Biological extraction of realgar by Acidithiobacillus ferrooxidans and its in vitro and in vivo antitumor activities

Abstract

Acidithiobacillus ferrooxidans is a Gram-negative, chemolithoautotrophic bacterium involved in metal bioleaching. It is used for the extraction of coarse medical realgar, which is converted into an aqueous solution. To prove its feasibility as an anticancer drug candidate, extracted realgar (ER/Af) was evaluated for its antitumor activities both in vitro and in vivo. In cytotoxicity tests, ER/Af displayed significant inhibition on cell proliferation of HepG2, SMMC7721, and H22 cells in a time and dose dependent manner. Remarkable tumor growth inhibition and survival time prolongation effects, along with no obvious toxicity, were observed in antitumor experiments against H22 cell-bearing mice. Apoptosis induction was also confirmed as one of the mechanisms involved in the efficacy of ER/Af both in vitro and in vivo. The most important observation is that ER/Af showed high selective affinity to tumor tissues with about eight-fold higher arsenic accumulations at the tumor site of mice than those of the arsenic trioxide (ATO)-treated group at the same dose (57.8 ± 3.34 μg/g dry tissue vs. 7.6 ± 0.88 μg/g dry tissue). In conclusion, A. ferrooxidans could be successfully used for the extraction of realgar and ER/Af was proved to be a promising anticancer drug candidate, which is valuable for further study and clinical trials.

Keywords::

• Acidithiobacillus ferrooxidans
• antitumor
• apoptosis
• bioleaching
• extraction
• realgar

Introduction

Since the 1990s, discovery that arsenic trioxide could induce complete remission in a high percentage of patients with acute promyelocytic leukemia (APL) (Zhang et al., 1996; Shen et al., 1997), the application of arsenic and its compounds for the therapy of human diseases has been attracting a great deal of attention. Realgar, one of the medical mineral ores and the main component of which is bi-arsenic bi-sulfide (As₂S₂), has been used as a traditional medicine in China and Europe for more than 1500 years (Wu et al., 2002). Employed either externally or orally, realgar is administered to treat a variety of diseases, including psoriasis, syphilis, malaria, and parasitic infections caused by Plasmodium and Schistosoma japonicum (Shen et al., 1997; Dilda & Hogg, 2007). In recent years, some researchers in China found that it was clinically effective to cure patients with APL and chronic myelogenous leukemia (CML) (Li et al., 2002; Lu & Wang, 2002; Lu et al., 2002). However, it has been reported that a large dose (0.75–3.75 g once a day) of realgar and a long period (2–9 weeks) of treatment was necessary for clinical complete release (CR) on both APL and CML because of its water insolubility. This also results in the problem of high toxicity, which greatly hampers the clinical application of coarse realgar (Prentis et al., 1988; Wu et al., 2002). Since realgar is an attractive and promising arsenic-containing traditional medicine for cancer therapy, urgent need still exists for the development of a specific and effective processing method, improving on the problems of realgar, i.e. solubility, bioavailability, curative effect, and toxicity.

In this study, in order to overcome the problems mentioned above and provide a possible strategy for cancer drug development, we attempted to establish a biological method for the extraction of realgar based on bioleaching, which is a well established hydrometallurgical technology that uses chemolithotrophic sulfur- and/or iron-oxidizing bacteria to convert insoluble metal sulfides into water soluble metal sulfates (Rawlings, 1998). In a recent study, we succeeded in the recovery of arsenic from realgar by pure and mixed indigenous Acidithiobacillus subsp. (Zhang et al., 2007). Acidithiobacillus ferrooxidans BY3 (A. ferrooxidans BY3), one strain of gram-negative, chemolithoautotrophic bacterium, was isolated and purified from an acid mine-drainage system in Northwestern China and was successfully used to convert coarse realgar into an aqueous solution. In order to demonstrate the potency of the extracted realgar (ER/Af) for cancer therapy, its in vitro and in vivo antitumor activities were examined on hepatic cancer due to this being one of the world’s most common malignant neoplasms (Mariko et al., 2002a).

Materials and methods

Mineral and drugs

Realgar, the main component of which is As₂S₂ and which is always mixed with the more toxic arsenic trioxide, was obtained from Shimen County, Hunan Province, China, and purified through traditional methods according to the Chinese Pharmacopeia (Zhen, 2005). Contents of arsenic and sulfur were 78.0% (w/w) and 21.1% (w/w), respectively (Zhang et al., 2007). Arsenic trioxide for injection (ATO; Yierda^®) with an arsenic concentration of 10 mg/mL was purchased from Yida Medicinal Ltd. (Harbin, China).

Microorganisms, cell lines, and animals

A. ferrooxidans BY3 (CCTCC-M 204057), which is a Gram-negative, chemolithoautotrophic bacterium involved in metal bioleaching, was isolated and purified from an acid mine-drainage system in Northwestern China. The strain was cultured in 9 K medium (Silverman & Lundgren, 1959): 3.0 g (NH₄)₂SO₄, 0.1 g KCl, 0.5 g K₂HPO₄, 0.5 g MgSO₄·7H₂O, 0.01 g Ca(NO₃)₂, and 44.78 g FeSO₄·7H₂O per 1000 mL, pH 1.8.

Human hepatocellular carcinoma HepG2 and SMMC7721 cell lines were obtained from the Key Laboratory of Preclinical Study for New Drugs of Gansu Province (Lanzhou, China). Mouse H22 cell line was purchased from the Institute of Cancer Research of Gansu Province (Lanzhou, China). All these cells were maintained in RPMI-1640 medium, supplemented with 10% heat-inactivated bovine serum, 100 U/mL penicillin, and 100 mg/mL streptomycin. Cells were maintained at 37°C in a humidified atmosphere consisting of 5% CO₂.

Healthy Kunming mice (Specific Pathogen Free grade, 6–8 weeks, weight 18–22 g) were purchased from the Laboratory Animal Center of Gansu College of Traditional Chinese Medicine (Lanzhou, China). All protocols were approved by the Lanzhou University Animal Care and Use Committee. Animals were segregated according to gender and housed five per plastic cage. They were maintained in a room at a temperature of 25 ± 1°C, with photoperiod of 12 h (from 06:00 to 18:00), and with frequent air changes. Mice had free access to tap water and food, except for a short fasting period before the treatment.

Biological extraction of realgar

Extraction of realgar was carried out in 250 mL flasks with 100 mL iron-free 9 K medium with 1.5 g realgar powder and 1 g Fe²⁺/L as additional energy source. A. ferrooxidans BY3 was cultured in 9 K medium to an exponential growth phase, and was inoculated at 20% (v/v). All flasks were shaken at 150 rpm and 30°C for 30 days. The supernatant was harvested by centrifuging at 10,000 rpm for 10 min, and then adjusted to pH 7.0 with 1 M NaOH and 2 M Na₂EDTA (ethylenediaminetetraacetic acid). Then, it was sieved through 0.22 μm filter membrane to remove bacteria and used as ER/Af throughout this study. The total arsenic concentration was measured by inductively coupled plasma-atomic emission spectrometry (ICP-AES; Jobin-Yvon Ultimate 2R).

Chemical characterization of ER/Af

The accurate concentration of each element in ER/Af was analyzed by ICP-AES. Capillary zone electrophoresis (CZE) was performed on a CE-L1 capillary electrophoresis system (CE Resources, Singapore). Standard separation compound of pyridine 2,6-dicarboxylic acid (PDC) was purchased from Fluka (Buchs, Switzerland). n-Hexadecyltrimethylammonium hydroxide (CTAOH) (25% in methanol) was purchased from Tokyo Kasei Kogyo (Tokyo, Japan). Potassium arsenate (KH₂AsO₄; iAs^V), dimethylarsenic acid [(CH₃)₂AsO(OH); DMA^V], and arsenic trioxide (As₂O₃) were purchased from Sigma (St. Louis, MO, USA). Sodium monomethyl arsonate (CH₄AsNaO₃·(3/2)H₂O; MMA^V) was purchased from Chem Service (West Chester, PA, USA). The aqueous stock solutions of iAs^V, MMA^V, and DMA^V, each at a concentration of 1000 ± 5 ppm, were respectively prepared by dissolution in Milli-Q water. The parameters were as follows: background electrolyte (BGE) composed of 10 mM chromate, 12.5 mM borate, and 0.5 mM CTAB (cetyltrimethylammonium bromide) at pH 9.4; U_setting = −25 kV and I_setting = 15 μA; detection wavelength 216 nm; temperature 20°C (Wu & Ho, 2004).

In vitro experiments

In vitro cytotoxicity of ER/Af and ATO on HepG2, SMMC7721, and H22 cell lines was determined by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay as described previously (Cetin & Bullerman, 2005) with some modifications. In brief, approximately 10⁴ cells per well were seeded in triplicate into 96-well plates, and then cells were incubated with a range of drug concentrations (0.15– 10 mg/mL) for 24, 48, and 72 h, respectively. Afterwards, 10 μL MTT (5 mg/mL) was added to a final concentration of 0.5 mg/mL in each well. Four hours later, an equal volume of dimethylsulfoxide (DMSO) was added to dissolve the blue crystals of formazan by gentle shaking of the plate. The optical density (OD) value was detected in the microplate reader (Multiskan MK3; Labsystems, Finland) at 570 nm. The inhibitory rate was calculated by Equation (1). Drug concentrations that caused a 50% reaction in cell proliferation (IC₅₀) were calculated according to the inhibitory rate and are expressed as the mean ± SD.

Apoptosis and cell cycle distributions for H22 cells were identified and quantified by flow cytometry (FCM). Cells were treated with various concentrations of ER/Af for 48 h. At the end of the incubation, all cells were collected and fixed in 70% ice-cold ethanol at a cell density of 1 × 10⁶/mL, and kept in the refrigerator (–20°C) at least overnight until analysis. Fixed cells were washed twice with phosphate buffered saline (PBS) and treated with 1 mg/mL RNase (DNase free) (Sigma Chemical Co.) for 30 min at 37°C. Propidium iodine (PI, ≥95% purity; Sigma Chemical Co.) was then added to the solution at a final concentration of 50 μg/mL. Samples were filtered through a 30 μm pore size nylon mesh before analysis. Cells with DNA content less than the cells in G1 phase (sub-G1) were taken as apoptotic cells.

In vivo experiments

All experiments in this section were performed at the Center of Medical Sciences, Lanzhou University, and approved by the Animal Care and Use Committee of Lanzhou University. Kunming mice bearing ascitic-type H22 cells were sacrificed 6–8 days after tumor implant. The skin was sterilized and removed and the ascitic fluid was drawn through the muscles of the abdominal wall using a sterile syringe, then diluted to a concentration of l × 10⁷ cells/mL in Hanks solution; 0.2 mL of diluted solution (about 2 × 10⁶ cells) was inoculated subcutaneously to Kunming mice (female:male, 1:1) at the axillary region to establish a solid type hepatic carcinoma H22 mouse model.

Ninety-six hours after inoculation, the H22 solid tumor bearing mice were randomized into five groups, i.e., groups receiving different doses of ER/Af at 1.5, 3, 6 mg/kg (milligrams arsenic per kilogram body weight of mouse), an ATO group (3 mg/kg), and a negative control (normal saline), with 10 mice in each group. All drugs were adjusted to proper concentration and 0.2 mL of each drug was administered intraperitoneally, once a day, for 7 consecutive days or until death of the mouse. General reactions were observed every day after administration. The body weight of the mouse was also monitored as an index of systemic toxicity. All mice were dissected on the 8th day post-inoculation and the tumors were excised and weighed. A portion of the tumor was cut into small pieces, fixed in 2.0% glutaraldehyde fluid, and then examined and photographed under a transmission electron microscope (TEM; Jeol JEM-2010 HR).

Each tested organ (heart, liver, spleen, lung, and kidney) was also excised, washed, accurately weighed, and digested with 10 mL of nitric acid per 0.5 g tissue overnight at room temperature and then boiled for 2 h. The digestions were diluted into 25 mL with ion-free H₂O and the arsenic concentrations were measured by ICP-AES. The accumulation of arsenic in each organ is expressed as micrograms arsenic per gram of dry organ.

Statistical analysis

All experimental data are expressed as mean value ± standard deviation (SD). Experimental data were analyzed using one-way analysis of variance (ANOVA) and significant differences between every two groups were assessed by the least significant difference (LSD) test.

Results

Extraction of realgar by A. ferrooxidans BY3

A. ferrooxidans is one of the most important bacterial strains used in hydrometallurgy, and has been successfully employed for the leaching of many kinds of metals such as manganese, gold, copper, uranium, and nickel from their sulfides (Rawlings, 1998). Concerning the activity of A. ferrooxidans and the main component of realgar (As₂S₂), an attempt was made on the extraction of realgar by A. ferrooxidans for the purpose of improving its biological activities. An extraction rate of 26.64% could be achieved in 30 days under optimal conditions (pH 2.2; pulp density, 0.5 wt-%; inoculum volume, 20%; ferrous content, 1 g/L; particle size, 180–200 mesh), while for the group at the same conditions without bacteria, a much lower extraction rate of realgar (3.22%) was reached (Figure 1). The supernatant of the bacterial extractions was simply treated and used as ER/Af in the following studies.

Figure 1.  Extraction of realgar by bacterial and acidic methods. Extraction rate of realgar was calculated according to the amount of arsenic in the supernatant which was measured by ICP-AES. Results are expressed as means of three separate experiments. Variability of data is small (standard deviation normally less than 10% of respective mean) and is not shown.

Chemical characterization of ER/Af

As an extraction of a mineral drug (realgar), the main components of ER/Af must be inorganic ions. Therefore, total elemental analysis was done by ICP-AES (Table 1). Zn, Cu, Mg, Mn, Fe, Se, Cd, Pb, S, and As were confirmed to be contained in ER/Af, while others were at levels lower than the detection limit and are not shown in Table 1.

Table 1.  Amounts of elements existing in ER/Af measured by ICP-AES.

Element	Concentration (μg/mL)^1
Zn	0.51 ± 0.05
Cu	2.31 ± 0.13
Mg	51.69 ± 3.25
Mn	42.38 ± 5.86
Fe	998.6 ± 11.15
Se	1.59 ± 0.25
Cd	0.39 ± 0.12
Pb	0.36 ± 0.06
S	2705 ± 248.5
As	824.5 ± 18.09

¹Concentration of each element in ER/Af.

Arsenic is the main active ingredient in realgar and must play a large role in the pharmacological effects of ER/Af. Concerning the species-dependent toxicity and activity of arsenic, CZE analysis was employed to identify the arsenic species in ER/Af. Results show that the main arsenic species contained in ER/Af were inorganic trivalent and pentavalent arsenic (iAs^III and iAs^V). A certain amount of monomethylarsonate [CH₄AsNaO₃·(3/2)H₂O; MMA^V] was also detected in ER/Af, possibly due to the bio-detoxification of arsenic in A. ferrooxidans.

Inhibition of cell proliferation on cancer cell lines in vitro

The growth of HepG2 cells was gradually inhibited by ER/Af in a concentration-dependent manner (Figure 2). The inset in Figure 2 demonstrates the typical sigmoidal growth curve for HepG2 cells in complete RPMI-1640 medium. In the presence of different concentrations of ER/Af, cell growth was obviously inhibited. Survival cell number even decreased from the third day after co-incubation with the highest concentration of ER/Af (2.0 μg/mL) (Figure 2). Similar profiles of inhibition were obtained on SMMC7721 and H22 cell lines. However, H22 cells exerted more resistance to ER/Af (data not shown).

Figure 2.  Effect of ER/Af on cell growth of HepG2 cells at various concentrations. Cell numbers after treatment with ER/Af were evaluated by the Trypan blue dye exclusion test. Results are expressed as mean ± SD (n = 3).

The MTT assay was carried out to estimate the IC₅₀ (drug concentration that causes a 50% reduction in cell proliferation) values of ER/Af on the three cell lines for 24, 48, and 72 h, respectively. Arsenic trioxide (ATO), which is commercially available and has been approved to be effective in treating patients with APL, was also used as a control. IC₅₀ values for ER/Af and ATO on different strains of hepatic cancer cell lines are shown in Table 2.

Table 2.  IC₅₀ values of ER/Af and ATO on HepG2, SMMC7721, and H22 cells.

	IC50 value (μg/mL)
	HepG2	SMMC7721	H22
ER/Af
24 h	6.08 ± 0.12^b	1.54 ± 0.04^b	0.51 ± 0.11^ab
48 h	0.98 ± 0.08^cd	0.38 ± 0.02^d	0.35 ± 0.03^bc
72 h	0.43 ± 0.06^d	0.22 ± 0.03^e	0.24 ± 0.08^c
ATO
24 h	8.46 ± 0.47^a	1.87 ± 0.12^a	0.63 ± 0.16^a
48 h	1.02 ± 0.22^c	0.52 ± 0.03^c	0.32 ± 0.03^c
72 h	0.58 ± 0.08^cd	0.34 ± 0.04^de	0.21 ± 0.02^c
LSD0.05 value	0.55	0.13	0.17

Differences obtained at levels of p < 0.05 were considered significant. Different letters in each cell line column indicate significant differences assessed by LSD test.

In vivo antitumor effect of ER/Af on H22 tumor bearing mice

In order to further demonstrate the tumor growth inhibitory effect of ER/Af, H22 cells in solid and ascitic tumor bearing models were established. Table 3 shows the tumor growth inhibitory effect and toxicity of ER/Af in solid tumor bearing mice. It suggests that both ATO and ER/Af at different dosages could significantly inhibit the growth of H22 cells transplanted in mice (p < 0.05). Additionally, except for the group of ER/Af (6.0 mg/kg), all the treatments showed no obvious toxic side effects, as indicated by the results for lethal toxicity and body weight gain of tumor bearing mice.

Table 3.  Tumor growth inhibition and toxicity of ER/Af on H22 solid tumor bearing mice.

Group	Lethal toxicity^1	Body weight gain (%)	Tumor weight (g)	TGIR^2 (%)
Normal saline	2/10	49.20	2.98 ± 0.93^a	0
ER/Af (1.5 mg/kg)	0/10	36.27	1.92 ± 0.53^b	35.56
ER/Af (3.0 mg/kg)	0/10	23.56	1.71 ± 0.29^b	42.78
ER/Af (6.0 mg/kg)	4/10	–8.22	1.07 ± 0.35^b	63.95
ATO (3.0 mg/kg)	2/10	19.44	1.73 ± 0.61^b	41.87

¹Number of dead mice/total number of mice.

²Tumor growth inhibitory rate versus control mice.

Differences obtained at levels of p < 0.05 were considered significant. Different letters in tumor weight column indicate significant differences assessed by LSD test (LSD_0.05 = 0.96).

The life extending effect of ER/Af and ATO on H22 ascitic tumor bearing mice was examined after treatment for 7 days, and the results are presented in Table 4. Treatments with ER/Af at the dosages of 1.5 and 3.0 mg/kg produced life extending rates of 66.07% and 118.8%, respectively. These values were also significantly different from that of the normal saline group (p < 0.05). However, there were no significant differences between the groups of ER/Af and ATO at the same concentration in both solid and ascitic tumor models.

Table 4.  Life extending effects of ER/Af and ATO on H22 ascitic tumor bearing mice.

Group	n	Survival time (days)	Life extending rate (%)
Normal saline	10	11.2 ± 1.17^c	—
ER/Af (1.5 mg/kg)	10	18.6 ± 3.23^b	66.07
ER/Af (3.0 mg/kg)	10	24.5 ± 5.24^a	118.8
ATO (3.0 mg/kg)	10	21.5 ± 4.06^ab	91.96

Differences obtained at levels of p < 0.05 were considered significant. Different letters in survival time column indicate significant differences assessed by LSD test (LSD_0.05 = 3.53).

Apoptosis induction as one of the mechanisms involved in the antitumor activity of ER/Af

Transmission electron microscopy (TEM) was used to examine the ultrastructural changes in tumor cells transplanted subcutaneously in mice. Compared to the normal saline group, the electron density was obviously lowered, and typical features of apoptosis could be seen in the ER/Af treated group (Figure 3). As shown in Figure 3B, the chondriosome and rough endoplasmic reticulum had totally disintegrated and dissolved, giving a vesicle-like structure. Obvious margination and fragmentation of nuclear chromatin was observed, with some parts disseminating in the cell plasma and other parts showing a loose, net-like structure.

Figure 3.  Effect of ER/Af on the ultrastructure of H22 cells transplanted in mice. (A) Tumor cells grew very well in the saline group (×5000 magnification). (B) In the ER/Af group (3.0 mg/kg), tumor cells showed typical features of apoptosis (×5000 magnification).

To further demonstrate the apoptosis-inducing effect of ER/Af on H22 cells, flow cytometry (FCM) analysis was employed for the quantification of cell cycle distributions after treatment with ER/Af in vitro (Figure 4). Cells in the sub-G1 phase were considered as apoptotic cells. The percentage of apoptotic cells in the untreated control (normal saline) was 8.7%. Cells treated with different concentrations of ER/Af (1.0 and 2.0 μg/mL of arsenic) had a significantly higher percentage of apoptotic cells (15.8 and 24.3%) after 48 h of treatment. Moreover, it can also be seen in Figure 4 that these changes were accompanied by a reduction of cells in G0/G1 phases. ATO was also used as a control in cell cycle analysis. Expectedly, ER/Af showed a comparative capacity for apoptosis induction on H22 cells in vitro.

Figure 4.  Effect of ER/Af and ATO on cell cycle distribution of H22 cells assessed by flow cytometry after 48 h of treatment. Percentages of sub-G1 and G0/G1 cells at relevant concentrations in ER/Af or ATO are indicated.

Discussion

Bioleaching is a well-established commercial technology for the recovery of metals from their sulfide minerals due to its ease of construction, use, expandability, maintenance, environmental soundness, and cost advantages as well (Rawlings, 1998; Gilbertson, 2000). In recent decades, mechanisms of this biological process have been fully studied. Two dissolution mechanisms are discussed: direct and indirect. The direct mechanism assumes the action of a metal sulfide-attached cell oxidizing the mineral by an enzyme system with oxygen to sulfate and metal ions. In contrast, the indirect mechanism basically comprises the oxidizing action of ferric ions dissolving a metal sulfide. The bacteria, i.e., Acidithiobacillus subsp., act as an oxidizer for ferrous ions, which are produced in the dissolution of metal sulfides (Rawlings, 1998; Sand et al., 2001).

The main component of realgar is As₂S₂, which is one of the typical metal sulfides. In our early studies, we succeeded in the recovery of arsenic from medical realgar by pure and mixed indigenous Acidithiobacillus subsp. (Zhang et al., 2007). As confirmed, realgar can be effectively dissolved by A. ferrooxidans, A. thiooxidans, and their mixed cultures as well. The dissolution of realgar is demonstrated to be highest in the mixed culture supplemented with 2 g/L ferrous sulfate in the bioleaching solution. In the mixed culture, ferrous sulfate is oxidized to ferric sulfate first by A. ferrooxidans. Then, realgar is oxidized by ferric ions to release arsenic and form sulfur, which can be absorbed on the surface of the mineral and inhibit the oxidation of realgar. A. thiooxidans can oxidize sulfur to sulfuric acid, which accelerates the reaction of ferrous to ferric iron, removes the diffusion barrier covering the surface, and allows ferric iron to further react with realgar. The highest dissolution rate of realgar in the mixed culture can be demonstrated by the following equations (Zhang et al., 2007):

In this study, the bioleaching solution of realgar by A. ferrooxidans under optimal conditions (pH 2.2; pulp density, 0.5 wt-%; inoculum volume, 20%; ferrous content, 1 g/L; particle size, 180–200 mesh) was selected for antitumor experiments. ATO was selected for control due to its defined mechanisms and effects on hepatocellular carcinoma (Mariko et al., 2002a, 2002b; Siu et al., 2002). Results show that ER/Af exerted a significant antitumor effect on hepatic cancer both in vitro and in vivo. It is worth mentioning that the primary contribution of the proposed biological method is complete settlement of the problems associated with insolubility, since ER/Af existed in a form of aqueous solution. Additionally, ER/Af may be administered easily in all kinds of dosage forms in both clinicial and preclinicial studies due to its existing form of aqueous solution. High bioavailability can also be achieved, which results in much higher antitumor activity than that of coarse realgar. The tumor growth inhibitory rate of ER/Af on H22 solid tumor bearing mice at the dose of 3.0 mg/kg reached 42.78% (Table 3). This is equivalent to those of realgar particles and nanometer size realgar at doses of 100 and 50 mg/kg, respectively (Xu et al., 2006).

Toxicity is the most concerning problem in the clinical use of realgar. Through biological extraction with A. ferrooxidans, the toxicity of realgar was obviously improved. As shown in Table 3, ER/Af at doses of 1.5–3.0 mg/kg displayed few side effects on body-weight change, and no abnormal behavior or acute lethal toxicity was observed in those treatment groups. When the experiments were finished, mice were sacrificed, each tested organ was excised, and tissue distributions of arsenic were analyzed by ICP-AES. As shown in Figure 5, there were no significant accumulations of arsenic in normal tissues (heart, liver, spleen, lung, and kidney) in both ER/Af and ATO treated groups. Interestingly, the arsenic accumulations at the tumor site of mice treated with ER/Af at a dose of 3.0 mg/kg were about eight-fold higher than those of the ATO treated group at the same dose (57.8 ± 3.34 μg/g dry tissue vs. 7.6 ± 0.88 μg/g dry tissue). There was no significant difference between the normal saline control and the ATO group. This result indicates that the coexistence of different ingredients mentioned above might greatly increase the selective affinity of arsenic to tumors, which probably reduced the toxicity of ER/Af to tumor bearing mice. The acute toxicity of ER/Af and ATO given as a single dose by the intraperitoneal route was then examined in healthy mice. The calculated LD₅₀ value for the intraperitoneal dose of ER/Af was 6.63 mg/kg. This was a little higher than that of ATO (6.22 mg/kg), although there were no significant differences.

Figure 5.  Tissue distribution of arsenic in normal tissues (heart, liver, spleen, lung, and kidney) and solid tumor. Results show the amounts of arsenic in 1 g of tissue and are expressed as mean ± SD (n = 5).

In conclusion, we have established an effective extraction method for realgar, in which realgar was biologically converted into an aqueous solution. This might be a complete settlement of the problems in the clinical use of realgar. ER/Af also displayed significant antitumor activity against hepatic cancer both in vitro and in vivo. The most important observation is that ER/Af showed far higher selective affinity to tumor tissues than did ATO. These results indicate that ER/Af is likely to be a promising anticancer drug candidate with high performance and low toxicity.

Declaration of interest: This work was funded by the Science and Technology Program of Gansu Province (Grant Nos. 2GS035-A52-008-01, 2GS064-A43-019-02), Science Foundation of the Ministry of Education (Grant No. 107108), and the International Cooperative Program of Gansu Province (Grant No. 0708WCGA150) in P. R. China.