Hepatocellular carcinoma (HCC) is the main type of liver cancer, accounting for more than half of all liver cancer cases [1, 2]. Currently, the therapeutic options for HCC patients are poor and mainly include surgical resection, liver transplantation, and several antitumour drugs [3, 4, 5]. However, the therapeutic effects are extremely limited.
The matrix metalloproteinases (MMPs), consisting of a series of subtypes in humans, are zinc-dependent endopeptidases that are involved in the digestion of the extracellular matrix (ECM) during cell invasion or migration [6, 7]. To date, more than 22 human MMPs and 25 vertebrate homologues have been identified . MMPs are produced as inactive zymogens and are maintained in the inactive form by the interaction between a cysteine in the propeptide domain and a zinc ion in the catalytic domain . Activated MMPs have been demonstrated to be associated with various cell physiological processes, including cell apoptosis, proliferation, invasion, and migration [9, 10]. Therefore, MMP activation is considered to play a critical role in the progression of human cancers. Recently, MMP2 was reported to participate in the tumourigenesis of multiple tumours, such as thyroid cancer, lung cancer, and ovarian cancer [11, 12, 13]. However, whether MMP2 is involved in the tumourigenesis of HCC remains largely unclear.
This study aimed to evaluate the underlying functions of MMP2 in HCC cells and to assess the effects of MMP2 on the sensitivity of HCC cells to cisplatin. We first synthesized an effective siRNA against MMP2 (si-MMP2), and two HCC cell lines (HepG2 and Huh7) were transduced with si-MMP2-carrying lentivirus to block MMP2 expression. In subsequent functional assays, we found that the knockdown of MMP2 significantly inhibited cell proliferation and invasion and promoted cell apoptosis, suggesting that MMP2 may act as an oncogene in HCC. In addition, we found that the knockdown of MMP2 in HCC cells enhanced the sensitivity of HepG2 and Huh7 cells to cisplatin, indicating that MMP2 inhibitors may serve as enhancers of antitumour drugs.
2 Materials and methods
2.1 Cell lines and culture
The HCC cell lines HepG2 and Huh7 were both purchased from the Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, and maintained in a humidified cell incubator containing 5% CO2 and 95% air at 37°C with RPMI 1640 medium supplemented with 100 U/ml penicillin, 100 U/ml streptomycin and 10% foetal bovine serum (FBS).
2.2 Experiments on optimal transduction conditions
The optimal transduction concentrations and times were determined using western blot assays. Three parallel experiments were used in each treatment.
Gene silencing was performed using a lentivirus-mediated method . The negative control siRNA (non-silencing siRNA with a scrambled sequence; si-con) lenti-virus and the MMP2-siRNA (si-MMP2; target sequence, AGU AGA UCC AGU AUU CAU UCC CUG C ) lentivirus were purchased from Sangon Biotech, Shanghai, China. Polybrene (Life Technologies; Carlsbad, CA, US) was used as the transduction enhancer. The lentivirus transduction procedures were performed according to the instructions indicated by the manufacturer (Sangon Biotech, China).
2.2.1 Optimal transduction concentration
HepG2 and Huh7 cells were seeded into 24-well plates at a density of 1×105 cells/well. A centrifuge tube (1.5 ml) was used, and serum-free culture medium without antibiotics at 150 μl was added. Afterwards, 15 μl of polybrene was added. The mixtures were applied to the cells to constitute the normal control, polybrene, si-con, 90 multiplicity of infection (MOI; lentivirus-cell ratio), 100 MOI, and 110 MOI groups. After 6 h of transduction, the complex culture medium was replaced with fresh medium. Then, the cells were incubated in 5% CO2 and saturated humidity at 37°C for 72 h.
2.2.2 Optimal transduction time
HepG2 and Huh7 cells were seeded at a density of 1×105 cells/well. Serum-free culture medium without antibiotics at 300 μl was added to a 1.5 ml centrifuge tube. Polybrene (15 μl) was added. si-MMP2 was applied at the optimal concentration. Afterwards, the mixtures were applied to the cells to constitute the normal control, polybrene, 12-h, 24-h, 48-h and 72-h groups according to different transduction times. Other culture conditions were the same as the conditions mentioned above. An si-control group at the same concentration as that of si-MMP2 was also used for the 72-h culture.
2.2.3 Western blot
Total proteins of the treated HCC cells were extracted using RIPA buffer (50 mM Tris, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, and 0.05% SDS, pH 7.4). After centrifugation by a refrigerated centrifuge (Eppendorf, TGL-18R) at 12,000 rpm/min at 4°C for 15 min, the supernatants were collected, and a protein assay reagent (Bio-Rad Laboratory, Hercules, CA, USA) was applied to determine the protein concentration. A total of 50 μg of protein was electrophoresed on a 10% SDS-PAGE gel and transferred onto PVDF membranes. After blocking with 5% non-fat dried milk at room temperature for 2 h, membranes were incubated with primary antibodies against MMP2 (mouse, 1:1,000, Ab86607, Abcam) and actin (mouse, 1:2000, Santa Cruz Biotechnology, USA) at 4°C overnight. The membranes were washed thrice for 15 min for each wash with Tris-buffered saline with Tween 20 (TBS-T; containing 1.65 ml of 20% Tween 20 and 700 ml of TBS). Then, the membranes were incubated with horseradish peroxidase-conjugated donkey anti-mouse IgG (1:3,000; Proteintech; Rosemont, IL, USA) for 45-60 min. After extensive washing, the blots were developed by ECL (Thermo Fisher Scientific Inc.; Waltham, MA, USA). Chemiluminescence was captured on a UVP BioImaging system (UVP; Cambridge, UK).
2.3 MMP2 knockdown experiments
Western blot assays showed that the optimal virus-cell ratio was 100 MOI for both cells and that the optimal transduction time for HepG2 and Huh7 cells was 48 h and 72 h, respectively (Supplementary Figures 1 and 2).
In the following experiments, the obtained optimal transduction conditions were used, with six parallel wells for each group.
2.3.1 Validation of knockdown efficiency
HepG2 and Huh7 cells were seeded into 24-well plates at a concentration of 1×105 cells/well and incubated at 37°C overnight. Transduction was performed according to the optimal transduction conditions obtained in this study. The knockdown efficiency was validated by real-time quantitative polymerase chain reaction (RT-qPCR) and western blot assays .
Total RNAs of treated HepG2 and Huh7 cells were prepared using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the protocol from the manufacturer. cDNA was synthesized using a Revert Aid First Strand cDNA Synthesis Kit (Thermo; USA) based on the total RNAs extracted from treated HCC cells. qRT-PCR was conducted on a 7900 HT Fast system (Applied Biosystems, California, USA) with the following protocol: 94°C, 5 min; 40 cycles of amplification (94°C, 30 s and 62°C, 40 s); and then 72°C, 10 min. The sequences of the MMP2 primers were 5’-AGACAGTGGATGATGCCTTTGC-3’ (forward) and 5’-GGAGTC
CGTCCTTACCGTCAAA-3’ (reverse), and those of actin were 5’-CATCCTGCGTCTGGACCTGG-3’ (forward) and 5’-TAATGTCACGCACGATTTCC-3’ (reverse). The results were analysed using the 2-ΔΔCt method , where ΔCt =ΔCt of the target or that of the reference (actin), and -ΔΔCt= average ΔCt of the control or that of the target. The 2-ΔΔCt method reflects the relative expression of the target gene to that of the control.
2.3.3 Western blot
The procedures were performed as described above.
2.3.4 Cell invasion assay (Transwell assay)
The cell invasive ability of treated HCC cells was determined using Transwell chambers (8-μm pore size; BD Biosciences, San Jose, CA, USA) with Matrigel coating (BD Biosciences). The upper chamber was filled with 200 μl of treated HepG2 and Huh7 cells (1×105) in serum-free medium, and 500 μl of medium containing 5% FBS was added into the lower chamber. After incubating at 37°C for 24 h, the non-invasive cells were removed, and the invasive cells were fixed and stained using crystal violet. Invasive cell numbers were counted under a microscope (×100). Five visual fields were randomly selected, and the average was obtained.
2.3.5 MTT assays
Cell viability was assessed by MTT assays following the procedures described in the literature . In this study, HepG2 and Huh7 cells were infected according to the optimal transduction conditions obtained in this study. The optimal virus:cell ratios for both cell lines were 100 MOI, and the optimal transduction times for HepG2 and Huh7 cells were 48 and 72 h, respectively. Then, the transduced cells were transferred to a 96-well plate for MTT assays at a density of 2×104 cells/well. After washing three times with phosphate buffered saline (PBS), cells were incubated with 20 μl of MTT solution at 5 mg/ml for 4 h at 37°C in a humidified CO2 incubator. The media were removed, and then 150 μl dimethylsulphoxide (DMSO) was added into each well and incubated at room temperature for 10 min. Finally, the absorbance was measured at 490 nm with a microplate reader (Sunrise™; Tecan Group Ltd., Switzerland).
2.3.6 Apoptosis analysis
The cell apoptosis rate of HCC cells infected with si-MMP2 lentivirus was determined by Annexin-V/PI double staining followed by flow cytometry analysis . Briefly, treated HepG2 and Huh7 cells were seeded in 24-well plates at a concentration of 2×104 cells/well and maintained overnight at 37°C. After washing with cold PBS, HepG2 and Huh7 cells were resuspended in 500 μl binding buffer containing 5 μl PI and 5 μl FITC-labelled Annexin-V and incubated for 10 min in the dark. Subsequently, apoptotic HepG2 and Huh7 cells were analysed with a FACScan flow cytometer and Kaluza analysis software (Beckman, Fullerton, CA, USA).
2.3.7 Cisplatin cytotoxicity assays
HepG2 and Huh7 cells were infected according to the optimal transduction conditions and were seeded in a 96-well plate at a density of 2×104 cells/well. Cisplatin was added into each well, and the final concentration of cisplatin was 50 μg/ml. The incubation with cisplatin was conducted following a procedure described in the literature . At 24 h after incubation, MTT assays were used to examine the cell viability of HCC cells, and the procedures were performed as described previously.
2.4 Statistical analysis
All the data are expressed as the mean ± standard deviation and processed with SPSS 24.0 software (SPSS, Inc., Chicago, USA). The normality of the data distribution was tested using the Kolmogorov-Smirnov method, and the homogeneity of variance was determined using Levene’s test. One-factor analysis of variance (ANOVA) followed by a post hoc least significant difference (LSD) test was performed for comparisons between groups. P<0.05 and P<0.01 were considered significant and very significant, respectively.
3.1 Knockdown efficiency of si-MMP2
The knockdown efficiency of si-MMP2 was evaluated using RT-qPCR and western blot assays. No significant differences in the mRNA expression of MMP2 were observed among the control, polybrene, and si-con groups in both cell lines (P>0.05); however, compared with that in any other group in the same cell line, the MMP2 mRNA expression in the si-MMP2 group was significantly downregulated, indicating a significantly increased MMP2 knockdown efficiency (P<0.01; Figure 1A and 1B). Western blot assays showed consistent outcomes with those based on RT-qPCR experiments (Figure 1C and 1D).
3.2 MMP2 knockdown decreased cellular proliferation and increased cellular apoptosis
To investigate the biological functions of MMP2 in HCC, we blocked MMP2 expression in HepG2 and Huh7 cells by si-MMP2, and cell proliferation and apoptosis were then assessed using MTT assays and flow cytometry analysis, respectively. Optical density (OD) values from the MTT assays indicate that, compared with that of the normal control cells, the proliferation of HepG2 and Huh7 cells infected with si-MMP2 lentivirus was significantly inhibited; however, the proliferation of cells from the remaining three groups did not significantly differ from that of the normal control cells (Figure 2).
Annexin V/PI double staining and flow cytometry analysis were performed to further evaluate the apoptosis of HepG2 and Huh7 cells after MMP2 knockdown (Figure 3). In both cell lines, no significant differences in the apoptotic rate were observed among the control, polybrene, and si-con groups in both cell lines (P>0.05). However, compared with that in any other group from the same cell line, apoptosis in the si-MMP2 group significantly increased (P<0.01).
3.3 MMP2 knockdown inhibited the cell invasive capacity of HCC cells
To further determine whether MMP2 was involved in the invasion of HCC cells, Transwell assays were performed (Table 1; Supplementary Figure 3). The invasive cell counts of the control, polybrene, and si-con groups were comparable to that reported in the literature . In both cell lines, no significant differences in the number of invasive cells were observed among the NC, polybrene, and si-con groups (P>0.05). However, compared with that in any other group from the same cell line, the number of invasive cells in the si-MMP2 group significantly decreased (P<0.01).
3.4 MMP2 knockdown increased the sensitivity of HCC cells to cisplatin
HCC cells were treated with si-MMP2 and cisplatin (50 μg/ml), and an MTT assay was then performed to assess cell viability (Figure 4). No significant differences in the OD values were observed among the NC+cisplatin, polybrene+cisplatin, and si-con+cisplatin groups in both cell lines (P>0.05). However, compared with that in any other group from the same cell line, the OD value in the si-MMP2+cisplatin group significantly decreased (P<0.01).
In this study, we demonstrated that knockdown of MMP2 in HepG2 and Huh7 cell lines by infecting cells with si-MMP2 lentivirus strongly inhibited cell proliferation and invasion and significantly promoted cell proliferation.
The process of invasion and migration of tumour cells requires degradation of the ECM and basement membrane, and MMPs that are expressed by almost all tumour cells perform this proteolytic process [21, 22, 23]. According to their exact functions and distributions in cells, MMPs can be divided into multiple subtypes. MMP2 is an important member of the MMP family that was found to be overexpressed in multiple human tumours, including HCC [24, 25]. Liu et al.  demonstrated that downregulation of MMP2 and MMP9 was induced by the main components of pomegranate, ellagic acid and luteolin, and suppressed the metastasis of ovarian cancer. Downregulation of MMP2 and MMP9 induced by TRIM31 knockdown was shown to suppress the proliferation and invasion of gallbladder cancer . In addition, Fang et al.  reported that downregulation of MMP2 induced by microRNA-29b overexpression significantly inhibited tumour angiogenesis, invasion, and metastasis of HCC. Our results were consistent with previously reported results .
Given the complexity of the pathogenesis of HCC, the current therapeutic options for HCC mainly include surgical resection, liver transplantation, and chemotherapy with several antitumour drugs, such as cisplatin, sorafenib, and brivanib [29, 30]. Currently, cisplatin remains the first-line chemotherapeutic agent for HCC . In this study, cisplatin was applied to HCC cells after MMP2 knockdown. The results showed that, compared with those of the other groups, the OD values of the MMP2 knockdown groups were substantially decreased, suggesting that MMP2 knockdown may have a synergistic reaction with cisplatin in inhibiting the proliferation of HCC cells. Recently, an inhibitor of MMP2 was shown to enhance the sensitivity of platinum-resistant ovarian cancer cells to cisplatin . In addition, Li et al. reported that genistein could reinforce the inhibitory effects of cisplatin on HCC by downregulating MMP2 expression . Our results are consistent with those previously reported, suggesting that MMP2 may serve as a target in the treatment of HCC in clinical practice.
This study has limitations. This study was conducted in vitro, and therefore, microenvironmental changes in vivo may alter the results. Furthermore, in this study, two HCC cell lines were employed. To further validate the roles of MMP2 in the proliferation and invasion of HCC cells, more cell lines and in vivo studies should be involved in future studies. In conclusion, MMP2 knockdown inhibits the progression of HepG2 and Huh7 cells and enhances the cytotoxicity of cisplatin to these cells, suggesting that MMP2 may act as a therapeutic target for HCC patients.
This work was supported by the Guizhou Province Science and Technology Fund Project; (No. J (Altekruse, et al.) 2113) and the Joint Fund Project of Guizhou Science and Technology Department (grant number, 2113).
Yen, C.W., Hsu, L.S., Chen, C.W., Lin, W.H., Hepatocellular carcinoma with thoracic metastases presenting as hemothorax: A case report and literature review, Medicine (Baltimore), 2018, 97, e10945 Web of SciencePubMedCrossrefGoogle Scholar
Kamo, N., Kaido, T., Yagi, S., Okajima, H., Uemoto, S., Liver transplantation for intermediate-stage hepatocellular carcinoma, Liver Cancer, 2018, 7, 179-189 CrossrefWeb of SciencePubMedGoogle Scholar
Niu, L.Z., Li, J.L., Zeng, J.Y., Mu, F., Liao, M.T., Yao, F., et al., Combination treatment with comprehensive cryoablation and immunotherapy in metastatic hepatocellular cancer, World J. Gastroenterol., 2013, 19, 3473-3480 Google Scholar
Schlachterman, A., Craft, W.W. Jr., Hilgenfeldt, E., Mitra, A., Cabrera, R., Current and future treatments for hepatocellular carcinoma, World J. Gastroenterol., 2015, 21, 8478-8491 Google Scholar
Lohi, J., Wilson, C.L., Roby, J.D., Parks, W.C., Epilysin, a novel human matrix metalloproteinase (MMP-28) expressed in testis and keratinocytes and in response to injury, J. Biol. Chem., 2001, 276, 10134-10144 CrossrefGoogle Scholar
Herrera, I., Cisneros, J., Maldonado, M., Ramírez, R., Ortiz-Quintero, B., Anso, E., et al., Matrix metallo-proteinase (MMP)-1 induces lung alveolar epithelial cell migration and proliferation, protects from apoptosis, and represses mitochondrial oxygen consumption, J. Biol. Chem., 2013, 288, 25964-25975 CrossrefGoogle Scholar
Kraljevic, D., Vukojevic, K., Karan, D., Rajic, B., Todorovic, J., Miskovic, J., et al., Proliferation, apoptosis and expression of matrix metalloproteinase-9 in human fetal lung, Acta Histochem, 2015, 117, 444-450. Web of SciencePubMedCrossrefGoogle Scholar
Laios, A., Mohamed, B.M., Kelly, L., Flavin, R., Finn, S., McEvoy, L., et al., Pre-treatment of platinum resistant ovarian cancer cells with an MMP-9/MMP-2 inhibitor prior to cisplatin enhances cytotoxicity as determined by high content screening, Int. J. Mol. Sci., 2013, 14, 2085-2103 CrossrefPubMedWeb of ScienceGoogle Scholar
Liu, J., Ping, W., Zu, Y., Sun, W., Correlations of lysyl oxidase with MMP2/MMP9 expression and its prognostic value in non-small cell lung cancer, Int. J. Clin. Exp. Pathol., 2014, 7, 6040-6047 Google Scholar
Maacha, S., Hong, J., von Lersner, A., Zijlstra, A., Belkhiri, A., AXL mediates esophageal adenocarcinoma cell invasion through regulation of extracellular acidification and lysosome trafficking. Neoplasia, 2018, 20, 1008-1022 PubMedWeb of ScienceCrossrefGoogle Scholar
Ni, X., Xia, T., Zhao, Y., Zhou, W., Wu, N., Liu, X., et al., Downregulation of miR-106b induced breast cancer cell invasion and motility in association with overexpression of matrix metalloproteinase 2. Cancer Sci., 2014, 105, 18-25 PubMedCrossrefWeb of ScienceGoogle Scholar
Zhang, C., Wang, G., Wang, J., Ji, Z., Dong, F., Chao, T., Analysis of differential gene expression and novel transcript units of ovine muscle transcriptomes. PLoS One, 2014, 9, e89817 CrossrefPubMedWeb of ScienceGoogle Scholar
Saha, P., Ghosh, I., Datta, K., Increased hyaluronan levels in HABP1/p32/gC1qR overexpressing HepG2 cells inhibit autophagic vacuolation regulating tumor potency. PLoS One, 2014, 9, e103208 PubMedCrossrefWeb of ScienceGoogle Scholar
Khantasup, K., Kopermsub, P., Chaichoun, K., Dharakul, T., Targeted small interfering RNA-immunoliposomes as a promising therapeutic agent against highly pathogenic Avian Influenza A (H5N1) virus infection. Antimicrob Agents Chemother, 2014, 58, 2816-2824 CrossrefPubMedWeb of ScienceGoogle Scholar
Kim, M., Jung, J.Y., Choi, S., Lee, H., Morales, L.D., Koh, J.T., et al., GFRA1 promotes cisplatin-induced chemoresistance in osteosarcoma by inducing autophagy. Autophagy, 2017, 13, 149-168 Web of ScienceCrossrefPubMedGoogle Scholar
Mao, H.H., In vitro experimental study on ant-FAK short hairpin RNA for metastasis of colorectal cancer. Dissertation. Fujian Medical University, 2010 Google Scholar
Chojnacki, M., Zajac, A., Piet, M., The involvement of matrix metalloproteinases in the development and progression of neoplasm diseases, Postepy. Biochem., 2017, 63, 277-286 Google Scholar
Das, K., Prasad, R., Ansari, S.A., Roy, A., Mukherjee, A., Sen, P., Matrix metalloproteinase-2: A key regulator in coagulation proteases mediated human breast cancer progression through autocrine signaling, Biomed. Pharmacother., 2018, 105, 395-406 CrossrefGoogle Scholar
Yadav, L., Puri, N., Rastogi, V., Satpute, P., Ahmad, R., Kaur, G., Matrix metalloproteinases and cancer-roles in threat and therapy, Asian Pac. J. Cancer Prev., 2014, 15, 1085-1091 Google Scholar
Giannelli, G., Bergamini, C., Marinosci, F., Fransvea, E., Quaranta, M., Lupo, L., et al., Clinical role of MMP-2/TIMP-2 imbalance in hepatocellular carcinoma, Int. J. Cancer, 2002, 97, 425-431 CrossrefPubMedGoogle Scholar
Ogasawara, S., Yano, H., Momosaki, S., Nishida, N., Takemoto, Y., Kojiro, S., et al., Expression of matrix metalloproteinases (MMPs) in cultured hepatocellular carcinoma (HCC) cells and surgically resected HCC tissues, Oncol. Rep., 2005, 13, 1043-1048 PubMedGoogle Scholar
Liu, H., Zeng, Z., Wang, S., Li, T., Mastriani, E., Li, Q.H., et al., Main components of pomegranate, ellagic acid and luteolin, inhibit metastasis of ovarian cancer by down-regulating MMP2 and MMP9, Cancer Biol. Ther., 2017, 18, 990-999 PubMedWeb of ScienceCrossrefGoogle Scholar
Li, H., Zhang, Y., Hai, J., Wang, J., Zhao, B., Du, L., et al., Knockdown of TRIM31 suppresses proliferation and invasion of gallbladder cancer cells by down-regulating MMP2/9 through the PI3K/Akt signaling pathway, Biomed. Pharmacother., 2018, 103, 1272-1278 CrossrefPubMedWeb of ScienceGoogle Scholar
Fang, J.H., Zhou, H.C., Zeng, C., Yang, J., Liu, Y., Huang, X., et al., MicroRNA-29b suppresses tumor angiogenesis, invasion, and metastasis by regulating matrix metalloproteinase 2 expression, Hepatology, 2011, 54, 1729-1740 PubMedCrossrefWeb of ScienceGoogle Scholar
Hatzaras, I., Bischof, D.A., Fahy, B., Cosgrove, D., Pawlik, T.M., Treatment options and surveillance strategies after therapy for hepatocellular carcinoma, Ann. Surg. Oncol., 2014, 21, 758-766 CrossrefGoogle Scholar
Raza, A., Sood, G.K., Hepatocellular carcinoma review: current treatment, and evidence-based medicine, World J. Gastroenterol., 2014, 20, 4115-4127 Google Scholar
Dasari, S., Tchounwou, P.B., Cisplatin in cancer therapy: molecular mechanisms of action, Eur. J. Pharmacol., 2014, 740, 364-378 Google Scholar
Nouso, K., Miyahara, K., Uchida, D., Kuwaki, K., Izumi, N., Omata, M., et al., Effect of hepatic arterial infusion chemotherapy of 5-fluorouracil and cisplatin for advanced hepatocellular carcinoma in the Nationwide Survey of Primary Liver Cancer in Japan, Br. J. Cancer, 2013, 109, 1904-1907 CrossrefPubMedWeb of ScienceGoogle Scholar
Chen, P., Hu, M.D., Deng, X.F., Li, B., Genistein reinforces the inhibitory effect of Cisplatin on liver cancer recurrence and metastasis after curative hepatectomy, Asian Pac. J. Cancer Prev., 2013, 14, 759-764 Google Scholar
About the article
Published Online: 2019-05-17
Conflict of interestConflicts of interest: The authors have no commercial or other associations that might pose a conflict of interest.
Citation Information: Open Medicine, Volume 14, Issue 1, Pages 384–391, ISSN (Online) 2391-5463, DOI: https://doi.org/10.1515/med-2019-0039.
© 2019 Jiangwei Liu et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. BY-NC-ND 4.0