Anti-hepatocellular carcinoma activity and mechanism of chemopreventive compounds: ursolic acid derivatives

Abstract

Context: Hepatocellular carcinoma (HCC) is a common cancer around the world, with high mortality rate. Currently, there is no effective drug for the therapy of HCC. Ursolic acid (UA) is a natural product which exists in various medicinal herbs and fruits, exhibiting multiple biological effects such as its outstanding anticancer and hepatoprotective activity, which has drawn many pharmacists’ attention.

Objective: This paper summarizes the current status of the hepatoprotective activity of UA analogues and explains the related mechanism, providing a clear direction for the development of novel anti-HCC drugs.

Methods: All of the data resources were derived from PubMed. By comparing the IC₅₀ values and analyzing the structure–activity relationships, we listed compounds with good pharmacological activity from the relevant literature, and summarized their anti-HCC mechanism.

Results: From the database, 58 new UA derivatives possessing wonderful anticancer and hepatoprotective effects were listed, and the relevant anti-HCC mechanism were discussed.

Conclusion: UA’s anti-HCC effect is the result of combined action of many mechanisms. These 58 new UA derivatives, particularly compounds 45 and 53, can be used as potential drugs for the treatment of liver cancer.

Keywords:

• HCC
• mechanisms
• structural modification
• structure activity relationship

Introduction

Along with the improvement of people’s living standards, the risk of liver cancer increases. HCC is the fifth most common cancer in men and the seventh in women (El-Serag 2012), and the third leading cause of cancer-related death, only behind the lung and the stomach cancer (Ferlay et al. 2010). Liver cancer is caused by many factors (Figure 1), including infection with HBV and HCV (Honda et al. 2001; Goossens & Hoshida 2015; Wang et al. 2015), nonalcoholic liver disease (Ekstedt et al. 2006), exposure to aflatoxin B₁ and cadmium (Liu & Wu 2010; Satarug 2012), hemochromatosis, alpha-antitrypsin deficiency, autoimmune hepatitis, some porphyrias, and Wilson′s disease (El-Serag 2011). Moreover, alcohol, tobacco and obesity are also independent risk factors for HCC in patients, and they interact synergistically to increase the risk of HCC (Marrero et al. 2005). In addition to the above factors, El-Serag et al. (2004) found that diabetes doubled the risk of chronic liver disease and hepatocellular carcinoma (HCC). Sasaki (2010) reported that insulin resistance was a consequence of HCC because it easily caused obesity and hepatic inflammation, both of which could promote hepatocarcinogenesis by themselves, through the production of cytokine and/or oxidative stress. Garcia-Compean et al. (2009) held that insulin resistance could decrease cure rates in patients with chronic hepatitis C (CHC) and enhance liver fibrosis. In genetics, lacking of the enzyme glucose-6-phosphatase, which is essential for glucose release from liver glycogen, is also a cause of HCC and hepatic adenomas (Bannasch 2010).

Throughout the ages, there are four main methods of treating liver cancer, including chemotherapy, drug treatment, liver resection and liver transplantation. Natural products, a rich source of compounds, was applied in various fields of medical treatment including cancer therapy (Cragg et al. 2009), and played a dominant role in the development of sophisticated traditional medicine systems. In our diet, pentacyclic triterpenoids are considered relatively nontoxic to humans (Kuttan et al. 2011). As we all known, ursolic acid (UA, compound 1) is the most predominant representative of pentacyclic triterpenoids and possesses considerable pharmacological activities, including: anti-HIV (Slamenová et al. 2006), anticancer (Rashid et al. 2013), anti-inflammatory (Jung et al. 2013), antidiabetic (Alqahtani et al. 2013; Lee et al. 2014; Wu et al. 2014), and antibacterial activities (Zhao et al. 2014). Importantly, it is also reported to have hepatoprotective activity (Liu 1995). Moreover, traditional Chinese medicine containing UA has been used for the treatment of liver disease for several decades. Here, we will discuss the anti-HCC activity and the mechanism of UA and its derivatives.

Figure 1. The possible cause of liver cancer and the structure of compound 1.

The anti-hepatocellular carcinoma mechanism of UA

UA widely exists in fruits and herbs with low toxicity, which makes it suitable for cancer metastatic chemoprevention (Novotny et al. 2001). Accumulating mechanistic studies indicate that the anti-HCC effect of UA is attributed to its direct inhibition of HCC cells growth processes and indirect suppression of carcinogenic factor (Figure 2). In terms of inhibition of HCC cell growth, UA can inhibit HCC cell proliferation, invasion and metastasis. Ramos et al. (2008) found that UA prevented DNA damage via increase of DNA repair and anti-proliferative properties in HepG2 cells. Lin et al. (2011) found that UA reversed the invasion and migration of Hep3B and Huh7 cell lines at 4 μmol/L. Son et al. (2013) reported that UA blocked HepG2 cellular metabolism via AMPK activation and GSK3b phosphorylation, and thus induced apoptosis of HepG2 cells. On the suppression of carcinogenic side, UA could prevent HCC from nonalcoholic fatty liver disease (NAFLD), insulin resistance, inflammation and oxidation stress. UA was demonstrated to decrease diet-induced obesity, glucose intolerance and NAFLD by increasing skeletal muscle mass, fibre size, brown fat weight and energy expenditure (Kunkel et al. 2012). NAFLD has high risk of liver fibrosis and cirrhosis, even hepatocarcinoma (Brunt 2005). But, nowadays many results showed UA may ameliorate disease progression in patients with NAFLD. Li et al. (2015) suggested that UA treatment (10–30 μg/mL) inhibited the development of NAFLD via increasing lipid β oxidation and inhibiting the hepatic endoplasmic reticulum stress. Jia et al. (2011) showed that UA was a PPAR-α agonist which regulated hepatic lipid metabolism and, therefore, significantly reduced intracellular triglyceride concentrations in hepatocytes. In addition, Wang et al. (2015) and Sundaresan et al. (2014) found that the combination of UA with artesunate or rosiglitazone exhibited the same pharmacological effect and free of toxic side effects. Li et al. (2014) indicated that UA reversed high-fat diet-induced hepatic steatosis. For HCC caused by insulin resistance, Wang et al. (2012) proved that UA ameliorated insulin resistance in liver of KKAy mice via activation of peroxisome proliferator-activated receptors α and γ. Castro et al. (2015) showed that UA as insulin secretagogue and insulin-mimetic also regulated the glucose balance in the liver of KKAy mice. Both liver inflammation and oxidation are regarded as hallmark of early-stage fibrosis, which aggravate fibrosis and cirrhosis and eventually deteriorate into liver cancer. However, Ma et al. (2014, 2015) demonstrated that UA protected mouse liver against CCl₄-induced oxidative stress and inflammation by the MAPK/NF-κB pathway, and meliorated the symptom of liver fibrosis via the Nrf2/ARE pathway. Liu et al. (2014) also revealed that UA, Viili exopolysaccharides (VEPS), Astragalus polysaccharides (APS) and any combination of these reduced the activities of COX-2, SOD, PGE₂ as well as markedly decreased MDA. Hence, the reduction of internal anti-oxidative elements created an over-oxidative environment for HepG2 cells, triggering HepG2 cells apoptosis and the inhibition of HepG2 cell proliferation.

As a promising antifibrosis agent, UA treats liver fibrosis via inhibition of hepatic stellate cells (HSCs) (Shyu et al. 2008). He et al. (2015) proved that UA reduced the accumulation of type I collagen in rat HSCs through attenuating the activity of NOX4 and downregulating ERK, PI3K/Akt, and p38 MAPK signalling pathways in HSCs. Meanwhile, successful antimicrobial therapy is good for liver cancer through declining the morbidity and mortality of HCC and reducing the risk of liver failure in patients with advanced fibrosis (Rutter et al. 2015). Kong et al. (2013) discovered that UA significantly suppressed the replication of HCV genotype 1b replicon and HCV genotype 2a JFH1 virus by suppressing HCV NS5B RdRp activity as noncompetitive inhibitors. Li et al. (2012) also confirmed that UA showed inhibitory activity for HBV, another pathogenic virus, with IC₅₀ values of 89.91 μM for HBsAg and 97.61 μM for HBeAg at no cytotoxicity. Wu et al. (2011) supported that UA could break hepatitis B virus X protein (HBx)-mediated tumourigenic effects. Besides that, UA could directly accelerate liver mass recovery and induce the hepatocyte proliferation after partial hepatectomy by the stimulation of C/EBPβ (Jin et al. 2012).

Figure 2. The anti-hepatocellular carcinoma mechanism of UA.

The modification of UA

UA exhibits low water solubility, rapid metabolism and low bioavailability, which limit its anti-HCC activity. Because of its structural constraints, considerable structural modification has been performed on UA and potentially important derivatives have emerged. Shao et al. (2011) concluded from a series novel UA derivatives (compounds 2–7, Table 1) that: (1) keeping a polar group at either the 3-OH and/or 17-COOH position was necessary for antitumour effect; (2) increasing the length of the carbon chain decreased the derivatives’ inhibitory effect on the cancer cells due to long chains could obstruct cells from contacting with the derivatives′ mother nucleus; (3) connecting diethanolamine at the 17-COOH position and acetylating the 3-OH was beneficial to the anticancer activities. Moreover, piperazine ring had better pharmacodynamic effects than the piperidine ring at the 17-COOH position; (4) introducing a more bulky group decreased UA derivatives’ binding ability to the target because of steric hindrance. Analyzing the mechanism of compound 3 demonstrated that it suppressed HepG2 cells by cell-cycle arrest at the S phase, and increased the activity of caspase-3 at the dosage of 100 mg/kg.

Table 1. UA derivatives with diethanolamine conjugation suppressed tumour growth.

Compound	R1	R2	IC50 (μM)
2	NHCH(CH2OH)COOCH3	CH3CO	35.26 ± 1.26
3	N(CH2CH2OH)2	CH3CO	20.25 ± 1.52
4		CH3CO	20.46 ± 2.53
5		CH3CO	15.26 ± 4.10
6		H	35.66 ± 0.33
7		H	36.71 ± 1.36

2-Deoxy-d-glucose (2-DG), as an essential inhibitor of glucose metabolism, successfully inhibits tumour growth in animal experiments and human clinical trials (Maschek et al. 2004; Zhong et al. 2009; Raez et al. 2013). Selective targeting of cancer glucose metabolism provided an alternative strategy of anticancer drug development with minimum side effects on normal cells. Dong et al. (2014) showed that piperazine moiety-modified UA (compound 8, Table 2) further enhanced the therapeutic effects of 2-DG probably through synergistic suppression of cancer cell glucose metabolism. Subsequently, Yang et al. (2015) demonstrated that UA bound to cyclins D1 (Cyc D1) and cyclin-dependent kinases (CDK6), while compound 9 targeted at glucokinase (GK), glucose transporter 1 (GLUT1) and ATPase. Therefore, modification at the C-28 position of UA with the piperazine moiety (compounds 9–11) exhibited better anticancer activity than against HepG2 by interfering with glucose metabolism in a dose-dependent manner. Wang et al. (2014) also synthesized a series of UA derivatives (compounds 12–19), and found that UA diamine derivative combined with 2-DG showed synergistic inhibition for hepatoma cell proliferation by dual-targeting of apoptosis and glycolysis. Furthermore, explicating its structure–activity relationship helped diamine connected at C-28 enhance efficacy and selectivity of UA; additionally, the length of six carbon chains (6C) of diamine might represent the most optimal length for UA to bind the active sites of hexokinase and show the most excellent competitiveness, hence downregulating the glycolysis metabolism of cancer cells. Xiang et al. (2015) further proved that the mechanism of compound 14 inhibiting HepG2 cell attributed to its suppression of cell-surface CAMs and integrin α6β1. The adjustment of ester metabolism and free radical scavenging also associated with the treatment of liver cancer. Recently, Kazmi et al. (2013) noticed that UA stearoyl glucoside (UASG, compound 20), a newly discovered triterpene in Lantana camara, significantly suppressed diethylnitrosamine (DENA)-induced hepatocellular carcinogenesis by scavenging the hydroxyl radicals, modulating the levels of lipid peroxidation (LPO) and increasing the endogenous antioxidant enzymes level. UASG was also able to repair the HCC-caused histological changes at concentrations of 40 and 80 mg/kg.

Table 2. UA derivatives inhibit HCC through glucose and ester metabolism pathway.

Compound	R1	R2	IC50(μM)
8		CH3CO	5.15 ± 0.31
9		CH3CO	4.78 ± 0.68
10		H	5.09 ± 0.76
11		H	10.18 ± 0.20
12	NH(CH2)2NH2	CH3CO	16.32 ± 1.76
13	NH(CH2)4NH2	CH3CO	15.97 ± 2.18
14	NH(CH2)6NH2	CH3CO	7.20 ± 2.04
15	NH(CH2)8NH2	CH3CO	17.34 ± 2.53
16	NH(CH2)2NH2	H	13.9 ± 0.78
17	NH(CH2)4NH2	H	14.44 ± 1.38
18	NH(CH2)6NH2	H	13.42 ± 2.31
19	NH(CH2)8NH2	H	26.65 ± 1.40
20	OH		NG

Thiourea derivatives have potential anti-cancer activity (Huang et al., 2013a, 2013b). Accordingly, Hua et al. (2015) synthesized a series of piperazine-thiourea UA derivatives (compounds 21–40, Table 3) by combining thiourea, piperazine and UA, and reached the following conclusions: (1) the synchronous introduction of piperazine-thiourea at C-28 and an acyl group at C-3 was significant for improving their activity, however, only introduction of an acyl piperazine thiourea at C-28 led to the opposite effect on anti-cancer activity; (2) the activity of introducing electron-donating group to phenyl ring was higher than electron-withdrawing group; (3) para-substituted phenyl ring showed better anti-cancer activity than meta-substituted. Among these piperazine-thiourea UA derivatives, trimethoxyphenyl thiourea UA (compound 38) showed the best efficiency. Moreover, compound 38 displayed higher levels of anti-proliferative activities compared with UA and 5-fluorouracil (5-FU) (IC₅₀ was 5.4 ± 0.79 μM) and exhibited much lower cytotoxicity than 5-FU in HepG2 cells line.

Table 3. Piperazine-thiourea UA derivatives.

Compound	R	IC50 (μM)
21		39.19 ± 0.93
22		45.21 ± 0.20
23		28.29 ± 1.43
24		36.93 ± 0.38
25		44.54 ± 0.29
26		58.74 ± 0.76
27		47.22 ± 0.76
28		54.14 ± 0.19
29		34.54 ± 0.59
30		38.69 ± 0.88
31		41.44 ± 0.47
32		51.58 ± 0.95
33		48.55 ± 0.65
34		32.36 ± 0.81
35		35.98 ± 0.85
36		26.11 ± 0.82
37		15.72 ± 0.84
38		5.40 ± 0.79
39		65.08 ± 0.98
40		37.44 ± 1.36

Carbazole, a nitrogen-containing aromatic heterocyclic compound, attracted more attention owing to its excellent anti-cancer activity (Shaikh et al. 2015). Prompted by the above facts, Gu et al. (2015) synthesized a series of novel carbazole derivatives of UA (compounds 40–46, Table 4). Among these derivatives, some compounds showed significant cytotoxic activity to HepG2 cells with IC₅₀ values below 10 μM. Especially in compound 45, IC₅₀ was only 1.26 ± 0.17 μM. This outstanding result indicated that: (1) the incorporation of carbazole moiety to the molecule was beneficial to the anticancer activity; 2) the N-(dimethylamino) propyl moiety was crucial for their cytotoxic activity. But, esterification at the C-28 carboxyl group led to reduction of activities; (3) the indole benzene rings with lipophilic substitutes exhibited stronger inhibitory activity than hydrophilic substitutes; (4) the indole benzene rings with p-substitution had higher cytotoxicity than o-substitution. Similarly, Leal et al. (2012) introduced a nitrogen-containing heterocyclic ring(s) to UA, such as imidazole and triazole, and obtained compounds 47–51 (Table 5) displaying better anti-proliferative effects than UA.

Table 4. Carbazole derivatives of UA.

Compound	R1	R2	IC50 (μM)
41	H	H	2.78 ± 0.21
42	F	H	6.03 ± 1.12
43	Cl	H	4.02 ± 0.86
44	Br	H	5.65 ± 1.43
45	CH3	H	1.26 ± 0.17
46	OCH3	H	6.25 ± 1.37

Table 5. Imidazole and triazole derivatives of UA.

Compound	R1	R2	IC50 (μM)
47	O		8.5 ± 0.06
48	O		7.9 ± 0.03
49	H		2.0 ± 0.03
50	H		4.0 ± 0.01
51	H		3.2 ± 0.03

Kalani et al. (2014) used an anti-HepG2 test and showed that 4-bromoanilamide derivative of UA (compound 52, Figure 3) was 212 times more efficient than the starting material, and 11 times more efficient than the positive control, paclitaxel. Amide derivatives were better than the ester derivatives against HepG2 cancer cell lines. A UA fluorolactone derivative (compound 53), 2-cyano-3-oxo-12a-fluoro-urs-1-en-13,28b-olide, improved the anti-proliferative activity against HepG2 cell lines, with IC₅₀ value lower than 1 μM (Leal et al. 2012). Functional liver cancer stem cells (LCSCs) could rapidly self-renew and infinitely proliferate. CD133⁺, as a surface marker, isolated LCSCs. UA chalcone (compound 54) targeted PLC/PRF/5 hepatoma cell line and Huh7 LCSCs, inhibiting the expression of CD133⁺in a dose- and time-dependent manner. The inhibition was significant at 50 μM and on day 8 (Lin et al. 2015). Concurrently, in natural, two new ursane-type triterpenes, microfokienoxane C (compound 55) and 3β, 28-dihydroxy-11α-methoxyurs-12-ene (compound 56), had recently been isolated from the leaves of Microtropis fokienensis and Ilex cornuta, respectively. Compound 55 was found to be the strongest cytotoxic activity against HepG2 and Hep3B cancer cell lines, with IC₅₀ values of 3.8 and 4.5 μg/ml, respectively. Compound 56 also exhibited cytotoxic activity against the HepG2 cell line, with IC₅₀ values of 4.6 μg/ml (Chen et al. 2006).

Figure 3. Structures of UA derivatives.

High concentrations of NO can promote HCC cells apoptosis via intervening with liver inflammatory. Thence, the combination of NO and UA probably protect hepatocytes from inflammation and toxicant-mediated liver damage. On this basis, Chen et al. (2011) connected UA with furoxan for a series of novel NO-releasing compounds, which could produce high levels of NO in the liver. Among these NO-releasing compounds, compound 57 showed the best anti-cancer activity against HepG2 cells, with IC₅₀ values of 3.2 μM. A novel codrug (compound 58), connecting lamivudine (LMV) with UA via ester linkage, had the dual action of anti-hepatitis B virus activity and hepatoprotective effects against acute liver injury. Moreover, in vitro, the high concentration of LMV and UA indicated that the ester bond of the codrug could be easily broken by esterases (Zhong et al. 2012).

Except chemical structural modification, great efforts also had been put in the modification of dosage form and drug delivery system. PEGylation is an effective method for increasing solubility in water, controlling permeability through biological barriers, longevity in the blood stream, and controlling release (Greenwald et al. 2003). Therefore, Zacchigna et al. (2014) synthesized mPEG-UA which increased solubility in water and stability in physiological buffer and delivered free and active UA in a prolonged manner in plasma. Zhao et al. (2015) found that PEG-modified UA liposomes not only possessed higher solubility but also lower cytotoxic effect than conventional UA liposomes. Nanotechnology-based drug delivery systems are effective methods to increase the solubility and dissolution rate of water-insoluble compounds. Then, Song et al. (2014) loaded UA into nanocrystals without stabilizer. Compared with the raw material, the UA nanocrystals exhibited improved aqueous dispensability and dissolution rate. Moreover, the suspension of UA nanocrystals showed better physically stability after storage at 4 °C for 7 weeks. Zhang et al. (2015) loaded UA into amphiphilic poly (N-vinylpyrrolidone)-block-poly (ɛ-caprolactone) nanoparticles (UA-NPs). The UA-NPs significantly delayed tumour growth and localized to the tumour site via altering the in vivo biodistribution and prolong the retention time of UA. The UA non-covalent complex with hydrophilic cyclodextrins not only improved water solubility but also optimized bioavailability and stability (Soica et al. 2008). However, in pharmacokinetic study, changing UA delivery carrier improved its bioavailability. UA liposomes and UA nanoliposomes were released and accumulated in the liver; therefore, Zhu et al. (2013) suggested UA nanoliposomes for safely dose levels within the range of 37–98 mg/m². In this drug concentration range, no drug accumulated was observed even with 14 days of continuous IV infusion. Moreover, the dose of UA liposomes for a phase II clinical trial recommended by Qian et al. (2015) is 98 mg/m².

Discussion

With many pharmacological activities, UA’s hepatoprotective effect possesses great potential in the future for the development of anti-HCC drugs. The mechanism of UA’s resistance to HCC was proved of great complexity. UA mainly blocks HCC cells growth and migration through inhibiting HCC cells glucose, ester metabolism and replication of genetic material. UA also has the function of promoting liver cells regeneration to repair damaged liver. Additionally, UA possesses anti-inflammatory, antioxidation, anti-HCV/HBV and antidiabetic activities which can assist in treating HCC. It could also relieve obesity, insulin resistance and hyperlipidemia-induced NAFLD, and therefore, decreases the incidence of HCC. However, in spite of the virtue of low toxicity, the clinical application of UA is still limited because of its low water solubility, which results in the compromised bioavailability and therapeutic efficacy. Therefore, there have been many research works on structural modification of UA due to its structural limitations. So far the modification of UA focused on the C-3 hydroxyl group, the C-28 carboxyl group and the double bond. Upon the modification of the C-3 hydroxyl group, some research works found that acylation was better than hydrolysis, and substituted nitrogen heterocyclic at C-3 was better than at C-28. In the C-28 carboxyl group, the anti-HCC activity of UA derivatives was improved by introducing nitrogen-containing group. But, attention still needed to be paid to the size and length of the added carbon chain, otherwise, the steric blocking would decrease the activity. C-1/C-21 and C-7/C-21 dihydroxylation is useless for anti-HCV activity, as well as C-1 hydroxylation and simultaneous introduction of carbonyl group at C-21 (Fu et al. 2013). In addition to chemical structural modification, covalent modification is also an important method to enhance the activity of UA, which includes conjugating highly water-soluble macromolecular groups, like PEG, NP and cyclodextrin to improve UA′s aqueous solubility as well as using nanoparticles, nanocrystals and nanosuspension to prolong the retention time and enhance dissolution velocity. However, the surface of polyamidoamine (PAMAM) conjugated with UA and folic acid (FA) showed no effect on the cellular uptake for HepG2 cells (Gao et al. 2015). Despite the low toxicity of UA, we should closely monitor their doses in clinic.

Conclusion

UA is a natural product with many pharmacological activities, including anticancer and hepatoprotective activity. To date, UA has been applied in liver disease treatment. However, its in vitro pharmacological activities reduced greatly because of its limited water solubility. Therefore, in order to find new effective anti-HCC compounds, many studies have focused on the structural modification of UA. The acetylation of the C-3 hydroxyl group and amidation of the C-28 carboxyl group are common and effective ways to improve its bioavailability and activities. Among these derivatives, compounds 45 and 53 showed outstanding anti-HCC activity with IC₅₀ values of 1.26 ± 0.17 μM and lower than 1 μM, respectively, which is worth developing further.

Disclosure statement

The authors report that they have no conflict of interest.

Funding

This research was supported by the National Natural Science Foundation of China (No. 81273537), Scientific Research Fund of Hunan Provincial Education Department (No. 12K095), the key disciplines of Hunan Province and the Project is sponsored by Zhengxing scholar program of the University of South China.