1 3
Arch Toxicol (2016) 90:2187–2200
DOI 10.1007/s00204-015-1600-z
INORGANIC COMPOUNDS
2,3,5,6‑Tetramethylpyrazine (TMP) down‑regulated
arsenic‑induced heme oxygenase‑1 and ARS2 expression
by inhibiting Nrf2, NF‑κB, AP‑1 and MAPK pathways
in human proximal tubular cells
Xuezhong Gong
1,2
· Vladimir N. Ivanov
2
· Tom K. Hei
2,3
Received: 26 May 2015 / Accepted: 10 September 2015 / Published online: 24 September 2015
© Springer-Verlag Berlin Heidelberg 2015
ERK, AP-1 and Nrf2 and block HO-1 protein expression.
The present study, furthermore, demonstrated arsenic-
induced expression of arsenic response protein 2 (ARS2)
that was regulated by p38 MAPK, ERK and NF-κB. To
our knowledge, this is the first report showing that ARS2
involved in arsenic-induced nephrotoxicity, while TMP
pretreatment prevented such an up-regulation of ARS2
in HK-2 cells. Given ARS2 and HO-1 sharing the similar
regulation mechanism, we speculated that ARS2 might also
mediate cell survival in this procession. In summary, our
study highlighted a role of HO-1 in the protection against
arsenic-induced cytotoxicity downstream from the primary
targets of TMP and further indicated that TMP may be used
as a potential therapeutic agent in the treatment of arsenic-
induced nephrotoxicity.
Keywords Sodium arsenite · Tetramethylpyrazine
(TMP) · Nephrotoxicity · Mitochondrial dysfunction ·
Inducible heme oxygenase-1 (HO-1) · Arsenic response
protein 2 (ARS2) · Nuclear factor erythroid derived-2
(Nrf2)
Abbreviations
AG Aminoguanidine
AKI Acute kidney injury
ARS2 Arsenic response protein 2
As Arsenic
Bay Bay 11-7082
CKD Chronic kidney disease
DHE Dihydroethidium
FACS Fluorescence-activated cell sorter
HO-1 Heme oxygenase-1
MAPK Mitogen-activated protein kinase
Nrf2 Nuclear factor erythroid derived-2
NAC N-acetylcysteine
Abstract Our recent study demonstrated that sodium
arsenite at a clinically relevant dose induced nephrotoxicity
in human renal proximal tubular epithelial cell line HK-2,
which could be inhibited by natural product 2,3,5,6-tetra-
methylpyrazine (TMP) with antioxidant activity. The pre-
sent study demonstrated that arsenic exposure resulted in
protein and enzymatic induction of heme oxygenase-1
(HO-1) in dose- and time-dependent manners in HK-2
cells. Blocking HO-1 enzymatic activity by zinc protopor-
phyrin (ZnPP) augmented arsenic-induced apoptosis, ROS
production and mitochondrial dysfunction, suggesting a
critical role for HO-1 as a renal protectant in this proces-
sion. On the other hand, TMP, upstream of HO-1, inhib-
ited arsenic-induced ROS production and ROS-dependent
HO-1 expression. TMP also prevented mitochondria dys-
function and suppressed activation of the intrinsic apoptotic
pathway in HK-2 cells. Our results revealed that the regu-
lation of arsenic-induced HO-1 expression was performed
through multiple ROS-dependent signal pathways and the
corresponding transcription factors, including p38 MAPK
and JNK (but not ERK), AP-1, Nrf2 and NF-κB. TMP
inhibited arsenic-induced activations of JNK, p38 MAPK,
* Xuezhong Gong
shnanshan@hotmail.com
1
Department of Nephrology, Shanghai Municipal Hospital
of Traditional Chinese Medicine, Shanghai University
of Traditional Chinese Medicine, 274 Zhijiang Middle Road,
Shanghai 200071, China
2
Center for Radiological Research, College of Physician
and Surgeons, Columbia University, 630 West 168th Street,
New York, NY 10032, USA
3
Department of Radiation Oncology, College of Physician
and Surgeons, Columbia University, 630 West 168th Street,
New York, NY 10032, USA
2188 Arch Toxicol (2016) 90:2187–2200
1 3
NF-κB Nuclear factor-κB
PARP Poly ADP-ribose polymerase
PI Propidium iodide
ROS Reactive oxygen species
SB SB203580
SDH Sucinate dehydrogenase
SP SP600125
TMP Tetramethylpyrazine
TUNEL Terminal deoxynucleotidyl transferase-mediated
dUTP-biotin nick end labeling
U0 U0126
ZnPP Zinc protoporphyrin
Introduction
Arsenic (As) is an important environmental contaminant
affecting more than 140 million people worldwide through
contaminated drinking water (Rodriguez-Lado et al. 2013).
Recent epidemiologic studies in a rural US population, in
some regions of China and especially in Bangladesh sug-
gest arsenic is a major risk factor for kidney disease (Chen
et al. 2011; Zheng et al. 2013). Kidney is one of the tar-
geted organs of arsenic cytotoxicity that could cause renal
dysfunction, proteinuria and chronic kidney disease (CKD)
(Yu et al. 2013; Michael 2013; Ruiz-Hernandez et al. 2015;
Zheng et al. 2014; Chen et al. 2014). Our recent study
demonstrated that sodium arsenite at a clinically relevant
dose induced cytotoxicity in human renal proximal tubu-
lar epithelial cell line, HK-2, which served as a repre-
sentative cell model for exposures of the human kidney to
As, drug-induced nephrotoxicity and acute kidney injury
(Huang et al. 2015; Peraza et al. 2003, 2006; Wang et al.
2013b). Furthermore, our recent results indicated that such
a nephrotoxicity was associated with a dramatic increase in
intracellular ROS production, mitochondrial dysfunction,
inflammation, apoptosis and autophagy (Gong et al. 2014).
However, the precise molecular mechanisms responsible
for arsenic nephrotoxicity remain largely unclear.
The inducible heme oxygenase-1 (HO-1) could exhibit
anti-apoptotic, anti-oxidative and anti-inflammatory prop-
erties and thus be a renal protectant in multiple kidney
injuries, such as acute kidney injury (AKI) induced by
ischemia or nephrotoxicity induced by cisplatin and con-
trasting solutions (Chang et al. 2014; Miyagi et al. 2014).
The critical role of HO-1 has also been reported in several
organ injuries (Wang and Dore 2007; Billings et al. 2014).
Given that HO-1 also has been identified as a response
biomarker for arsenic exposure in various types of cells,
we were very interested: (1) What is the role of HO-1 in
As-induced nephrotoxicity (Gong et al. 2014) at clini-
cally relevant doses? (2) What are the intricate molecular
mechanisms involving in the regulation of HO-1 induction
during As nephrotoxicity?
Furthermore, we have previously identified 2,3,5,6-tetra-
methylpyrazine (TMP), a compound extracted from the
Chinese medicinal plant Ligusticum wallichi (Chuanxiong)
as a protective agent against arsenic nephrotoxicity, which
could attenuate ROS production, inflammation and cell
death (Gong et al. 2014). One of the main aims of the cur-
rent study was to further elucidate a potential relationship
between HO-1 production and the renal protection by anti-
oxidant TMP in arsenic nephrotoxicity, which is not well
understood.
Arsenic response protein 2 (ARS2, also known as Srrt)
was first isolated as a gene product conferring resistance
to arsenite and arsenate in Ass/S5 cell line (Rossman and
Wang 1999). Based on the very limited published data,
ARS2 has been shown to be essential for the development
of plants and mammals and also acts as a transcriptional
regulator of Sox2 in neural stem cell (Kiriyama et al. 2009;
Wilson et al. 2008; Andreu-Agullo et al. 2012). However,
the precise biological functions of ARS2 in mammalian are
largely unknown (Wilson et al. 2008; Andreu-Agullo et al.
2012). The previous work from our laboratory has shown
an up-regulation of ARS2 expression in human neural stem
cell after arsenic exposure (Ivanov and Hei 2013), which
suggested that ARS2 might be involved in arsenic-induced
cytotoxicity and supported the previous suggestion that
ARS2 has essential functions (Wilson et al. 2008). How-
ever, the signaling mechanism regulating ARS2 induction
is still unclear, and a role of ARS2 in arsenic nephrotoxicity
has not been reported so far.
In the present study, we have further investigated the
potential relationships between HO-1 induction, TMP-
mediated renal protection and ARS2 expression in the sup-
pression of arsenic nephrotoxicity.
Materials
All chemicals were purchased from Sigma (St. Louis,
MO, USA) unless otherwise stated. NF-κB inhibitor Bay
11-7082 (Bay), MAPK p38 inhibitor SB203580 (SB) and
ERK inhibitor U0126 (U0) were obtained from Calbio-
chem (La Jolla, CA, USA), and JNK inhibitor SP600125
(SP) was obtained from Biomol (Plymouth Meeting, PA,
USA).
Cell culture and treatment
The human proximal tubular cell line HK-2 (Ameri-
can Type Culture Collection, Manassas, VA, USA)
was grown in culture medium (keratinocyte serum-free
2189Arch Toxicol (2016) 90:2187–2200
1 3
medium + 5 ng/ml epidermal growth factor and 50 μg/
ml bovine extract + 100 U/ml penicillin and 100 μg/
ml of streptomycin) at 37 °C and 5 % CO
2
humidified
environment.
The next stock solutions were prepared: 50 mM sodium
arsenite, antioxidant N-acetylcysteine (NAC, 10 mM),
TMP (50 μM, 100 μM) in PBS, NF-κB inhibitor Bay
(5 μM), MAPK p38 inhibitor SB (10 μM), ERK inhibi-
tor U0 (10 μM) and JNK inhibitor SP (10 μM) in DMSO,
and HO-1 inhibitor zinc protoporphyrin (ZnPP, 2 μM) in
methanol. NAC, TMP and other inhibitors were added into
media 30 min before As.
Intracellular ROS detection
Dihydroethidium (DHE, Invitrogen, Eugene, OR) method
to detect intracellular superoxide production was used.
After 24-h As treatment, cells were exposed to 2 μM DHE
45 min at 37 °C in the dark and then washed twice with
PBS. Finally, fluorescence-activated cell sorter (FACS)
analysis was performed (Becton–Dickinson, Franklin
Lakes, NJ) using the CellQuest program. Samples were
analyzed in triplicate, and all experiments were repeated
three times independently.
Analyses of apoptosis by PI staining and FACS assay
and TUNEL staining
Apoptotic cells were identified by diminished DNA content
in the sub-G1 population of normal diploid cells by FACS
assay after staining with propidium iodide (PI) (Gong et al.
2010). Approximately 20,000 counts were made for each
sample. Finally, the data were collected and analyzed with
CellQuest program combined with FACS machine. Apop-
totic levels were calculated by evaluating the percentage of
events accumulated in the sub-G1 position. Samples were
analyzed in triplicate and repeated 3 times.
To further confirm apoptotic cell death, terminal deoxy-
nucleotidyl transferase-mediated dUTP-biotin nick end
labeling (TUNEL) staining was performed using a Click-
iT TUNEL Alexa Fluor 488 Imaging Assay (Invitrogen,
Grand Island, NY) according to the manufacturer’s instruc-
tions except that PI replaced Hoechst 33342 to mount cells
and label all nuclei. Stained nuclei were analyzed by a
Nikon confocal microscope (Nikon TE200-C1) at 24 °C
room temperature. TUNEL-positive cell numbers from 20
different fields (a total of 2000–2500 cells) were counted to
get an average number of cells per field.
Mitochondrial network morphology assay
Mitochondrial network morphology and activity were vis-
ualized by using MitoTracker Green (Molecular Probes,
Invitrogen). In brief, cells were grown on 6-well plate and
incubated with 200 nM MitoTracker Green for another
20 min at 37 °C after 6-h As treatment. After three-time
washes with PBS, cells were fixed with 4 % paraformal-
dehyde. Confocal fluorescence microscopy images were
captured, and mitochondrial network morphology (tubu-
lar and non-tubular) was quantified by Image J software
(NIH). Each treatment was randomly selected 20 non-
contiguous fields for further observation and analysis,
generally, each field containing 20–25 cells with mito-
chondrial networks. Percentages of normal (tubular) and
abnormal (non-tubular) mitochondrial network morpholo-
gies were counted.
Mitochondrial function assay: cytochrome c
oxidase (Cox) and succinate dehydrogenase (SDH)
histochemistry
Cox and SDH histochemistry were monitored as described
previously (Gong et al. 2014). In brief, cells were cul-
tured on glass cover inside a 6-well plate. After 6 h indi-
cated treatment, cells on glass cover were allowed to dry
at room temperature for 1 h and then followed by 15-min
preincubation at room temperature with 1 mM CoCl
2
and
50 μl DMSO in 50 mM Tris–HCl, pH 7.6, containing
10 % sucrose. All samples were rinsed once in PBS and
incubated for another 3 h with incubation medium (10 mg
cytochrome c, 10 mg of DAB hydrochloride, 2 mg of cata-
lase and 25 μl DMSO resolved in 10 ml 0.1 M phosphate
buffer, pH 7.6). After further three-time rinse, all samples
were mounted on warm glycerin gelatin and observed
under Nikon LABOPHOT-2 microscope to capture images
with SPOT Basic ™ software. Quantification of histochem-
ical staining was performed with Image J software (NIH).
Camera light settings were standardized, and color images
were captured with 40× objective.
Western blotting
After the various treatments, whole cell lysates were pre-
pared by incubation in RIPA buffer (Invitrogen). For
nuclear transcription factor Nrf2 immunoblotting analysis,
nuclear extracts were prepared using methods described
previously (Schreiber et al. 1989). Protein concentrations
were determined with Bio-Rad DC protein assay (Bio-Rad
Laboratories, Calif., USA) using bovine serum albumin
as the standard. The resulting protein samples underwent
SDS-PAGE gel electrophoresis and were transferred to
PVDF membrane.
The specific primary antibodies included the fol-
lowing rabbit Abs: anti-HO-1 (Enzo Life Sciences),
anti-Bcl-xl (Cell Signaling), anti-Bax (Cell Signaling),
anti-PARP (Cell Signaling), anti-pro-caspase-9 (Cell
2190 Arch Toxicol (2016) 90:2187–2200
1 3
Signaling), anti-Nrf2 (Cell Signaling), anti-JNK (Cell
Signaling), anti-phospho-JNK (Cell Signaling), anti-
ERK (Cell Signaling), anti-phospho-ERK (Cell Signal-
ing), anti-p38 MAPK (Cell Signaling), anti-phospho-p38
MAPK (Cell Signaling), anti-ARS2 (Santa Cruz), anti-
histone H3 (Cell signaling) and mouse anti-beta actin
(Sigma).
Statistical analysis
The data were presented as mean ± SD for a minimum
of three independent experiments. All comparisons were
made using either one-way ANOVA or a two-tailed t test
analysis depending on how many conditions were com-
pared in each experiment. One-way ANOVA was followed
by Tukey’s post hoc test. A value of p < 0.05 was consid-
ered significant.
Results
2,3,5,6‑Tetramethylpyrazine (TMP) inhibited
arsenic‑induced ROS‑dependent HO‑1 expression
Arsenic (As) has been identified as inducer of heme oxy-
genase-1 (HO-1) expression in many cells and tissues
(Teng et al. 2013; Li et al. 2013). Consistent with these
reports, we demonstrated that the induction of HO-1 pro-
tein expression was dose and time dependent in HK-2 cells
after As exposure. Since the As dose range of 2.0–10 μM
is successfully applied for treating acute promyelocytic
leukemia (APL) and multiple myelomas (Ivanov and Hei
2004, 2005; Shen et al. 1997), we therefore choose this
dose range for our present study. Furthermore, antioxidants
N-acetylcysteine (NAC) and TMP significantly inhibited
arsenic-induced HO-1 expression (Fig. 1a, b). In contrast,
Fig. 1 TMP inhibited arsenic-induced ROS-dependent HO-1 expres-
sion in HK-2 cells. a, b HK-2 cells were exposed to sodium arsenite
at indicated doses for 3, 6 and 24 h alone or in a combination with
either 10 mM NAC or 50–100 μM TMP. Arsenite (As) exposure
induced HO-1 protein expression in a dose- and time-dependent man-
ner, whereas NAC and TMP pretreatment sufficiently blocked HO-1
up-regulation. c, d Compared with 10 mM NAC, 10 μM AG, the
inhibitor of iNOS, failed in preventing HO-1 up-regulation. β-actin
was used as loading control. Quantitative densitometry of protein
bands was performed. Values are mean ± SD (n = 3), (**) p < 0.01
versus control
2191Arch Toxicol (2016) 90:2187–2200
1 3
aminoguanidine (AG), the iNOS inhibitor, failed in pre-
venting HO-1 up-regulation (Fig. 1c, d).
Blocking HO‑1 with ZnPP augmented As‑induced
apoptosis and ROS production
To verify the role of HO-1 in arsenic (As) nephrotoxic-
ity, zinc protoporphyrin (ZnPP), a known inhibitor of
HO-1 enzymatic activity, was used. Interestingly, ZnPP
aggravated sodium arsenite (10 μM)-induced cytotoxicity
and apoptosis that was confirmed by increased percent-
age of cells accumulated in the sub-G1 position (35.7 vs.
12.5 %) determined by FACS analysis of PI-stained cell
nuclei (Fig. 2a, b) and by % TUNEL-positive staining cells
(Fig. 2c, d).
Several studies have shown that the induction of HO-1
by As is ROS dependent (Fan et al. 2010; Teng et al. 2013).
Our previous data have demonstrated that As exposure
elevated ROS production in HK-2 cells (Gong et al. 2014).
The present results demonstrate that ROS production is up-
regulated by 2.19-fold with 10 μM sodium arsenite treat-
ment. Furthermore, combined treatments with ZnPP and
10 μM sodium arsenite resulted in a 5.35-fold increase in
ROS production highlighting antioxidant role of HO-1 acti-
vation (Fig. 2e, f).
Fig. 2 Blocking HO-1 activity with ZnPP augmented arsenic-
induced apoptosis and ROS production. Apoptosis levels and intra-
cellular ROS production were measured 24 h after arsenic (As) treat-
ment with or without 2 μM ZnPP pretreatment of HK-2 cells. a, b
Levels of apoptosis were calculated by evaluating the percentage of
cells accumulated in the sub-G1 position after PI staining DNA. c,
d Apoptosis was determined by TUNEL staining, and average per-
centages of TUNEL-positive cells were assessed in each group. e, f
100 μM TMP and 10 mM NAC inhibited As-induced ROS genera-
tion at 24 h, while 2 μM ZnPP augmented ROS generation. Values
are mean ± SD (n = 3), (*) p < 0.05 versus Con; (**) p < 0.01 versus
Con; (
##
) p < 0.01 versus 10 μM As