2017 Volume 40 Issue 1 Pages 82-87
The human breast cancer resistance protein (BCRP/ABCG2), a member of the ATP-binding cassette transporter family, is a drug transporter restricting absorption and enhancing excretion of many compounds including anticancer drugs. The cis-regulatory elements in the BCRP promoter include a hypoxia response element, i.e., the DNA binding site for hypoxia-inducible factor-1 (HIF-1). In this study, we investigated the effect of cobalt chloride, a chemical inducer of HIF-1α, on the expression and function of BCRP in human renal proximal tubular cell line HK-2. Cobalt chloride treatment significantly increased the mRNA expression of not only glucose transporter 1 (GLUT1), a typical HIF-1 target gene mRNA, but also ABCG2 mRNA in HK-2 cells. The BCRP inhibitor Ko143-sensitive accumulation of BCRP substrates such as Hoechst33342 and mitoxantrone was significantly enhanced by cobalt chloride treatment. In addition, treatment with cobalt chloride significantly increased the Ko143-sensitive accumulation of fluorescein isothiocyanate-labeled methotrexate in HK-2 cells. Furthermore, cobalt chloride treatment attenuated the cytotoxicity induced by mitoxantrone and methotrexate, which might be, at least in part, due to the increase in BCRP-mediated transport activity via HIF-1 activation. These findings indicate that HIF-1 activation protects renal proximal tubular cells against BCRP substrate-induced cytotoxicity by enhancing the expression and function of BCRP in renal proximal tubular cells.
The human breast cancer resistance protein (BCRP/ABCG2), which was initially found in a multidrug-resistant breast cancer cell line, is a member of the ATP-binding cassette (ABC) transporter family. Like P-glycoprotein, BCRP plays an important role in disposition and distribution of various compounds including xenobiotics and endogenous metabolites. BCRP can transport a wide range of structurally and functionally diverse substrates, including anticancer agents such as mitoxantrone, topotecan and methotrexate, Hoechst 33342, estrone-3-sulfate, urate, and porphyrins such as heme.1–3)
BCRP protein is observed in the plasma membranes of different tissues, including the brush-border membrane of placental syncytiotrophoblasts, the apical membrane of the small and large intestines, the canalicular membrane of hepatocytes, and the luminal membrane of the capillary endothelium of the brain.1,3) BCRP has been also reported to be localized in the apical membrane of renal proximal tubular cells,4) but the expression level of BCRP in the kidneys is relatively lower than that in the liver and small intestine.
The expression of BCRP is reported to be regulated at the transcriptional level. The promoter region of the ABCG2 gene is TAT A-less, contains several SP1 sites and is downstream of a putative CpG island.5) The cis-regulatory elements in the BCRP promoter have been shown to include a hypoxia response element (HRE), an estrogen response element (ERE), a progesterone response element (PRE), an antioxidant response element (ARE), an aryl hydrocarbon response element (AhRE), and a nuclear factor κB (NFκB) response element.1,6) Thus, the ABCG2 gene could be modulated under physiological and pathophysiological conditions.1,3,7)
Hypoxia-inducible factor (HIF), a basic helix–loop–helix transcription factor, is composed of an inducible α-subunit (HIF-1α) and a constitutive β-subunit (HIF-1β).8,9) In normoxia, HIF-1α is hydroxylated by oxygen-dependent HIF prolyl-4-hydroxylases and binds the von Hippel Lindau (VHL) protein, leading to HIF-1α ubiquitination and degradation by the 26S proteasome. In hypoxia, the activity of prolyl hydroxylase domain enzymes is reduced, resulting in the prevention of HIF-1α binding to the VHL protein. The stabilization of HIF-1α protein is followed by nuclear translocation, dimerization with HIF-1β and binding to hypoxia-response elements in the promoter of target genes. Thus, HIF-1α is an essential regulatory factor that facilitates cellular adaptation to hypoxia. In addition to hypoxic activation, the expression of HIF-1α protein is reportedly induced by platelet-derived growth factor, insulin, insulin-like growth factor, angiotensin II, and serum albumin in oxygen-independent manners.10–13)
HIF activation has been shown to be efficient in various kidney disease models such as ischemic or nephrotoxic acute kidney injury and chronic progressive glomerulonephritis.14–20) In contrast, the activation of HIF signaling in renal epithelial cells leads to the development of chronic kidney diseases including renal fibrosis.21–23)
It has been reported that the expression and function of BCRP were enhanced via HIF-1 activation pathway in bone marrow cells and culture tumor cells exposed to hypoxia.6) However, there are few reports on HIF-1-mediated modulation of BCRP expression and function in other tissues including kidney. As described above, the expression of BCRP protein was detected in the apical membrane of renal proximal tubular cells, but it has not been fully clarified yet how HIF-1 activation affects BCRP in renal proximal tubular cells. In this study, we investigated that the effect of cobalt chloride, a HIF-1 activator, on BCRP expression and function in human renal proximal tubular cells. To the best of our knowledge, this is the first report showing HIF-1-mediated induction of the expression and function of BCRP in a human renal proximal tubular epithelial cell line.
Cobalt(II) chloride hexahydrate and Hoechst33342 were obtained from Nacalai Tesque (Kyoto, Japan). Cell proliferation reagent WST-1 was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Mitoxantrone dihydrochloride was obtained from LKT laboratories, Inc. (St. Paul, MN, U.S.A.). Fluorescein methotrexate (FITC-MTX) was obtained from Setareh Biotech LLC (Eugene, OR, U.S.A.). Methotrexate hydrate was purchased from Tokyo Chemical Industry, Co., Ltd. (Tokyo, Japan). All other chemicals used in the experiments were commercial products of the highest purity available.
Human renal proximal tubular cell line HK-2 cells were purchased from the American Type Culture Collection (Manassas, VA, U.S.A.). The cells were cultured in a 1 : 1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F-12 (Nacalai Tesque, Kyoto, Japan), containing 10% fetal bovine serum (FBS), under an atmosphere of 5% CO2–95% air at 37°C, and were subcultured every 7 d. The medium was replaced with fresh medium every 2 or 3 d.
HK-2 cells were cultured in medium containing 10% FBS. At 3 d after seeding, the medium was changed to serum-free medium, and the cells were maintained for a further 24 h. Then, the cells were treated with 300 µM cobalt chloride for 48 h. Cobalt chloride was dissolved in serum-free medium.
Western blot analysis was performed as described previously.12) The primary antibody used in this study was as follows: monoclonal anti-HIF-1 alpha antibody produced in mouse (clone ESEE122) (Sigma-Aldrich) (1 : 1000), anti-BCRP/ABCG2 antibody (BXP-21) (Abcam, Cambridge, MA, U.S.A.) (1 : 100), monoclonal anti-β-actin antibody produced in mouse (clone AC-74) (Sigma-Aldrich) (1 : 5000). The corresponding secondary antibody was anti-mouse immunoglobulin G (IgG) linked with horseradish peroxidase (Kirkegaard & Perry Laboratories Inc., Gaithersburg, MD, U.S.A.) (1 : 5000).
Total RNA was extracted using an RNeasy®Plus Mini Kit (QIAGEN GmbH, Hilden, Germany). Real-time PCR was performed using THUNDERBIRD SYBR qPCR Mix (TOYOBO Co., Ltd., Osaka, Japan). The specific primers for glucose transporter 1 (GLUT1), ABCG2, and β-actin genes were designed as described previously.13) The PCR conditions consisted of initial denaturation at 95°C for 30 s, followed by amplification for 45 cycles of 10 s at 95°C, 15 s at 60°C, and 15 s at 72°C using a Thermal Cycler Dice® Real Time System Lite TP700 (TaKaRa Bio Inc., Shiga, Japan). The threshold cycle (Ct) value for each mRNA was determined using the second derivative maximum method.
After HK-2 cells were treated with cobalt chloride as described above, each well was washed and preincubated with phosphate-buffered saline (PBS) (in mM, 137 NaCl, 3 KCl, 8 K2HPO4, 1.5 KH2PO4, 1 CaCl2 and 0.5 MgCl2) containing 5 mM D-glucose [PBS(G)]. Then, the accumulation of Hoechst33342 (10 µM), mitoxantrone (20 µM), and FITC-MTX (10 µM) was determined. Briefly, PBS(G) buffer containing Hoechst33342, mitoxantrone or FITC-MTX with or without Ko143 (1 µM), a specific inhibitor of BCRP, was added to each well, and the cells were incubated at 37°C for 120 min. At the end of the incubation, the uptake buffer was aspirated off and the wells were rinsed rapidly three times with ice-cold PBS buffer. To each well 0.1% Triton X-100 in PBS buffer without CaCl2 or MgCl2 [PBS(−)] was added, and the cells were scraped off with a cell scraper. The wells were rinsed again to improve the recovery of the cells. The cells were solubilized in 0.1% Triton X-100 in PBS(−) buffer for 30 min at room temperature and then centrifuged at 10000 rpm for 5 min. The supernatant was used for fluorescence and protein assays. The fluorescence was measured using an EnSpire® Multimode Plate Reader (PerkinElmer, Inc., Waltham, MA, U.S.A.) at excitation/emission wavelengths of 360/450 nm (for Hoechst33342), 614/689 nm (for mitoxantrone), and 500/520 nm (for FITC-MTX). In order to estimate BCRP-mediated transport, the accumulation of a fluorescent substrate in the absence of Ko143 was subtracted from that in the presence of Ko143. Protein contents were determined by the Bradford method with bovine serum albumin as a standard.24) The accumulation of the fluorescent substrate was normalized as to the protein content of the cells in each well.
HK-2 cells were treated with cobalt chloride as described above. After the treatment, the cells were further treated with mitoxantrone or methotrexate for 24 h. Then the cells were washed twice, 0.5 mL of a buffer including a WST-1 solution was added to each well, and then the cells were incubated at 37°C for 60 min. After transfer of the reaction solution to a microplate well, the quantity of the formed formazan dye was measured using an EnSpire® Multimode Plate Reader at a wavelength of 440 nm. The reference absorbance (nonspecific readings) was measured at a wavelength of 600 nm.
The half-maximal inhibitory concentration (IC50) value was determined by means of the Hill equation using the KaleidaGraph program (Version 4.5, Synergy Software, PA, U.S.A.) for curve-fitting. Statistically significant differences were determined by Student’s t-test, or one way or two way ANOVA with the Tukey’s honestly significant difference (HSD) test for post hoc analysis. A p value of less than 0.05 was considered statistically significant.
Cobalt chloride is thought to stabilize HIF-1α by decreasing the activity of prolyl hydroxylase domain enzymes (PHDs), which comprise a family of enzymes that play a key role in oxygen-dependent degradation of HIF-1α.25) First, we examined the changes in HIF-1α protein expression in HK-2 cells treated with 300 µM cobalt chloride for 48 h. Cobalt chloride treatment induced the expression of HIF-1α protein in HK-2 cells (Fig. 1A). We next examined whether or not the increase in HIF-1α protein expression leads to one in a HIF-1 target gene mRNA. A typical HIF-1 target gene GLUT1 mRNA26) was markedly enhanced in cobalt chloride-treated HK-2 cells, indicating that transcription factor HIF-1 is activated by cobalt chloride treatment (Fig. 1B). Furthermore, ABCG2 mRNA, another HIF-1 target gene, was significantly increased by the treatment with cobalt chloride (Fig. 1C). Western blot analysis revealed that cobalt chloride enhanced the protein expression of BCRP but not that of β-actin as a loading control (Fig. 1A).
(A) Western blot analysis of HIF-1α, BCRP and β-actin proteins was performed with cell lysates derived from HK-2 cells treated without (control) or with 300 µM cobalt chloride for 48 h. The mRNA levels of GLUT1 (B) and ABCG2 (C) in HK-2 cells treated without or with 300 µM cobalt chloride for 48 h were measured by real-time PCR analysis. The relative level of each target gene mRNA was determined after normalization as to β-actin mRNA. The mRNA levels without cobalt chloride were set to 100%. Values are expressed as the mean±S.E. for three monolayers. * p<0.05, significantly different compared with the value for the cells without cobalt chloride.
To examine the effect of HIF-1 activation on the function of BCRP, an uptake study was performed using Hoechst33342, mitoxantrone and FITC-MTX as BCRP substrates. As shown in Fig. 2A, Hoechst33342 accumulation in HK-2 cells treated with cobalt chloride was more enhanced by Ko143, a specific inhibitor of BCRP, than without cobalt chloride. The Ko143-sensitive transport of Hoechst33342, which was calculated by subtracting the accumulation of Hoechst33342 without Ko143 from that with Ko143, was significantly increased in HK-2 cells treated with cobalt chloride as comparted to ones not treated with cobalt chloride (control) (Fig. 2B). In addition, the increase in mitoxantrone accumulation induced by Ko143 was more clearly observed in the cobalt chloride-treated cells than the untreated cells (Fig. 3A). Thus, the Ko143-sensitive transport of mitoxantrone was significantly greater in the cells with cobalt chloride than in those without cobalt chloride (Fig. 3B). Furthermore, the enhancing effect of Ko143 on FITC-MTX accumulation was observed in the cells treated with cobalt chloride, but not in those without cobalt chloride (Fig. 4A). There was a clear increase in the Ko143-sensitive transport of FITC-MTX in cobalt chloride-treated cells compared to untreated cells (Fig. 4B).
(A) Cells were treated without (control) or with 300 µM cobalt chloride for 48 h, and then the accumulation of Hoecsht33342 in 120 min at 37°C was measured in the buffer in the absence (open column) or presence (gray column) of 1 µM Ko143. (B) Ko143-sensitive transport of Hoechst33342 was evaluated by subtracting the accumulation of Hoecsht33342 in the absence of Ko143 from that in the presence of Ko143 in control and cobalt chloride-treated cells. Each symbol represents the mean±S.E. for four to six monolayers. * p<0.05, significantly different from the value for w/o Ko143 (A) or control (B).
(A) Cells were treated without (control) or with 300 µM cobalt chloride for 48 h, and then the accumulation of mitoxantrone in 120 min at 37°C was measured in the buffer in the absence (open column) or presence (gray column) of 1 µM Ko143. (B) Ko143-sensitive transport of mitoxantrone was evaluated by subtracting the accumulation of mitoxantrone in the absence of Ko143 from that in the presence of Ko143 in control and cobalt chloride-treated cells. Each symbol represents the mean±S.E. for six monolayers. * p<0.05, significantly different from the value for w/o Ko143 (A) or control (B).
(A) Cells were treated without (control) or with 300 µM cobalt chloride for 48 h, and then the accumulation of FITC-MTX in 120 min at 37°C was measured in the buffer in the absence (open column) or presence (gray column) of 1 µM Ko143. (B) Ko143-sensitive transport of FITC-MTX was evaluated by subtracting the accumulation of FITC-MTX in the absence of Ko143 from that in the presence of Ko143 in control and cobalt chloride-treated cells. Each symbol represents the mean±S.E. for four monolayers. * p<0.05, significantly different from the value for w/o Ko143 (A) or control (B).
As described above, treatment with cobalt chloride significantly enhanced the BCRP-mediated efflux of cytotoxic drugs including mitoxantrone and methotrexate, resulting in a decrease in the cellular accumulation of these agents. Next, we examined whether or not the enhanced function of BCRP induced by cobalt chloride protects against the cytotoxic drug-induced cytotoxicity. Figure 5 shows the changes in cell viability with mitoxantrone at various concentrations in HK-2 cells treated without or with cobalt chloride. Mitoxantrone decreased the cell viability in a concentration-dependent manner, and cobalt chloride treatment significantly attenuated the cytotoxic effect of mitoxantrone. The apparent IC50 values of mitoxantrone in cobalt chloride-untreated and treated cells were 2.5 and 37.2 µM, respectively. Furthermore, methotrexate reduced the cell viability with increasing concentration, and cobalt chloride treatment dramatically weakened the cytotoxic effect of methotrexate (Fig. 6). The apparent IC50 values of methotrexate in cobalt chloride-untreated and treated cells were 205.3 and 610.2 µM, respectively.
Cells were treated without or with 300 µM cobalt chloride for 48 h, and then the cells were further treated without (control) or with mitoxantrone at various concentrations for 24 h. Then, cell viability was determined by means of WST-1. Each symbol represents the mean±S.E. for three monolayers. * p<0.05, significantly different from the value for each control. †p<0.05, significantly different from the value for cobalt chloride-untreated cells treated with mitoxantrone at the same concentration.
Cells were treated without or with 300 µM cobalt chloride for 48 h, and then the cells were further treated without (control) or with methotrexate at various concentrations for 24 h. Then, cell viability was determined by means of WST-1. Each symbol represents the mean±S.E. for three monolayers. * p<0.05, significantly different from the value for each control. † p<0.05, significantly different from the value for cobalt chloride-untreated cells treated with methotrexate at the same concentration.
An earlier study showed that the expression of BCRP protein was undetectable in normal human kidneys.27) In subsequent study, it was demonstrated that BCRP is expressed in the apical membrane of human renal proximal tubules using the BXP-9 antibody.4) However, the expression level of BCRP in normal human kidneys is relatively low compared to in the placenta, small intestine, liver and colon. Since the expression of BCRP is highly regulated by external factors including hypoxic conditions, it is suggested that BCRP may play a protective role under pathological conditions. The expression of Bcrp protein in mouse kidneys is reportedly upregulated after ischemic reperfusion injury though the Abcg2 mRNA level is downregulated.28) In addition, serum albumin overload increased ABCG2 mRNA levels in HK-2 cells.12,13) These findings might indicate the important role of BCRP in a variety of disease conditions.
In this study, we employed HK-2 cells as an in vitro model of human renal proximal tubular cells. On real-time PCR analysis, the Ct (threshold cycle) values for GLUT1 and ABCG2 mRNAs from HK-2 cells were 21.7±0.1 and 31.7±0.2, respectively (n=3, the mean±standard error (S.E.)), indicating that the mRNA level of ABCG2 in HK-2 cells is much lower than that of GLUT1. In addition, the Ct value of ABCG2 mRNA from Caco-2 cells, which are often used as an in vitro model of human intestinal epithelial cells, was 27.4±0.3 (n=3, the mean±S.E.), showing that ABCG2 mRNA is more abundant in Caco-2 cells than HK-2 cells. This relationship of the ABCG2 mRNA expression level between Caco-2 cells and HK-2 cells is relatively well correlated with that between the intestines and kidneys. Thus, it is likely that the expression level of ABCG2 mRNA in HK-2 cells possibly reflects that in normal human kidneys.
In association with the low level of ABCG2 mRNA in HK-2 cells, BCRP-mediated transport activity, which was evaluated as BCRP inhibitor Ko143-sensitive transport of BCRP substrate, was almost not detected in the untreated HK-2 cells. In contrast, hypoxia mimetic cobalt chloride treatment induced a significant increase in the BCRP-mediated transport activity of Hoechst33342, mitoxantrone and FITC-MTX. It is likely that this enhanced BCRP transport activity is due to the induction of BCRP protein expression via the HIF-1 pathway since the treatment with cobalt chloride significantly increased the expression of not only HIF-1α protein but also ABCG2 mRNA. An HIF-mediated increase in ABCG2 mRNA expression has also been reported in HK-2 cells, which were treated with fatty acid-bearing serum albumin.13)
BCRP substrates are reported to include a wide range of cancer chemotherapeutic agents as well as physiological compounds.1,3) The anticancer drugs among BCRP substrates include mitoxantrone, methotrexate, camptothecin derivatives, and tyrosine kinase inhibitors such as imatinib and gefitinib. Since BCRP is a transporter that pumps substrates out from cells, the overexpression of BCRP is suggested to be responsible for the multidrug resistance of cancer cells by decreasing the cellular accumulation of the anticancer drugs.2,3) In this study, cobalt chloride treatment clearly attenuated the cytotoxic effects of mitoxantrone and methotrexate, BCRP substrates, in HK-2 cells. The observed resistance to the two anticancer drugs in cobalt chloride-treated cells is, at least in part, due to the decrease in cellular accumulation of the cytotoxic drugs, following the increase in BCRP-mediated transport induced by cobalt chloride treatment.
On the other hand, cobalt chloride increased the cellular accumulation of BCRP substrates in contrast with the reduced cytotoxicity of the BCRP substrate anticancer drugs. The apparent discrepancy is not clear at present. It might result from an increase in the nonspecific binding of the BCRP substrates to the plasma membrane rather than that in intracellular accumulation because the binding of cobalt, a multivalent cation, to the cell membrane might decrease the cell surface charge, leading to an increase in adsorption of the substrates to the cell membrane. Alternatively, the differences in incubation time with BCRP substrates between the two experiments might be related to the apparent inconsistency. Briefly, the accumulation was evaluated in the cells that was incubated with BCRP substrates for 120 min, at which the accumulation might not reach the steady-state conditions, while the cytotoxicity was assayed after 24 h incubation with BCRP substrates. Further studies are needed to explain the observed inconsistency in this study.
In addition to cancer chemotherapeutic agents, BCRP also transports endogenous compounds including heme/porphyrins, folic acid and uric acid.1,7,29) Furthermore, conjugated organic anions such as estrone-3-sulfate and phosphorylated nucleotides are effluxed out by BCRP from cells.1,30) Induction of BCRP transport activity via HIF-1 activation under disease conditions may play an important role in control of the intracellular and extracellular milieu of renal proximal tubular cells.
In conclusion, a HIF-1 inducer, cobalt chloride, induced the expression and function of BCRP in HK-2 cells. This is the first report showing the induction of BCRP-mediated transport via HIF-1 activation in renal proximal tubular cells.
This work was supported in part by JSPS KAKENHI Grant Number 26460218, the Takeda Science Foundation and the Nakatomi Foundation.
The authors declare no conflict of interest.