The Activity of AQP9 Is Mediated MAPK in Arsenic-Treated Mouse Model of Hepatocellular Carcinoma ()
1. Background
Arsenic trioxide (As2O3, more commonly known as arsenic), recorded in both traditional Chinese and ancient Greek medicine, has been applied to the care of various disorders such as psoriasis, syphilis, and rheumatism for more than 1500 years. In the 1970s, Chinese scholars first applied it to acute promyelocytic leukemia (APL) and achieved a significant curative effect [1] , later to gynecological tumors [2] , liver cancer [3] , lung cancer [4] , and pancreatic cancer [5] . Since then, the anti-tumor effect of arsenic has attracted more and more attention. The mechanistic study demonstrated that arsenic achieves the therapeutic purpose by degrading PML/RARA fusion protein, down-regulating BCL2 gene expression, and inducing apoptosis of leukemia cells [6] . In this way, the simple understanding of “fighting poison with poison” in traditional Chinese medicine has been bridged to modern medical molecular biology. Arsenic-induced apoptosis has been elucidated from the perspective of genetics, making arsenic one of the first cell apoptosis-inducing drugs. However, tumor cells are often desensitized to arsenic along with repeated treatment, this drug resistance phenomenon will hinder its therapeutic effect, there are few studies on the resistance mechanism of arsenic agents so far. Three arsenic-resistant tumor cell lines [7] were established including HepG2/AS, SGC7901/AS, and SMMC7721/AS. The known ABC transporter family members ABCB1, ABCC1, and ABCC2 are found increased in these resistant cell lines. Our previous work also found that AQP9, a membrane transporter, plays an important role in the cellular uptake of arsenic [8] [9] , affecting the accumulation of arsenic in cells.
AQP (Aquaporin) is a member of the major intrinsic protein (MIP) family. Eleven AQP isoforms, AQP0 - AQP10, have been identified in mammals by far [10] . They are basically involved in the transport of intracellular and extracellular fluids as channel proteins. In the previous studies in our laboratory, we found that arsenic can regulate the phosphorylation of the aquaporin AQP9 (aquaglyceroporin9, AQP9), thereby affecting the quantity of arsenic entering cells [8] [9] . However, the mechanism of phosphorylation regulation of AQP9 is still unclear. In the study of arsenic tolerance in yeast [11] , Hog1p (p38 homolog, whose gene sequence is 52.3% similar) can regulate the phosphorylation of AQP9 homologous analog Fps1p, thereby affecting the entry of arsenic into cells.
Previous studies in our laboratory also showed that p38 protein kinase has a certain regulatory effect on AQP9 phosphorylation in normal hepatocyte L-02 [12] ).
Mitogen-activated protein kinases (MAPKs), as a family of signaling pathways, mediate extracellular stimulation to intracellular responses, which regulate cell proliferation and differentiation, apoptosis, and the interaction between cells [13] . It is mainly divided into extracellular-signal-regulated protein kinase (ERK), c-Jun amino-terminal kinase (JNK), and P38 [14] [15] . Studies have shown that MAPK signaling is involved in the regulation of ROS on carcinogenesis, cancer cell apoptosis, and cancer metastasis [16] [17] [18] [19] ; many anticancer drugs also play a therapeutic effect by increasing reactive oxygen species and inducing the corresponding tumor cell apoptosis. Therefore, it is very meaningful to explore the mechanism of the MAPK signaling pathway to control the accumulation of arsenic in cells by regulating AQP9 to achieve an anti-tumor effect or generate drug resistance. Studies had shown that the MKK3/6-p38 signaling cascade is initiated by arsenic trioxide (As2O3) to control the anti-leukemia effect mediated by As2O3 in a negative feedback regulation manner [16] .
A clinic study [17] showed that inhibition of p38 MAPK could enhance the efficacy of the combined use of As2O3 and thalidomide (bortezomib) in the treatment of myeloma, suggesting the p38 MAPK signaling pathway is involved in arsenic-induced apoptosis. In recent years, many studies have confirmed that arsenic in vivo induces apoptosis through the JNK signaling pathway [19] [20] . In addition, it was [18] found that JNK kinase inhibitor can significantly reduce the phosphorylation level of Bcl-2 induced by arsenic, which leads to mitochondrial apoptosis in cervical cancer cells. In short, P38 and JNK play important roles in the ROS-induced apoptosis signaling pathway.
This study will focus on the alteration in the expression and phosphorylation of AQP9 and p38 and JNK in mouse models of hepatocellular carcinoma by arsenic, and further, reveal that AQP9 phosphorylation mediated by MAPK-related signaling pathways is resistant to arsenic antitumor resistance. In addition, the localization and distribution of phosphorylated AQP9 in HepG2 cells will be revealed by laser confocal experiments.
2. Material & Methods
2.1. Animal and Cell Lines
C57 mice were purchased from Shanghai Southern Model Biotechnology Co., Ltd. Human hepatoma cell line HepG2 is cryopreserved in our laboratory. NaAsO2 (Sigma, USA) Cell culture HepG2 cell line was cultured in RPM1640 medium; containing 10% fetal bovine serum, 100 U∙mL−1 penicillin and streptomycin) (GIBCO, USA), placed in a saturated humidity incubator with 5% CO2 volume at 37˚C. Mouse model of hepatocellular carcinoma (HCC) was induced with 5 × 106 H22 cells injected subcutaneously on the lateral side of each right axilla in C57BL/6 mice. When the tumor grows to about 100 mm3, the modeling is successful.
2.2. Western Blotting
C57BL/6 wild-type mice and tumor model mice were selected as a group of three mice, and each group was intraperitoneally injected with NaAsO2 at a concentration of 8 μmol/L at a dose of 10 mL/kg. The non-NaAsO2 group was set as the normal control group. It was injected once a day, and after ten consecutive days, the liver tissues of normal mice and model mice were taken and stored in liquid nitrogen for freezing. After grinding, the cell lysate was made by RIPA method, then subjected to quantification, SDS-PAGE, transferred to nitrocellulose membrane. Blotting were performed with appropriate diluted primary antibodies (mouse anti-human AQP9 antibody, mouse anti-human GAPDH antibody (Santa Cruz company), rabbit anti-human p38, JNK antibody, mouse anti-human p38, JNK phosphorylated antibody and mouse anti-pan-phosphorylated antibody (abcam company), followed by horseradish peroxidase (HRP) labeling secondary antibody, visualized by ECL kit (Amersham Bioscience company). Here GAPDH functions as an internal control. Immunoprecipitation was performed using protein A + G agarose with Anti-AQP9 as described previously.
2.3. Plasmids Construction
The full-length AQP9 cDNA was subcloned into pEGFP-N3 (Invitrogen) to obtain the pEGFP-N3-AQP9 plasmid. In addition, site-directed mutagenesis pEGFP-N3-AQP9 S11A (phosphorylated expression inhibition) and pEGFP-N3-QP9 S11D (phosphorylated constitutive expression) plasmids were constructed accordingly.
2.4. Confocal Immunofluorescence
Confocal immunofluorescence was performed on HepG2 transfected with different constructs of plasmids treated with 8 µmol∙L−1 of NaAsO2 for 12 with a confocal microscope (Carl Zeiss AG, German).
2.5. Statistical Analysis
SPSS13.0 statistical software was used to analyze the data, and one-way analysis of variance (ANOVA) was used for multiple comparisons between means. GraphPad Prism software draws the graphs.)
3. Results
3.1. Arsenic Stimulates AQP9 Phosphorylation in Mice Model of Liver Tumor
The C57BL/6 wild-type mice and tumor mice were intraperitoneally injected with 8 μmol/L of NaAsO2, and the liver tissues of the mice were collected after ten consecutive days, and the unmedicated group was used as the control. Immunoprecipitation and Western blot were used to detect AQP9 expression and phosphorylation status. The results showed that there was no significant difference in the expression of AQP9 among the groups (Figure 1(a) and Figure 1(b)). The phosphorylation of AQP9, however, was significantly increased in the arsenic treatment group and tumor model group (Figure 2(a) and Figure 2(c)).
3.2. Arsenic Stimulates P38 Phosphorylation in Liver Specimen from C57 Tumor Model Mice
Since arsenic affects p38 expression and phosphorylation, this study also detected the changes of arsenic on p38 and its phosphorylation in the liver of C57
![]()
Figure 1. AQP9 expression and phosphorylation in livers of mice treated with NaAsO2. (a) 1 WT mice; 2 WT mice treated with NaAsO2; 3 Hepatoma mouse models; 4 Hepatoma mice treated with NaAsO2. The graph represents the densitometry measurement of the AQP9 and P-AQP9 ((b), (c)). *p < 0.05, **p < 0.01.
![]()
Figure 2. p38 expression and phosphorylation in livers of mouse treated with NaAsO2. (a) 1 WT mice; 2 WT treated with NaAsO2; 3 hepatoma mouse model; 4 hepatoma mice treated with NaAsO2. The graph represents the densitometry measurement of the p38 and p-p38 ((b), (c)). *p < 0.05.
tumor model mice. Western blotting experiments showed that p38 protein expression was not significantly different among the groups (Figure 2(a) and Figure 2(b)). The expression level of P-p38 was significantly increased in the arsenic-treated groups (Figure 2(a) and Figure 2(c)).
3.3. Arsenic Increases JNK Phosphorylation in C57 Tumor Model Mice
The changes of JNK expression and its phosphorylation with arsenic treatment in the liver of C57 tumor model mice were further detected. Western blotting experiments showed that there was no significant difference in JNK protein expression among the groups (Figure 2(a) and Figure 2(b)) just like p38. But p-JNK was significantly increased in the arsenic-treated groups (Figure 2(a) and Figure 2(c)). Compared with the wild type, the tumor model group had increased expression (Figure 3(a) and Figure 3(c)).
3.4. Active AQP9 Is Predominantly Expressed in the Membrane in HepG2 Cells
The constructed pEGFP-N3-AQP9, pEGFP-N3-AQP9 S11A (phosphorylation inhibition), and pEGFP-N3-AQP9 S11D (phosphorylation constitutive expression) were transfected into HepG2 cells and treated with 8 µmol∙L−1 NaAsO2 12 h, the fluorescence of HepG2 cells transfected with the wild-type vector was
![]()
Figure 3. JNK expression and phosphorylation in livers of mouse treated with NaAsO2. (a) 1 WT wild group; 2 WT + NaAsO2 treatment; 3 tumor model group; 4 tumor model treated with NaAsO2 group. The graph represents the densitometry measurement of the JNK and P-JNK ((b), (c)). *p < 0.05.
![]()
Figure 4. AQP9 localization in HepG2 cells. (a) HepG2 cell transfected with pEGFP-N3-AQP9; (b) HepG2 cell transfected with pEGFP-N3-AQP9-S11A; (c) HepG2 cell transfected with pEGFP-N3-AQP9-S11D. All of cells were treated with 8 µmol/L NaAsO2 for 12 h.
mainly distributed in the cell membrane under the laser confocal fluorescence microscope (Figure 4(a)). The fluorescence of HepG2 cells transfected with the pEGFP-N3-AQP9 S11A mutant vector was mainly in the cytoplasm as seen (Figure 4(b)), while the fluorescence of HepG2 cells transfected with pEGFP-N3-AQP9 S11D mutant vector is mainly distributed on the cell membrane, and compared with the cells transfected with the wild-type vector, it shows stronger fluorescence on the membrane (Figure 4(c)).
4. Discussion
In the current study, using the C57BL/6 mouse model of hepatoma, we found that arsenic can activate aquaporin AQP9 by phosphorylation, and translocate predominantly its expression to the cell membrane; and further, arsenic could stimulate the phosphorylation of kinases p38 and JNK.
Previous studies have shown that AQP9 is the channel for arsenic to enter/exit cells, and its phosphorylation is the main way to regulate its entry into cells [8] [9] . The hypothesis that protein kinases p38, JNK, etc. may regulate the phosphorylation of AQP9 is proposed accordingly. Preliminary experiments show that p38 has a certain regulatory effect on arsenic intake [12] . However, there are few related research reports on animal experiments. In one study in a mouse model with type 1 diabetes, arsenic uptake was recorded significantly higher in the liver in comparison with normal literate, in concert with higher expression of AQP9 and Glucose transport 1 (GLUT1) in the liver [21] . This paper focuses on the effect of arsenic on the protein expression and phosphorylation expression level of AQP9 and protein kinase p38, JNK in animal experiments.
The experiments showed that under arsenic stimulation, the expression of AQP9 protein in the liver of normal wild-type mice did not increase significantly, but the phosphorylation level increased significantly; the phosphorylation of AQP9 also increased in the liver tumor model in mice (Figure 1), and their phosphorylation levels were higher than those in wild-type mice, which is in parallel with the previous findings [8] . In addition, the expression of kinase p38 did not change significantly after arsenic treatment, but the phosphorylation level increased (Figure 2), in vitro hepatocyte L-02 model also witnessed the similar findings [8] [12] . The expression and phosphorylation pattern is similar in another protein kinase JNK with increased phosphorylation but no change in protein expression (Figure 3). In short, phosphorylation of AQP9 may be a major way to affect the intake of arsenic, while the phosphorylation of AQP9 may be affected by the phosphorylation of protein kinases p38 and JNK. This requires more experiments to further verify.
In addition, in the laser confocal experiment, we observed that the phosphorylated AQP9 was more localized on the cell membrane of HepG2 cells after being stimulated by arsenic. Due to the common characteristics of organisms seeking advantages and avoiding disadvantages, when they are stimulated by the outside world, they generally choose to fight and block harmful substances or expel them. Therefore, we believe that the increased expression of AQP9 phosphorylation on the cell membrane should be related to the inhibition of arsenic entry into cells or expel it. This implies that phosphorylation of AQP9 is a way for this channel protein to transport arsenic into and out of cells. AQP9 may act as a two-way channel for arsenic into or out of cells, in which its phosphorylation may be the main regulatory mode that determines whether it allows arsenic into or out of cells. This also lays the foundation for the later research work.
Funding
The present study was supported by Natural Science Foundation of Zhejiang Province [grant number LY19H260002], to Cao Yifei.