cell line (HUVEC) using CCK-8 assay (Table 3). Compounds 10, 14j and 14k at the concentrations of 125, 25, 5, 1, 0.2, 0.04 µM all showed no cytotoxic effect on HUVEC cell viability ( Figure 3). These results indicated that the cytotoxicity of compounds 10, 14j and 14k against cancer cells was much higher than that against HUVEC normal cells. HepG2 and HL-60) using the MTT assay (Table 3). 3β-acetoxy derivative 14k was more active than 3β-hydroxy derivative 13k against A549, A2780, HL-60 and MCF-7 cell lines, and was commensurately active with 14k against HepG2 cell lines. Combined with the results against the HCT116 cell line shown in Table 1, 3β-acetoxy derivatives 10 and 14j exhibited increased activity compared to the corresponding 3β-hydroxy derivatives 9 and 13j. These combined results suggested that 3β-acetoxy moiety improved the cytotoxic activity, in the case of a lack of amide hydrogen atoms at C-23. Additionally, the cytotoxic activities of compounds 10, 14j and 14k were evaluated against one human normal cell line (HUVEC) using CCK-8 assay (Table 3). Compounds 10, 14j and 14k at the concentrations of 125, 25, 5, 1, 0.2, 0.04 μM all showed no cytotoxic effect on HUVEC cell viability ( Figure 3). These results indicated that the cytotoxicity of compounds 10, 14j and 14k against cancer cells was much higher than that against HUVEC normal cells.

General
Asiatic acid was purchased from Xi'an Shouhe Biotechnology Co., Ltd. All reagents and solvents were obtained from Sinopharm. IR spectra were recorded on a THERMO IS5. NMR spectra ( 1 H, 13 C, DEPT, HSQC and HMBC) were recorded on a Bruker Ascend-400 400 MHz or Bruker Ascend-500

General
Asiatic acid was purchased from Xi'an Shouhe Biotechnology Co., Ltd. All reagents and solvents were obtained from Sinopharm. IR spectra were recorded on a THERMO IS5. NMR spectra ( 1 H, 13 C, DEPT, HSQC and HMBC) were recorded on a Bruker Ascend-400 400 MHz or Bruker Ascend-500 500 MHz at room temperature. HRMS were recorded on an Agilent 6530-Q-TOF mass spectrometer equipped with an Agilent 1260-HPLC.

In-Vitro Cytotoxicity of Human Normal Cell Line
HUVEC cell lines were incubated in DMEM with 10% heat-inactivated FBS in a humidified atmosphere of 5.0% CO 2 at 37 • C. The in-vitro cytotoxicity of AA derivatives was determined using CCK-8. A 100-µL volume of culture medium at a concentration of 3 × 10 4 cell/mL was added to 96-well plate wells, and these were maintained in a humidified atmosphere of 5.0% CO 2 at 37 • C. After 24 h, 10 µL solution of AA derivatives dissolved in DMSO and PBS and then diluted with DMEM under a concentration gradient was added to the wells for final concentrations of 125, 25, 5, 1, 0.2, 0.04 µM. After 72 h incubation in a humidified atmosphere of 5.0% CO 2 at 37 • C, 10 µl CCK-8 solution was added into each well and the cells were incubated for another 1 h, and the absorbance was read at 450 nm. All cell viability levels were averaged from three independent experiments.

Conclusions
In this paper, we reported a highly selective oxidation method on the C-23 hydroxyl group in an AA derivative, and then, for the first time with AA derivatives, the C-23 carboxyl group was modified to synthesize a series of novel AA derivatives. The selective oxidation method provided a high-yield route for more modification in the C-3 hydroxyl, C-23 aldehyde and C-23 carboxyl groups in AA.
The 23-amide AA derivatives 10, 14j and 14k exhibited good cytotoxic activities against tumor cell lines but were significantly less cytotoxic against normal human cell line, showing excellent selectivity.
We studied the intramolecular hydrogen bonds of the 23-amide AA derivatives via exhaustive NMR characterization, and combined with their cytotoxic activities against SKOV-3 and HCT116 cell lines, which revealed a distinctive SAR associated with the intramolecular hydrogen bonding of the amide moiety at C-23. We found that the 23-amide AA derivatives, which fixed the amide hydrogen atom in the AA skeleton through intramolecular hydrogen bonding, showed rather good cytotoxic activities. Intramolecular hydrogen bond has been actively applied in drug design for its effects on receptor binding and its improved molecular properties, such as membrane permeability, water solubility, and lipophilicity. When a donor and an acceptor are in proximity on the same molecule, an equilibrium may exist between closed conformations in which an intramolecular hydrogen bond is formed, creating a temporary ring system, and open conformations in which the polar groups are exposed to solvent. The closed forms, hiding polarity from the environment, should be more lipophilic and might display a higher membrane permeability, whereas the open forms should be more water-soluble [33]. Especially for the triterpenoids, with a rigid skeleton that often leads to poor water solubility, dynamic and proper water solubility and lipophilicity may be crucial for activity and the pharmacokinetic profile. In this paper, these active AA derivatives with intramolecular hydrogen bonds in the A ring, can function as ideal starting points for further modification to develop compounds with better activities and improved molecular properties.

Conflicts of Interest:
The authors declare no conflict of interest.