Original Article / Liver

Harmine induces apoptosis in HepG2 cells via

mitochondrial signaling pathway
Ming-Rong Cao, Qiang Li, Zhi-Long Liu, Hui-Hui Liu, Wei Wang, Xiao-Li Liao,
Yun-Long Pan and Jian-Wei Jiang
Guangzhou, China

BACKGROUND: Harmine has antitumor and antinociceptive G0/G1 and increasing the proportion in S and G2/M. Harmine
effects, and inhibits human DNA topoisomerase. However, induced apoptosis in a concentration-dependent manner, with
no detailed data are available on the mechanisms of action of rates of 20.0%, 32.7% and 64.9%, respectively. JC-1 revealed a
harmine in hepatocellular carcinoma. This study aimed to decrease in Ψm. Apoptosis of HepG2 cells was associated with
investigate the effects of harmine on proliferation and apoptosis, caspase-3 and caspase-9 activation, down-regulation of Bcl-2,
and the underlying mechanisms in the human hepatocellular Mcl-1, and Bcl-xl, and no change in Bax.
carcinoma cell line HepG2.
CONCLUSIONS: Harmine had an anti-proliferative effect
METHODS: The proliferation of HepG2 cells was determined in HepG2 cells by inducing apoptosis. Mitochondrial signal
by the cell counting kit-8 (CCK-8) assay and the clone pathways were involved in the apoptosis. The cancer-specific
formation test. The morphology of HepG2 cells was examined selectivity shown in this study suggested that harmine is a
using fluorescence microscopy after Hoechst 33258 staining. promising novel drug for human hepatocellular carcinoma.
Annexin V/propidium iodide (PI) was used to analyze
apoptosis and PI to analyze the cell cycle. Western blotting (Hepatobiliary Pancreat Dis Int 2011; 10: 599-604)
was used to assess expression of the apoptosis-regulated KEY WORDS: hepatocellular carcinoma;
genes Bcl-2, Bax, Bcl-xl, Mcl-1, caspase-3, and caspase-9. harmine;
Mitochondrial transmembrane potential (Ψm) was determined Bcl-2 protein;
using JC-1. caspase-3;
RESULTS: Harmine inhibited the proliferation of HepG2 cells apoptosis
in a dose-dependent manner. Hoechst 33258 staining revealed
nuclear fragmentation and chromosomal condensation, cell
shrinkage, and attachment loss in HepG2 cells treated with Introduction
harmine. The percentage of the sub/G1 fraction was increased

H
epatocellular carcinoma (HCC) is a common and
in a concentration-dependent manner, indicating apoptotic
cell death. PI staining showed that harmine changed the cell aggressive malignant tumor worldwide. Primary
cycle distribution, by decreasing the proportion of cells in liver cancer, which consists predominantly of
HCC, is the fifth most common cancer worldwide and the
third most common cause of cancer mortality. In China,
HCC accounts for 90% of primary liver cancer, which is
the second most common cause of death.[1] Recent data
indicate that the mortality of HCC in China is tending to
increase, severely threatening the health and life of people.
Currently, the treatments for HCC are mainly surgery
Author Affiliations: Department of General Surgery, First Affiliated and chemotherapy, but the curative effect of existing
Hospital, Jinan University, Guangzhou 510632, China (Cao MR, Li
Q, Liu ZL and Pan YL); Department of Anesthesiology, First Affiliated chemotherapeutic drugs is not good enough and they
Hospital, Guangzhou University of TCM, Guangzhou 510632, China (Liu have numerous side-effects, including myelosuppression,
HH); Department of Biochemistry, Medical College of Jinan University, neutropenia, and thrombocytopenia.[2, 3] Therefore, it has
Guangzhou 510632, China (Wang W, Liao XL and Jiang JW)
become a focus to search for drugs which are capable of
Corresponding Author: Jian-Wei Jiang, MD, Department of Biochemistry, preventing and treating HCC and other malignancies. One
Medical College of Jinan University, Guangzhou 510632, China (Tel:
86-20-85220256; Email: [email protected]) possible way to increase the efficacy of anticancer drugs
and decrease toxicity or side-effects is to develop traditional
© 2011, Hepatobiliary Pancreat Dis Int. All rights reserved.
doi: 10.1016/S1499-3872(11)60102-1 medicines, especially from medicinal plants.[4-7] Herbal

Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com • 

 Hepatobiliary & Pancreatic Diseases International

medicines are an important source of novel agents with (300 cells/well) in 2.5 mL of complete RPMI-1640 for
pharmaceutical potential. Harmine is a major component 24 hours. Then the cells were treated with the indicated
isolated from Peganum harmala L. (Zygophyllaceae) seed concentrations harmine for 7 days. Finally, the cells
extract. The pharmacological characterization shows that were stained in crystal violet for 20 minutes. Images of
harmine has antitumor[8, 9] and antinociceptive effects,[10] the colonies were captured by a digital camera.
and inhibits human DNA topoisomerase.[11] However, no
detailed data are available on the mechanisms of action of Fluorescence microscopy assay
harmine in HCC. We investigated its effect on the growth Harmine-induced apoptosis in HepG2 cells was
of human HepG2 cells and the underlying intracellular assessed by Hoechst 33258 staining. Treated with 5
signal transduction pathways involved in regulating μg/mL harmine for 48 hours, the cells were harvested
apoptosis. and smeared on slides. The slides were air-dried, fixed
in methanol-acetone (3/1, v/v), and stained with
Hoechst 33258 (5 μg/mL) at 37 ć for 20 minutes.
Methods Nuclear morphology was examined under fluorescence
Materials microscopy (DFC480; Leica Microsystems, Wetzlar,
Harmine was from Xi'an Feida Bio-Tech Co., Ltd. Germany) to identify cells undergoing apoptosis.
The following were all from Cell Signaling (Danvers,
Massachusetts, USA): antibodies for immunoblots anti- Flow cytometry
caspase-3 (#9662), anti-caspase-9 (#9502), anti-Bcl-2 HepG2 cells at a density of 20 000 cells/well were
(#2870), anti-Mcl-1 (#4572), anti-Bcl-xl (#2764), anti- incubated in 6-well plates for 48 hours with DMSO, 0, 5,
Bax (#2772), and GAPDH (#3683); and cell lysis buffer. 10, or 20 μg/mL harmine. After incubation, the cells were
collected and analyzed by flow cytometry. Then the cells
Cell culture and treatment with harmine were harvested, washed with phosphate buffer solution
Cells were cultured in RPMI-1640 medium (PBS), and fixed in 70% ice-cold ethanol overnight.
supplemented with 10% fetal bovine serum at 37 ć The fixed cells were incubated with 20 U/mL RNase I
in a humidified atmosphere containing 5% CO2. The and 50 μg/mL propidium iodide (PI) for 30 minutes.
harmine was dissolved in dimethyl sulfoxide (DMSO), The DNA content was determined by flow cytometry
and diluted to appropriate concentrations with culture (Beckman Coulter, Fullerton, CA, USA). Apoptotic
medium. The final concentration of DMSO in the cells were identified by the sub-G1 phase in the cell-
culture medium did not exceed 0.1%. cycle distribution. For assessment of the apoptotic rate,
Annexin V-FITC/PI (Becton Dickinson, USA) staining
CCK-8 assay was performed according to the manufacturer's protocol.
The CCK-8 test was used to monitor cell proliferation. The cells stained with Annexin V but not with PI were
Cells were plated at a density of 5000 cells/well in 96-well defined as apoptotic. The apoptotic rate was measured
plates. After 24 hours in culture, the cells were treated by flow cytometry (FCM, Becton Dickinson, USA) using
with harmine at the final concentrations 0, 0.625, 1.25, CellQuest software.
2.5, 5, 10, or 20 μg/mL for 48 hours. Control cells were
treated with DMSO. After addition of test compounds, Western blotting
10 μL CCK-8 was added to each well. Absorbance was Total proteins were extracted by incubation of cell
detected with an enzyme calibrator at 570 nm and then pellets with lysis buffer. The protein concentration
optical density (OD) values were measured. Inhibition of the extracts was determined by bicinchoninic acid
of cell growth was computed by the percentage of viable (Sigma) according to the manufacturer's instructions.
cells compared with control. Percentage (%)=(ODC- Cell lysates (50 μg/well) were electrophoresed in 12%
ODT)/ODC×100%. ODT is the OD value of the treated SDS-PAGE and then transferred onto nitrocellulose
sample, and ODC is the OD value of the control sample. membranes. After blotting in 5% non-fat dry milk
Experiments were done in triplicate. in TBST buffer, the membranes were incubated with
primary antibodies in 5% non-fat milk overnight at
Clone formation assay 4 ć, and then secondary antibodies conjugated with
A clone formation assay was used to evaluate the horseradish peroxidase for 1 hour at room temperature.
effects of harmine on the proliferation of HepG2 Immunoreactive bands were visualized by enhanced
cells.[12] Cells were first cultured in 12-well microplates chemiluminescence regents (Amersham, UK). To assess

 • Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com

Harmine induces apoptosis in HepG2 cells

the presence of comparable amounts of proteins in each harmine on proliferation, the cells were treated with the
lane, the membranes were stripped to assess GAPDH. indicated concentrations for 7 days. The result showed that
harmine strongly inhibited their proliferation (Fig. 1B).
Membrane potential of mitochondria (Ψm)
Changes of Ψm were monitored by determination of Harmine induces apoptosis in HepG2 cells
JC-1. HepG2 cells were treated with 5 μg/mL harmine Apoptotic nuclear morphology was observed after
for 48 hours, then the cells were harvested, washed Hoechst 33258 staining using fluorescence microscopy.
with PBS, and fixed in JC-1 at 37 ć in the dark for 30 After treatment with 5 μg/mL harmine for 48 hours,
minutes, when they were harvested and smeared on HepG2 cells began to exhibit apoptotic characteristics,
slides. Changes in Ψm were measured after staining such as cell shrinkage, nuclear condensation, and
under fluorescence microscopy. fragmentation. In the control group, the cells were
regular in morphology and grew fully in patches and
Statistical analysis were confluent, rarely sloughing off (Fig. 2).
Data were analyzed by ANOVA and Student's t test. Apoptosis was also detected through fluorescein
[13]
These analyses were performed using SPSS 13.0 software. Annexin V-FITC/PI double labeling. The staining
Differences with P<0.05 were considered significant. of cells with Annexin V-FITC and PI was used to

Results
Harmine inhibits HepG2 cell proliferation
Cell viability was decreased remarkably by harmine
treatment (Fig. 1A). Harmine inhibited the growth of
HepG2 cells in a dose-dependent manner (P<0.05 vs
control), with an IC50 of 9.80 μg/mL. To assess the effect of

Fig. 1. A: Effects of harmine on viability of HepG2 cells. The

indicated concentrations of harmine were added, and the cells
were incubated for 48 hours. Viability was measured by CCK-8
assay. Results represented 3 independent experiments. B: Clone Fig. 2. Harmine-induced apoptosis in HepG2 cells was assessed
formation assay was used to evaluate the effects of harmine on by Hoechst 33258 staining. Morphology of HepG2 cells exposed
the proliferation of HepG2 cells. HepG2 cells were treated with to harmine at different concentrations photographed under
0, 0.625, 1.25, 2.5, or 5 μg/mL harmine for 7 days and stained in a f luorescence microscope (original magnification ×200). A:
crystal violet for 20 minutes. DMSO; B: 0 μg/mL; C: 5 μg/mL.

Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com • 

 Hepatobiliary & Pancreatic Diseases International

distinguish and quantitatively determine the percentage Effect of harmine on the protein expression level
of apoptotic cells. HepG2 cells underwent apoptosis Loss of Ψm is a crucial step in the apoptotic process
after exposure to harmine at 5, 10, and 20 μg/mL for and is lethal to cells because it leads to the release of
48 hours. The percentage of apoptotic cells stained diverse pro-apoptotic factors from the mitochondria
by Annexin V-FITC is shown in Fig. 3A. PI staining into the cytoplasm.[14] In this study, we used the
and flow cytometry were used to investigate whether unique cationic dye JC-1 to determine the status of
harmine inhibited the growth of HepG2 cells by mitochondria. In non-apoptotic cells, JC-1 enters the
initiating the apoptotic pathway. HepG2 cells underwent negatively-charged mitochondria where it aggregates and
apoptosis after being exposed to harmine at 5, 10, and turns red. However, in cells undergoing apoptosis, where
20 μg/mL for 48 hours (Fig. 3B). The percentage of Ψm has collapsed, JC-1 exists as monomers in the cytosol
apoptotic cells in the sub-G1 phase of the cell cycle and turns green. Our results showed that harmine
increased. The percentage of the sub/G1 fraction, which induced a depletion of Ψm in HepG2 cells (Fig. 5A).
was indicative of apoptotic cell death in harmine-treated The Bcl-2 family of proteins are key regulators of
cells, increased in a dose-dependent manner. mitochondrial permeability.[15] Therefore, we investigated
whether the mitochondria-mediated apoptosis in HepG2
Harmine changes the cell cycle in HepG2 cells cells induced by harmine was modulated by Bcl-2 family
Flow cytometry with only PI staining showed that members. Harmine suppressed the expression of Bcl-2,
treatment of HepG2 cells with 5, 10, and 20 μg/mL Mcl-1, and Bcl-xl, and did not change the expression
harmine for 48 hours resulted in a higher number of of Bax (Fig. 5B). As a result of these changes, the ratios
cells in S phase and G2/M phase compared with the of Bcl-2/Bax, Mcl-1/Bax, and Bcl-xl/Bax were reduced
control group (Fig. 4). This increase was coupled with significantly during apoptosis. To delineate the possible
the decreased percentage of cells in G0/G1 phase. signaling pathways by which harmine induced HepG2 cell

Fig. 3. Staining cells with Annexin V-FITC and PI was used to distinguish and quantitatively determine the percentage of apoptotic
cells. The rate of apoptosis increased in a dose-dependent manner.

Fig. 4. PI staining showed characteristic features of apoptosis in harmine-exposed cells. The number of cells in S phase and G2/M
phase increased and the percentage of cells in G0/G1 phase decreased compared with the control group.

 • Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com

Harmine induces apoptosis in HepG2 cells

We demonstrated that harmine dose-dependently

inhibited the proliferation of HepG2 cells, and moreover,
it also induced apoptosis and arrested the cell cycle.
Morphological changes in apoptotic characteristics, such
as cellular shrinkage, rounding, poor adherence, and
round floating shapes in harmine-treated cells were also
observed by fluorescence microscopy. Flow cytometry
analysis revealed that harmine treatment resulted in an
increase of apoptotic cells. These results suggested that
the growth inhibition of HepG2 cells by harmine is due
to its ability to induce apoptosis.
Mitochondria play an essential role in apoptosis,
since both the intrinsic and extrinsic apoptosis
pathways converge at the mitochondrial level and
trigger mitochondrial membrane permeabilization.[15]
Mitochondria-mediated apoptosis is precisely regulated
by Bcl-2 family proteins, which can be divided into two
Fig. 5. Harmine induced apoptosis of HepG2 cells through a
mitochondrial signaling pathway. A: Change of Ψm was evaluated
subfamilies:[15, 18, 19] one is anti-apoptotic such as Bcl-2 and
by fluorescence microscopy. HepG2 cells were treated with 5 μg/mL Bcl-xl, the other is pro-apoptotic like Bax, Bad, and Bid.[20]
harmine for 48 hours. B, C: Harmine activated Bcl-2 family The expression of Bcl-2 and Bax is important in the
members, caspase-3 and caspase-9. HepG2 cells were treated balance of pro-apoptotic and anti-apoptotic signals at the
with 0, 1.25, 2.5, and 5 μg/mL harmine for 48 hours. The effect mitochondrial level.[21] In our experiments, harmine had
of harmine on the protein expression level was evaluated by
no effect on expression of the pro-apoptotic Bax protein
immunoblotting. Immunoblotting data were quantitated from
at least three separate experiments. but decreased the expression levels of anti-apoptotic
Bcl-2, Bcl-xl, and Mcl-1, leading to up-regulation of the
Bax/Bcl-2, Bax/Bcl-xl, and Bax/Mcl-1 ratios. This might
be responsible for the concomitant execution phase of
apoptosis, we determined the changes in the expression apoptosis, which included disruption of Ψm.
levels of various apoptosis-regulating proteins. Caspase, a family of cysteine proteases, is an
Caspase, a family of cysteine acid proteases, is integral part of the apoptotic pathway. Caspase-9 is
known to act as an important mediator of apoptosis the apical caspase in the intrinsic or mitochondria-
and contributes to the overall apoptotic morphology initiated apoptosis pathway and requires the release
by cleavage of various cellular substrates. Exposure of of cytochrome C from the mitochondria. Activation
HepG2 cells to harmine resulted in cleavage of caspase-3 of caspase-3 correlates with activation of caspase-9.
and caspase-9 (Fig. 5C). Caspase-3 in particular, when activated, has many
cellular targets.[22] During apoptosis, caspase-3 is one of
the key executioners of apoptosis in response to various
Discussion stimuli.[23] In many studies, it has been determined that
Although contemporary therapeutic strategies have a variety of chemotherapeutic agents induce apoptosis
shown remarkable anticancer ability, severe side- through the activation of caspase.[16] Consistent with
effects are unavoidable. The search for new antitumor an increase in the ratio of Bax/Bcl-2 and disruption of
agents that are more effective but less toxic has kindled Ψm, this study showed that harmine resulted in dose-
great interest. Since apoptosis plays an important dependent activation of caspase-9 and caspase-3.
role in developmental processes, homeostasis, and In conclusion, we found that harmine-induced
the elimination of damaged cells, accumulating data apoptosis was accompanied by modulation of the Bcl-2
indicate that many anticancer drugs cause the death family, mitochondrial dysfunction and activation
of tumor cells through the induction of apoptosis. [16, 17] of caspase in HepG2 cells, implying that harmine
Although harmine shows effective antitumor activity, induced apoptosis via the mitochondrial death pathway.
its effects on human hepatocytes are still unclear. Therefore, we believe that harmine might be a promising
Therefore, the purpose of the present study was to reveal molecule in cancer chemoprevention or chemotherapy
the underlying molecular mechanism by which harmine and further efforts to explore this therapeutic strategy
induces apoptosis in the human HCC cell line HepG2. are needed.

Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com • 

 Hepatobiliary & Pancreatic Diseases International

Funding: This study was supported by grants from the Sci-Tech status on harmaline metabolism, pharmacokinetics and
Project Foundation of Guangdong Province, China (2010B031600248) pharmacodynamics, and a pharmacogenetics-based phar-
and the National Natural Science Foundation of China (30772131). macokinetic model. Biochem Pharmacol 2009;78:617-624.
Ethical approval: Not needed. 11 Singh Y, Palombo M, Sinko PJ. Recent trends in targeted
Contributors: CMR, LQ and JJW proposed the study and wrote anticancer prodrug and conjugate design. Curr Med Chem
the first draft. LQ analyzed the data. All authors contributed to the 2008;15:1802-1826.
design and interpretation of the study and to further drafts. JJW is 12 Shao XD, Wu KC, Guo XZ, Xie MJ, Zhang J, Fan DM.
the guarantor. Expression and significance of HERG protein in gastric
Competing interest: No benefits in any form have been received cancer. Cancer Biol Ther 2008;7:45-50.
or will be received from a commercial party related directly or 13 Chen S, Cheng AC, Wang MS, Peng X. Detection of apoptosis
indirectly to the subject of this article. induced by new type gosling viral enteritis virus in vitro
through fluorescein annexin V-FITC/PI double labeling.
World J Gastroenterol 2008;14:2174-2178.
14 Marzetti E, Wohlgemuth SE, Lees HA, Chung HY,
References Giovannini S, Leeuwenburgh C. Age-related activation of
1 Chen JA, Shi M, Li JQ, Qian CN. Angiogenesis: multiple mitochondrial caspase-independent apoptotic signaling in rat
masks in hepatocellular carcinoma and liver regeneration. gastrocnemius muscle. Mech Ageing Dev 2008;129:542-549.
Hepatol Int 2010;4:537-547. 15 Wang JB, Qi LL, Zheng SD, Wu TX. Curcumin induces
2 Chau GY, Lui WY, Tsay SH, Chao Y, King KL, Wu CW. apoptosis through the mitochondria-mediated apoptotic
Postresectional adjuvant intraportal chemotherapy in pathway in HT-29 cells. J Zhejiang Univ Sci B 2009;10:93-102.
patients with hepatocellular carcinoma: a case-control study. 16 Lah JJ, Cui W, Hu KQ. Effects and mechanisms of silibinin
Ann Surg Oncol 2006;13:1329-1337. on human hepatoma cell lines. World J Gastroenterol 2007;
3 Lo CM, Liu CL, Chan SC, Lam CM, Poon RT, Ng IO, et al. 13:5299-5305.
A randomized, controlled trial of postoperative adjuvant 17 Edderkaoui M, Odinokova I, Ohno I, Gukovsky I, Go VL,
interferon therapy after resection of hepatocellular carcinoma. Pandol SJ, et al. Ellagic acid induces apoptosis through
Ann Surg 2007;245:831-842. inhibition of nuclear factor kappa B in pancreatic cancer cells.
4 Dai ZJ, Wang XJ, Li ZF, Ji ZZ, Ren HT, Tang W, et al. World J Gastroenterol 2008;14:3672-3680.
Scutellaria barbate extract induces apoptosis of hepatoma 18 Suzuki S, Yokoyama U, Abe T, Kiyonari H, Yamashita N, Kato
H22 cells via the mitochondrial pathway involving caspase-3. Y, et al. Differential roles of Epac in regulating cell death in
World J Gastroenterol 2008;14:7321-7328. neuronal and myocardial cells. J Biol Chem 2010;285:24248-
5 El Gendy MA, Somayaji V, El-Kadi AO. Peganum harmala 24259.
L. is a candidate herbal plant for preventing dioxin mediated 19 Ghibelli L, Diederich M. Multistep and multitask Bax
effects. Planta Med 2010;76:671-677. activation. Mitochondrion 2010;10:604-613.
6 Astulla A, Zaima K, Matsuno Y, Hirasawa Y, Ekasari 20 Roos DH, Puntel RL, Lugokenski TH, Ineu RP, Bohrer D,
W, Widyawaruyanti A, et al. Alkaloids from the seeds of Burger ME, et al. Complex methylmercury-cysteine alters
Peganum harmala showing antiplasmodial and vasorelaxant mercury accumulation in different tissues of mice. Basic Clin
activities. J Nat Med 2008;62:470-472. Pharmacol Toxicol 2010;107:789-792.
7 Li H, Wang LJ, Qiu GF, Yu JQ, Liang SC, Hu XM. Apoptosis 21 Hetz CA. ER stress signaling and the BCL-2 family of
of Hela cells induced by extract from Cremanthodium proteins: from adaptation to irreversible cellular damage.
humile. Food Chem Toxicol 2007;45:2040-2046. Antioxid Redox Signal 2007;9:2345-2355.
8 El Gendy MA, El-Kadi AO. Peganum harmala L. differentially 22 Bialik S, Zalckvar E, Ber Y, Rubinstein AD, Kimchi A.
modulates cytochrome P450 gene expression in human Systems biology analysis of programmed cell death. Trends
hepatoma HepG2 cells. Drug Metab Lett 2009;3:212-216. Biochem Sci 2010;35:556-564.
9 Farouk L, Laroubi A, Aboufatima R, Benharref A, Chait 23 Lüthi AU, Martin SJ. The CASBAH: a searchable database of
A. Evaluation of the analgesic effect of alkaloid extract caspase substrates. Cell Death Differ 2007;14:641-650.
of Peganum harmala L: possible mechanisms involved. J
Ethnopharmacol 2008;115:449-454. Received December 27, 2010
10 Wu C, Jiang XL, Shen HW, Yu AM. Effects of CYP2D6 Accepted after revision May 19, 2011

 • Hepatobiliary Pancreat Dis Int，Vol 10，No 6 • December 15，2011 • www.hbpdint.com