Tumor Biol. DOI 10.1007/s13277-016-4957-6

ORIGINAL ARTICLE

miR-145 sensitizes gallbladder cancer to cisplatin by regulating multidrug resistance associated protein 1 Ming Zhan 1 & Xiaonan Zhao 1 & Hui Wang 1 & Wei Chen 1 & Sunwang Xu 1 & Wei Wang 1 & Hui Shen 1 & Shuai Huang 1 & Jian Wang 1

Received: 1 December 2015 / Accepted: 2 February 2016 # International Society of Oncology and BioMarkers (ISOBM) 2016

Abstract Gallbladder cancer (GBC) is the most common malignancy in biliary tract with poor prognosis. Due to its high chemoresistance, systemic chemotherapies have had limited success in treating GBC patients. MicroRNAs (miRNAs) are emerging novel regulators of chemoresistance, which modulate the expression of drug resistance-related genes. In this study, we investigated the association between miR-145 expression and cisplatin sensitivity by both in vivo and in vitro analysis. Quantitative PCR (q-PCR) analysis indicated an increased miR-145 expression in GBC tissues. In addition, studies on GBC cell lines suggested an increased cisplatin efficacy with miR-145 overexpression, whereas decreasing miR-145 expression reduced cisplatin sensitivity. Further, we found that miR-145 accelerated MRP1 mRNA degradation by directly targeting its 3′-UTR and therefore caused increased cisplatin toxicity in GBC cells. Moreover, lower miR-145 and higher MRP1 expression levels predicted poor prognosis in GBC patients who received chemotherapy. Collectively, our findings established a rationale for using miR-145 expression as a biomarker to identify cisplatin-resistant GBC patients and propose that treatment strategies increasing the expression of miR-145 could be a new therapeutic approach for GBC patients.

Ming Zhan and Xiaonan Zhao contributed equally to this work. * Jian Wang [email protected]

1

Department of Biliary-Pancreatic Surgery, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, 1630 Dongfang Road, Shanghai 200127, China

Keywords Gallbladder cancer . miR-145 . MRP1 . Chemoresistance . Cisplatin

Introduction Gallbladder cancer (GBC) is the most common malignancy of the biliary tract and accounts for an estimated 40 % of all biliary tract carcinomas [1, 2]. With an overall 5-year survival rate of less than 5 %, GBC is a poorly understood aggressive malignancy with poor prognosis [3]. In contrast to available standard adjuvant therapy for patients with other malignant neoplasms, such as lung cancer and hepatocellular carcinoma, systemic chemotherapies have had limited success in GBC p a t ie n t s , w h i c h i s m a i n ly a t tr i b u te d t o i t s h i g h chemoresistance [4]. To date, only a few potential targets and signaling pathways underlying GBC chemoresistance have been revealed [5, 6]. In order to find potential new targets for improving therapeutic efficacy, a thorough and deeper understanding of the drug resistance mechanisms of GBC is urgently needed. Recent discoveries have found that multiple regulators are involved in increasing cancer cell chemoresistance, including enhancement of DNA repair [7], inefficient cellular drug uptake and accumulation [8], activation of the antioxidant glutathione system for detoxification [9], upregulation of the antiapoptosis pathway [10], and increase of cancer stem cells [11]. Among these, reduced intracellular drug accumulation has been proven to be an important mechanism in the development of chemoresistance. Moreover, overexpression of drug efflux pump-related proteins, such as P-glycoprotein (MDR1), multidrug resistance-related protein 1 (MRP1), multidrug resistance-related protein 2 (MRP2), and the breast cancer resistance protein (ABCG2), significantly promotes cancer chemoresistance [12–16]. In our previous study, we have

Tumor Biol.

demonstrated that emodin, a traditional Chinese herbal medicine, enhances the anticancer efficacy of cisplatin by inhibiting MRP1 expression in GBC cells [17]. However, the precise mechanism of how MRP1 is regulated in GBC tumor remains largely unknown. MicroRNAs (miRNAs) are a group of evolutionarily conserved, small, endogenous, single-stranded noncoding RNAs consisting of an average of 22 nucleotides, which can bind the 3′-untranslated regions (3′-UTRs) of their target genes, resulting in translational repression or degradation of their target mRNAs [18]. A series of recent studies have shown that miRNAs play critical roles in a variety of physiological and pathological processes including cell proliferation, differentiation, metabolism, apoptosis, tumorigenesis, and tumor progression [19–22]. Several studies have investigated functions of miRNA in chemotherapy efficiency, but how miRNAs are involved in chemoresistance of GBC is still unclear [23–25]. miR-145, a tumor suppressor miRNA, has been found to be downregulated in several cancer types, such as GBC [26], prostate cancer [27], and bladder cancer [28]. However, whether downregulation of miR-145 also contributes to GBC chemoresistance has never been studied. Here, using two miRNA databases (TargetScan and miRanda), we found that the 3′-UTR of MRP1 mRNA is a potential target of miR-145. An inverse relationship between miR-145 and MRP1 mRNA expression was also found in GBC tissues. Further, we found that miR-145 is downregulated in GBC tissues and identified a mechanism of direct regulation of MRP1 by miR-145. Targeting of MRP1 by miR-145 leads to enhanced cisplatin sensitivity in vitro and in vivo. Moreover, we show that miR-145 and MRP1 are prognostic markers in GBC patients. Taken together, our findings suggest using miR-145 mimic to sensitize GBC cells to cisplatin as a potential therapeutic strategy.

Materials and methods Tissue samples Formalin-fixed, paraffin-embedded (FFPE) cancer tissues were collected from 82 patients harboring histologically confirmed GBC who underwent surgical resection of the gallbladder and postoperative adjuvant chemotherapy at the Department of Pathology (Renji Hospital) from January 2004 to December 2013 retrospectively. Fresh GBC tissues and the corresponding noncancerous gallbladder (CNG) tissues were also obtained from 36 patients among the 82 GBC patients. All fresh tumor samples were collected immediately after the surgical removal and snap-frozen in liquid nitrogen, then stored at −80 °C until total RNA was extracted. Postoperative survival was calculated from time of surgery to time of last follow-up or death. The collection and analysis

of patient samples were approved by the Ethical Committee of Renji Hospital, Shanghai Jiao Tong University School of Medicine, and written informed consent was obtained from all patients. Cell culture The human embryonic kidney 293 cells (HEK293FT) and human GBC cell lines of GBC-SD were maintained in Dulbecco’s Modified Eagle Medium (DMEM), and SGC996 were maintained in RPMI-1640 medium, with all media containing 10 % fetal bovine serum (FBS) and antibiotics (Gibco, Grand Island, NY, USA). Cells were maintained at 37 °C in a humidified atmosphere consisting of 5 % CO2. GBC-SD and SGC-996 cells were provided by the Academy of Life Sciences, Tongji University (Shanghai, China), and HEK293FT cells were purchased from Invitrogen (MD, USA). HEK293FT cells were used for adenovirus amplification. Cisplatin was dissolved in dimethyl sulfoxide (DMSO). The GBC cells were treated with cisplatin (4 μM) or control DMSO. Cell transfection Human miR-145 expression construct was generated by insertion of the coding sequence (CDS) of miR-145 into pCDHCMV-MCS-EF1-copGFP (System Biosciences, CA, USA). Recombinant lentiviruses were produced by transient transfection of HEK293FT cells, along with package vectors, using Lipofectamine 2000 (Invitrogen). After transfection for 48 h, the viruses were harvested and viral titers were determined. Then, GBC-SD cells were infected with lentiviruses in the presence of 4 μg/ml polybrene (Sigma), followed by puromycin selection (2 μg/ml). MRP1 expression vector was generated by insertion of the CDS of MRP1 into a pcDNA 3.1 vector (Invitrogen). GBC-SD and SGC-996 cells were transfected with the MRP1 expression plasmid using Lipofectamine 2000 (Invitrogen) transfection reagent according to the protocols. The pcDNA3.1 empty vector was used as negative control (vector). The miRNA-145 mimic, miRNA145 inhibitor, and siRNA of MRP1 were purchased from GenePharma (Shanghai, China). Cells were cultured to 60– 70 % confluence in six-well plates and then transfected using Lipofectamine 2000 (Invitrogen). Quantitative real-time PCR analysis Total RNA and miRNA were isolated from fresh tissues and cells using TRIzol reagent (Invitrogen) and miRNeasy Mini Kit (Qiagen, Hilden, Germany), and miRNAs were extracted from FFPE samples using miRNeasy FFPE Kit (Qiagen) according to the manufacturer’s instructions. After synthesizing cDNAs with Reverse Transcriptase M-MLV kit, the

Tumor Biol.

expression levels of miR-145 and MRP1 were analyzed using SYBR Premix Ex Taq (Takara, Shiga, Japan) and run with Applied Biosystems ViiA™ 7 Real-Time PCR System (Applied Biosystems, Foster City, CA). Data were analyzed by 2−ΔΔCT method [29] and presented relative to the expression of GAPDH for MRP1 and in relation to the expression of small nuclear U6 RNA for miR-145. The primer sequences used for qPCR are listed in Table 1. Cytotoxicity, cell apoptosis, cell proliferation, and cell migration assays Cell viability (GBC-SD and SGC-996) was identified by 3-(4, 5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4sulfophenyl)-2H-tetrazolium assay (MTS; Promega, Madison, WI, USA). Briefly, cells were plated (5 × 103 cells/ well) on 96-well plates and incubated overnight to allow cell attachment. Then, the cells were treated with cisplatin at concentrations of 0.125, 0.25, 0.5, 1, 2, 4, 8, 16, and 32 μM for 48 h. Subsequently, the MTS reagent (20 μl) was added to each well, followed by incubation at 37 °C in a humidified, 5 % CO2 atmosphere for 2 h. Finally, the absorbance was read at 490 nm by using a Synergy 2 (BioTek, VT, USA) plate reader. The cell viability was indicated as a percentage relative to the untreated control. Cell proliferation was also analyzed using MTS assay (Promega) when GBC-SD and SGC-996 cells were seeded into a 96-well plate (1 × 103 cells/well) and cultured for 72 h. Cell apoptosis of GBC-SD and SGC-996 were analyzed using Annexin V/PI Apoptosis Detection Kit (BD Biosciences, MA, USA) according to the manufacturer’s instructions. Cells were seeded in six-well plates and grown to approximately 60 % confluence, followed by treating with cisplatin (4 μM) for 48 h. Then, the floating and attached cells were harvested and incubated with Annexin V/PI for 15 min in the dark, followed by fluorescence-activated cell sorting (FACS) analysis. All the assays were carried out four times.

Table 1

U6 MRP1 GAPDH

Reporter vector constructs and dual luciferase reporter assay The fragment from MRP1-3′-UTR containing the predicted miR-145 binding site was amplified by PCR and then cloned into a pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega) to form the reporter vector MRP1-3′-UTR wild type. The putative binding site of miR-145 in the MRP13′-UTR was mutated by using a site-directed mutagenesis kit from Fast Mutagenesis System (TransGen Biotech, Beijing, China), and the new reporter vector was named as MRP1-3′UTR mutant. The miR-145 mimic and vector were cotransfected into GBC-SD cells, and Renilla luciferase reporter plasmid (pRL-TK) was also co-transfected as the internal reference. After transfection for 48 h, GBC-SD cells were lysed in passive lysing buffer, and then firefly and Renilla luciferase activities were analyzed using the Dual-Luciferase Reporter Assay System (Promega). The results of firefly luciferase activity were normalized to the Renilla luciferase activity. Western blot analysis For protein isolation from GBC-SD and SGC-996 cells, RIPA buffer supplemented with proteinase inhibitor cocktail was used. The protein concentration was determined using the BCA assay. Equal amounts of cell lysates were loaded on a 10 % sodium dodecyl sulfate-polyacrylamide gel for

Primers of qPCR

Genes (Homo sapiens)

miR-145

The migration abilities of the GBC-SD and SGC-996 cells were tested using 24-well transwell chambers with 8-μm pore size polycarbonate membrane (Corning, NY, USA). Cells (5 × 103 cells/well) were suspended in serum-free DMEM medium and seeded into the upper chamber of each insert, while the lower chamber contained DMEM medium with 10 % FBS. After incubation at 37 °C for 24 h, the migrated and invaded cells were fixed with methanol and stained for 30 min in a 0.1 % crystal violet solution in PBS.

Primers

Sequences

Anchor RT primer

CGACTCGATCCAGTCTCAGGGTCCGAGGTATT CGATCGAGTCGCACTTTTTTTTTTTTV

Forward Reverse Forward Reverse Forward Reverse Forward Reverse

3′-TCCCTAAGGACCCTTTTGACC-5′ 5′-AGTCTCAGGGTCCGAGGTATTC-3′ 5′-CTCGCTTCGGCAGCACA-3′ 5′-AACGCTTCACGAATTTGCGT-3′ 5′-CGCTCTGGGACTGGAATGT-3′ 5′-ACCCACACTGAGGTTGGTTA-3′ 5′-GAAGGTGAAGGTCGGAGTC-3′ 5′-GAAGATGGTGATGGGATTTC-3′

Tumor Biol.

antibodies of MRP1 (1:100, Santa Cruz). Positive staining cells were visualized by DAB systems and counterstained with hematoxylin. The stained sections were photographed and converted to a digital image using light microscopy equipped with camera (Olympus, Tokyo, Japan). The scoring of immunohistochemistry (IHC) is based upon the staining intensity (I) and the proportion of stained quantity (q) of tumor cells to obtain a final score (Q) defined as the product of I × q and was performed by two independent pathologists. The scoring system for I was 0 = negative, 1 = low, 2 = moderate, 3 = intense immunostaining; and for q was 0 = negative, 1 = 1—9 % positive, 2 = 10—39 % positive, 3 = 40—69 % positive, and 4 = 70—100 % positive cells. Cell apoptosis of xenograft sections was also detected by using In Situ Cell Death Detection Kit, POD (Roche, Basel, Switzerland) according to the manufacturer’s instruction. The sections were visualized with DAB and counterstained with hematoxylin. The number of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells was randomly

electrophoresis (SDS-PAGE) and transferred to PVDF membranes (Millipore, IL, USA). The membranes were blocked for 1 h at room temperature using Tris-bufferred saline with 0.05 % Tween 20 (TBST) and 5 % skimmed milk, and then the following primary antibodies were applied overnight at 4 °C: anti-MRP1 (Santa Cruz, CA, USA) and anti-β-actin (Sigma). After washing three times with TBST, the membranes were incubated with secondary antibody at room temperature for 2 h and washed again with TBST. Images of target proteins were detected by chemiluminescence HRP substrate kit (Millipore). Immunohistochemistry and terminal deoxynucleotidyl transferase dUTP nick end labeling assays All specimens from patients and subcutaneous xenografts fixed in 10 % buffered formalin were embedded in paraffin blocks. Consecutive 4-μm thick sections were analyzed using a standard immunohistochemistry protocol and stained by

B

1

80

Vector IC50: 4.45 µM miR-145

60 IC50: 2.51 µM Mimic-Con

IC50: 4.58 µM 40

Mimic-miR-145

IC50: 2.00 µM As-Con

20 IC50: 4.61 µM

**

60

*

40

0.125

20

2 8 Cisplatin (µM)

0.5

D

GBC-SD

80

Mimic-Con

IC50: 3.13 µM Mimic-miR-145

40

IC50: 2.08 µM As-Con

20 IC50: 3.05 µM 0

32

n.s

8

IC50: 4.89 µM 0.125

0.5

E

GBC-SD

10

2 8 Cisplatin (µM)

GBC-SD

n.s

6 4

Mimic-Con

Mimic-miR-145

As-Con

As-miR-145

2

45

on

-1 iR

45

-1

M

im

As

ic

-m

-m

As

iR

ic im M

-m As

ic im

-C

on

-1

-C

45

on iR

-1 -m

As

iR

-C

45

on -C ic im M M

G

80

Cell number ( 103)

SGC-996

100

*

60

**

40 20

H

SGC-996

10 8

SGC-996

n.s n.s

6 4

Mimic-Con

Mimic-miR-145

As-Con

As-miR-145

2

45

on iR

-1

-C

-m As

-m ic im M

As

45

-1 iR

ic im M

iR

-1

-C

45

on -C

-m As

-1 iR -m ic

im M

As

45

on -C ic

on

0

0

im

Annexin V+ cells (%)

miR-145

60 IC50: 1.96 µM

0

0

F

Vector IC50: 3.43 µM

As-miR-145

IC50: 7.50 µM

GBC

100

80

As-miR-145

0

CNG

Annexin V+ cells (%)

100 Relative cell viability (%)

***

0

C

SGC-996

GBC-SD 100

Relative cell viability (%)

Relative miR-145 expression

2

Cell number ( 103 )

A

M

Fig. 1 Effect of miR-145 on the sensitivity of GBC cells to cisplatin. a miR-145 expression was determined by qPCR in 36 pairs of human GBC tissue and CNG tissues. n = 36; bar, SEM. b Cell viability was measured by MTS assay to calculate IC50 of cisplatin in GBC-SD and SGC996 cells transfected with miR145 construct and miR-145 mimic or inhibitor. n = 4; bar, SEM. c, f Annexin V/PI staining to detect cell apoptosis rate of GBC-SD and SGC-996 cells transfected with miR-145 mimic or inhibitor, followed by exposure to cisplatin (4 μM) for 48 h. n = 4; bar, SEM. d, g Cell proliferation was measured by MTS assay at 72 h after GBC-SD and SGC-996 cells transfected with miR-145 mimic or inhibitor. n = 4; bar, SEM. e, h Cell migration was determined by Transwell assay in GBC-SD and SGC-996 cells transfected with miR-145 mimic or inhibitor. *P < 0.05; **P < 0.01; ***P < 0.001; n.s, no significant, Student’s t test

32

Tumor Biol.

counted in five fields, and the apoptosis index for each field was calculated as the percent of TUNEL-positive cells relative to the total cells.

Statistics Data are expressed as mean ± standard error of mean (SEM). Two-group comparisons were performed with unpaired twotailed Student’s t test. Survival probabilities were determined using Kaplan-Meier analyses and compared by the log-rank test. Each experiment consisted of at least four replicates per condition. SPSS17.0 software was used for all statistical analysis. P < 0.05 was considered statistically significant.

In vivo studies Animal maintenance and experimental procedures were strictly performed following the guidelines of the Animal Care and Use Committee of Shanghai Jiao Tong University. A total 1 × 10 6 GBC-SD/pcDNA3.1 miR-145 or GBC-SD/ pcDNA3.1 empty vector cells in 60 μl medium were subcutaneously transplanted into 4-week-old male nude mice of each group (group 1, vector; group 2, miR-145; group 3, vector + cisplatin; group 4, miR-145 + cisplatin; n = 6/group). When the average tumor size reached approximately 0.1 cm3, cisplatin was administered via intraperitoneal injection at a dose of 6 mg/kg at one dose every 16 days for a total of two doses. Saline was given as a placebo in nontreatment groups (group 1 and group 3). Tumor volumes were examined using external caliper once every 4 days and were calculated based on the equation: V = (length × width2) / 2 [30]. All mice were sacrificed at the 44th day, and the tumors were dissected out for hematoxylin and eosin (H&E) staining, IHC staining, and TUNEL staining.

2 1 0

C

12 9

* 6 3 0

CNG

D

Relative miR-145 expression

IHC score of MRP1 protein

Relative MRP1 mRNA expression

3

**

A previous study had suggested that miR-145 exhibits reduced expression in human GBC tissues [26], but how it is involved in GBC tumorigenesis and if it associated with GBC chemoresistance have never been investigated. We first examined expression of miR-145 in 36 GBC and CNG tissues by qPCR. Consistent with previous reports, miR-145 was found to be significantly decreased in GBC tissues (Fig. 1a). To assess the contribution of miR-145 to cisplatin resistance in

CNG

B

4

miR-145 is downregulated in GBC tissues and is associated with cisplatin sensitivity in GBC cell lines

GBC

A

Results

GBC

CNG

IHC staining

GBC

E

Pearson's r = -0.623 P < 0.001

2

1

0 0.0 0.4 0.8 1.2 1.6 Relative MRP1 mRNA expression

F GBC-SD

80

SGC-996

100

***

***

Relative cell viability (%)

100 Annexin V+ cells (%)

3

60 40 20 0

n.s

n.s

Con si-MRP1 Con si-MRP1 Con si-MRP1 Con si-MRP1 DMSO Cisplatin (4µM) DMSO Cisplatin (4µM)

Fig. 2 MRP1 expression in GBC tissues and its effect on chemosensitivity of GBC cells. a qPCR analysis of MRP1 mRNA levels in GBC and CNG tissues. n = 36; bar, SEM. b Semi-quantitative analysis of IHC staining for MRP1 protein in GBC and CNG tissues. n = 36; bar, SEM. c The IHC staining of MRP1 protein in GBC and CNG tissues. Scale bars, 10 μm. d MRP1 mRNA expression is decreased following forced expression of miR-145 in GBC and CNG tissues. Pearson’s correlation analysis was used. The expression level of

GBC-SD si-Con IC50: 4.34 µM si-MRP1 60 IC50: 1.77 µM

80

40 SGC-996 si-Con 20 IC50: 3.38 µM si-MRP1 IC50: 2.12 µM 0 0.125 0.5

2 8 Cisplatin µM)

32

small nuclear RNA U6 and GAPDH was used to normalize the qPCR results. e Annexin V/PI staining to detect cell apoptosis rate of GBC-SD and SGC-996 cells transfected with MRP1 siRNA (si-MRP1) or scrambled siRNA (Con), followed by exposure to cisplatin (4 μM) for 48 h. n = 4; bar, SEM. f MTS assay to calculate IC50 of cisplatin in GBCSD and SGC-996 cells transfected with MRP1 siRNA (si-MRP1) or scrambled siRNA (Con). n = 4; bar, SEM. *P < 0.05; **P < 0.01; ***P < 0.001, Student’s t test

Tumor Biol.

GBC cell lines, we inhibited miR-145 using miR-145 inhibitor and overexpressed it using miR-145 mimic or miR-145 construct in GBC-SD and SGC-996 cells, respectively, and analyzed cell sensitivity to cisplatin. Increased miR-145 expression accelerated the cytotoxic effect of cisplatin in GBCSD and SGC-996, with 1.05–2.58 μM reduction in IC50 (Fig. 1b). In contrast, reduced miR-145 expression enhanced their cisplatin resistance ability, with 1.84–2.89 μM increase in IC50 (Fig. 1b). In addition, Annexin V staining and FACS analysis revealed an increase in apoptotic cells in miR-145upregulated GBC-SD and SGC-996 cells and a decrease in apoptosis in miR-145-downregulated cells with cisplatin treatment (Fig. 1c, f). No significant change in cell proliferation or metastasis was observed after alteration in miR-145 expression (Fig. 1d, e, g, h). These results suggest that reduced miR145 expression potentially reduced the chemosensitivity of GBC to cisplatin.

A

B

SGC-996

GBC-SD

5

Relative MRP1 mRNA expression

MicroRNAs mainly exert their function through binding the 3′-UTR of their target genes to regulate their expression. Using TargetScan Release 7.0, we found that the 3′UTR of MRP1 gene is predicted to possess a putative binding site for miR-145. We then analyzed the expression of MRP1 in GBC and corresponding CNG tissues and found that MRP1 mRNA and protein expression levels were markedly increased in GBC tissues (Fig. 2a–c). Interestingly, Pearson correlation analysis showed that there was a significant inverse correlation between miR145 and MRP1 mRNA expression (P = 0.012, r = −0.623) (Fig. 2d), which suggested that miR-145 might play an important role in regulating MRP1 expression in GBC. We also verified that reduction in MRP1 increases cisplatin sensitivity of GBC cells (Fig. 2e, f).

*

***

4

GBC-SD

MRP1 -actin

3

SGC-996

2

**

***

MRP1

1

-actin

Luciferase

MRP1-3'UTR

hsa-miR-145

3'-UCCCUAAGGACCCUUUUGACCUG-5'

MRP1-wide type

5'-....UUUGUAAUGACUUAC ACUGGA -3' ...

hsa-miR-145

3'-UCCCUAAGGACCCUUUUGACCUG-5'

MRP1-mutant

5'-....UUUGUAAUGACUUAC UC AGCA -3' ...

150

**

80

n.s

60

**

40

**

20 0

-1 n.s

* 100 50 0 Mutant

SGC-996

100 Annexin V+ cells (%)

100

iR

Mimic-Con Mimic-miR-145

F

GBC-SD

45

on -C

GBC-SD

200

Wild type

E

-m

As

As

45

on M

im

ic

M

-m

im

ic

iR

-1

-C

45

-m iR -1

-C on

As

As

45

-m iR -1 45 M im ic -C M on im ic -m iR -1

-C on

As

D

TSS

Relative luciferase activity (%)

C

As

45

M im ic -C M on im ic -m iR -1

0

**

80

n.s

**

60

**

40 20

M RP 1 M im M ic RP -m + 1 iR -1 45

si -M R M si- P1 im M ic RP -m + 1 iR -1 45 Ve ct or

si -C on

M RP 1 M M im ic RP -m + 1 iR -1 45

si -M R M si- P1 im M ic RP -m + 1 iR -1 45 Ve ct or

0

si -C on

Annexin V+ cells (%)

Fig. 3 miR-145 directly regulates MRP1 in GBC cells. a, b The mRNA and protein levels of MRP1 in GBC-SD and SGC996 cells transfected with control mimic, miR-145 mimic, miR-145 inhibitor, and miR-145 inhibitor were measured by qPCR and western blotting assays. n = 4; bar, SEM. c The predicted miR145 binding sites in the 3′-UTR region of MRP1 (MRP1-wide type) and the designed mutant sequence (MRP1-mutant) are indicated. d Luciferase activity analysis of MRP1-3′-UTR were performed after co-transfection with MRP1-wild type or MRP1mutant pGL3 constructs and miR145 mimic by using the DualLuciferase Reporter Assay System. n = 4; bar, SEM. e, f Annexin V/PI staining to detect cell apoptosis rate of GBC-SD and SGC-996 cells after cotransfection with miR-145 mimic and MRP1 siRNA or MRP1 construct, followed by exposure to cisplatin (4 μM) for 48 h. n = 4; bar, SEM. The expression level of small nuclear RNA U6 and GAPDH was used to normalize the qPCR results. *P < 0.05; **P < 0.01; ***P < 0.01; n.s, no significant, Student’s t test

MRP1 is a target gene of miR-145 in GBC cell lines

Tumor Biol.

To further investigate the association between miR-145 and MRP1, we measured the expression level of MRP1 in GBCSD and SGC-996 cells treated with miR-145 inhibitor or mimic. Reduced mRNA and protein levels of MRP1 were observed in miR-145 mimic-treated cells, whereas MRP1 mRNA and protein levels were significantly increased in miR-145-downregulated cells (Fig. 3a, b). We then cloned the predicted miR-145 binding site on MRP1 gene into a pGL3 luciferase reporter vector and examined luciferase activity in GBC-SD cells supplemented with miR-145 mimic or control mimic, respectively. Also, a mutant MRP1-3′-UTR

Tumor volume (mm3)

A

1500

B

Cisplatin treatment (6mg/kg) Vector miR-145 Vector+Cisplatin miR-145+Cisplatin

1200 900

Vector

miR-145

Vector + Cisplatin

miR-145 + Cisplatin

600

*** 300 0

60 40 20

n.s at

in

or

pl

45

+C

+C

is

is

pl

-1

at

45

or iR

-1 iR m

Ve

iR m

Vector + Cisplatin

in

0

ct

45 -1

***

80

ct

Apoptosis rate (%)

in pl is

pl

+C

is or

Ve

ct

-1 iR m

miR-145

100

at

in at

45 -1

pl is

0

iR

at

in

or ct Ve

MRP1

H&E

Vector

TUNEL

3

45

+C

+C

is

iR

pl

-1

at

45

or ct m

Ve

F

in

0.0

**

6

or

0.5

9

ct

***

1.0

**

Ve

1.5

E

12

m

D

2.0

Ve

12 16 20 24 28 32 36 40 44 48 Day

+C

8

IHC score of MRP1

C

4

m

0

Tumor weight (g)

Fig. 4 miR-145 enhances the antitumor efficacy of cisplatin in vivo. a. Nude mice (n = 6) were subcutaneously injected with 1 × 106 cells stably transfected with pcDNA3.1 miR-145 or pcDNA3.1 empty vector and treated with cisplatin (6 mg/kg) at the 16th and 32nd day. Saline was given as a placebo in nontreatment groups. Tumor size was measured every 4 days after cisplatin treatment. n = 4; bar, SEM. b Representative photographs of tumors formed at the 44th day after subcutaneous transplantation are displayed. c Tumor weight analysis of the four paired groups. n = 6; bar, SEM. d, e Quantification of MRP1 expression and percentage of apoptotic tumor cells in paired GBC-SD tumor xenografts from five distinct images of each tumor sample. n = 6; bar, SEM. f H&E staining, MRP1 staining, and TUNEL staining in paraffin sections of the GBC-SD tumor xenografts. Representative images from six separate samples were shown. Original magnification, ×400; scale bars 50 μm. **P < 0.01; ***P < 0.01; n.s, no significant, Student’s t test

reporter construct harboring three point mutations in the center of the putative miR-145 response element was generated (Fig. 3c). In contrast to its WT counterpart, which exhibited strongly suppressed luciferase activity when treated with miR145 mimic in the GBC-SD cell line, the mutant MRP1-3′UTR failed to suppress luciferase reporter. (Fig. 3d). Furthermore, decreased rate of apoptosis in GBC-SD and SGC-996 caused by MRP1 overexpression was reversed by miR-145 mimic treatment, but the increased cytotoxic effect by reducing MRP1 expression in GBC-SD and SGC-996 cells could not be strengthened by upregulating miR-145 (Fig. 3e,

miR-145 + Cisplatin

Tumor Biol.

of miR-145 and MRP1 using qPCR and IHC, in 82 GBC patients who had received chemotherapy. After semiquantitative IHC analysis of these GBC samples, the Kaplan-Meier survival curves indicated that MRP1-low (IHC score <4, n = 36) patients had significantly enhanced postoperative survival times when compared to MRP1-high (IHC score ≥4, n = 46) patients (log-rank P = 0.017) (Fig. 5a). Further, GBC tissue samples were categorized as miR-145-high (n = 41) or miR-145-low (n = 41) using the median miR-145 mRNA expression level as threshold. Our data showed that low miR145 expression was highly correlated with the poor postoperative survival (log-rank P = 0.008) (Fig. 5b). Taken together, these data strongly suggest that decreased miR-145 expression, which is inversely correlated with MRP1 expression, is linked to poor prognosis in GBC patients after receiving chemotherapy.

f). This regulation was further supported by the observation that chemoresistance in GBC induced by miR-145 downregulation is mediated by MRP1. Overexpression of miR-145 sensitizes GBC to cisplatin treatment in vivo To investigate the effect of miR-145 on cisplatin sensitivity of GBC in vivo, GBC-SD cells were stably transfected with a miR-145 plasmid construct or control plasmid and s.c. into male nude mice. Until tumor presentation, one group was treated with cisplatin (6 mg/kg per 16 days with two doses in total) administered by i.p. injection, and the other with the carrier. Tumor volumes were monitored every 4 days. Xenograft tumor growth curves showed that miR-145 expression did not confer growth advantage to tumor compared to vector control (Fig. 4a). However, miR-145 expression induced a marked difference in response of tumors to cisplatin treatment. GBC-SD tumors with miR-145-expression were very sensitive to cisplatin treatment and failed to grow after receiving drugs, whereas tumors with vector-only expression were relatively resistant to cisplatin treatment (Fig. 4a). In addition, compared with vector tumors, tumors expressing miR-145 showed considerably smaller reduction in tumor size and weight, after 5-FU treatment (Fig. 4b, c). IHC and TUNEL analysis of tumor specimens showed diminished MRP1 expression and increased apoptosis rate in miR-145expresing tumors, in response to cisplatin, compared to empty vector-only tumors (Fig. 4d–f). These results from tumor xenograft mouse models provide further evidence that miR-145 enhances cisplatin-induced apoptosis by targeting MRP1.

Discussion Often diagnosed at an advanced stage, the high chemoresistance of GBC has made it a difficult malignancy to cure [5]. A thorough understanding of the precise mechanisms underlying GBC chemoresistance as well as identification of novel regulators to improve chemotherapeutic efficacy are therefore of critical importance. In this study, we identified miR-145 as a potential candidate biomarker of chemoresistance in GBC. miR-145 is downregulated in GBC patients and is correlated with poor prognosis. Altering miR-145 expression effectively modulated the chemoresistance of GBC cell lines. Further investigation into the mechanism revealed that miR-145 can directly target MRP1 gene and reduce its expression. This reduction in expression enhanced GBC chemosensitivity both in vivo and in vitro. miRNAs are a group of regulators capable of modulating target gene expression and are involved in both physiological and pathological processes [19–22]. miRNAs negatively

miR-145 and MRP1 expression are associated with prognosis in chemotherapy-administered GBC patients To determine whether lower miR-145 expression is associated with poor patient outcome, we analyzed the expression levels

A

B

Log-rank test: P = 0.017

100

Cumulative survival (%)

Cumulative survival (%)

Fig. 5 Downregulation of miR145 and upregulation of MRP1 in GBC are correlated with poor prognosis. Kaplan-Meier analysis of the postoperative survival of GBC patients stratified by the expression of miR-145 (a) or MRP1 (b) from tumor tissues. miR-145-low, n = 41; miR-145high, n = 41; MRP1-low, n = 36; MRP1-high, n = 46. The P value was calculated by a log-rank test

MRP1 low expression MRP1 high expression 50

0 0

20

40 60 80 Months after surgery

100

MRP1 expression Low High Median overall 19.2 survival time (month) 1-year survival rate 74.2%

10.7 45.5%

Log-rank test: P = 0.008 miR-145 high expression miR-145 low expression

100

50

0 0

20

40 60 80 Months after surgery

100

miR-145 expression Low High Median overall 11.1 survival time (month) 1-year survival rate 46.2%

16.3 69.4%

Tumor Biol.

regulate gene expression either by promoting degradation of mRNA or by interfering with translation of target messenger RNAs. To date, about 28,645 human miRNAs have been recorded in the miRBase database [31], and many are correlated with cancer chemoresistance [32, 33]. Modulation of miRNA expression levels has been proven to increase the efficacy of genotoxic drugs in various preclinical cancer studies. However, whether these miRNA regulators are also involved in GBC development and chemotherapeutic inefficiency is poorly understood. Expression of miR-145 is found to be reduced in several types of tumors and is associated with patients’ poor prognosis [26–28]. miR-145 is known to directly target important genes required for self-renewal and pluripotency, but whether it can also regulate genes involved in chemoresistance has never been investigated. To find the candidate genes responsible for GBC chemoresistance, we applied open-target prediction programs and found that miR145 could directly target MRP1. Our studies in GBC cell line confirmed that a regulatory mechanism exists between miR145 and MRP1. Modulating miR-145 expression level changed MRP1 expression and was responsible for GBC chemoresistance both in vivo and in vitro. Based on these findings, we demonstrate that upregulating miR-145 expression enhances cancer chemosensitivity and could serve as a therapeutic biomarker for GBC. MRP1 is a key component in chemoresistance development and is upregulated in various cisplatin-unresponsive cancers [34]. We found that MRP1 is also upregulated in GBC tissues. An inverse relationship between MRP1 and miR-145 was also identified in clinical CNG and GBC tissue samples. miR-145 could directly bind to the 3′-UTR sites of MRP1 by luciferase assay. Although various targets have been identified for miR-145, this is the first time we demonstrated that it could also affect cisplatin sensitivity by regulating expression of MRP1 in GBC cells. In addition, the prognostic potential of miR-145 and MRP1 was highlighted by the Kaplan-Meier analysis which suggests that the detection of these markers could potentially merit further GBC patient treatment. Although miR-145 has been reported to be involved in regulation of various genes associated with cancer cell proliferation and metastasis, it was surprising to find that altered miR-145 expression itself has no effect on GBC cell proliferation, apoptosis, and metastasis. Interestingly, a recent study in lung cancer suggested that tumor-specific deletion of miR143/145 in an autochthonous mouse model of lung adenocarcinoma did not affect tumor development [35]. Instead, stromal miR-143/145 expression could, on the contrary, promote tumorigenesis. As cancer is a heterogeneous disease, it is likely that the precise regulating mechanisms may be different and are context-dependent. Taken together, the present work therefore demonstrates for the first time that miR-145 could significantly affect GBC chemosensitivity through negative regulation of MRP1.

Therefore, targeting miR-145-MRP1 signaling may be a potential strategy for reversing chemoresistance in combating GBC. Acknowledgments This work was supported by the National Science Foundation of China (81072011, 81272748, and 81472240), National Key Technology R&D Program (2012BAI06B01), Foundation of Science and Technology Commission of Shanghai Municipality (12XD1403400), and Foundation of Shanghai Municipal Health Bureau (XBR2011035). Compliance with ethical standards Conflicts of interest None

References 1.

2. 3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Castro FA, Koshiol J, Hsing AW, Devesa SS. Biliary tract cancer incidence in the united states-demographic and temporal variations by anatomic site. Int J Cancer. 2013;133:1664–71. de Groen PC, Gores GJ, LaRusso NF, Gunderson LL, Nagorney DM. Biliary tract cancers. N Engl J Med. 1999;341:1368–78. Misra S, Chaturvedi A, Misra NC, Sharma ID. Carcinoma of the gallbladder. Lancet Oncol. 2003;4:167–76. Lazcano-Ponce EC, Miquel JF, Munoz N, Herrero R, Ferrecio C, Wistuba II, et al. Epidemiology and molecular pathology of gallbladder cancer. CA Cancer J Clin. 2001;51:349–64. Shibata T, Kokubu A, Gotoh M, Ojima H, Ohta T, Yamamoto M, et al. Genetic alteration of Keap1 confers constitutive Nrf2 activation and resistance to chemotherapy in gallbladder cancer. Gastroenterology. 2008;135:1358–68. 1368 e1351-1354. Zhan M, Wang H, Chen T, Chen W, Yang L, He M, et al. NOX1 mediates chemoresistance via HIF1alpha/MDR1 pathway in gallbladder cancer. Biochem Biophys Res Commun. 2015;468:79–85. Srivastava AK, Han C, Zhao R, Cui T, Dai Y, Mao C, et al. Enhanced expression of DNA polymerase eta contributes to cisplatin resistance of ovarian cancer stem cells. Proc Natl Acad Sci U S A. 2015;112:4411–6. Chen Q, Li W, Wan Y, Xia X, Wu Q, Chen Y, et al. Amplified in breast cancer 1 enhances human cholangiocarcinoma growth and chemoresistance by simultaneous activation of Akt and Nrf2 pathways. Hepatology. 2012;55:1820–9. Sau A, Pellizzari Tregno F, Valentino F, Federici G, Caccuri AM. Glutathione transferases and development of new principles to overcome drug resistance. Arch Biochem Biophys. 2010;500: 116–22. Li PL, Zhang X, Wang LL, Du LT, Yang YM, Li J, et al. MicroRNA-218 is a prognostic indicator in colorectal cancer and enhances 5-fluorouracil-induced apoptosis by targeting BIRC5. Carcinogenesis. 2015;36:1484–93. Ge G, Zhou C, Ren Y, Tang X, Wang K, Zhang W, et al. Enhanced slc34a2 in breast cancer stem cell-like cells induces chemotherapeutic resistance to doxorubicin via slc34a2-bmi1-abcc5 signaling. Tumour Biol. 2015 Liang X, Xu X, Wang F, Li N, He J. E-cadherin increasing multidrug resistance protein 1 via hypoxia-inducible factor-1alpha contributes to multicellular resistance in colorectal cancer. Tumour Biol. 2015. Wang Y, Liu L, Liu X, Zhang H, Liu J, Feng B, et al. Shugoshin1 enhances multidrug resistance of gastric cancer cells by regulating MRP1, Bcl-2, and Bax genes. Tumour Biol. 2013;34:2205–14. Lee TC, Ho IC, Lu WJ, Huang JD. Enhanced expression of multidrug resistance-associated protein 2 and reduced expression of

Tumor Biol.

15.

16.

17.

18. 19.

20. 21. 22. 23.

24.

25.

aquaglyceroporin 3 in an arsenic-resistant human cell line. J Biol Chem. 2006;281:18401–7. Sicchieri RD, da Silveira WA, Mandarano LR, Goncalves de Oliveira TM, Carrara HH, Muglia VF, et al. Abcg2 is a potential marker of tumor-initiating cells in breast cancer. Tumour Biol. 2015. Ding XW, Wu JH, Jiang CP. ABCG2: a potential marker of stem cells and novel target in stem cell and cancer therapy. Life Sci. 2010;86:631–7. Wang W, Sun YP, Huang XZ, He M, Chen YY, Shi GY, et al. Emodin enhances sensitivity of gallbladder cancer cells to platinum drugs via glutathion depletion and MRP1 downregulation. Biochem Pharmacol. 2010;79:1134–40. He L, Hannon GJ. MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev Genet. 2004;5:522–31. Li Z, Yu X, Shen J, Jiang Y. MicroRNA dysregulation in uveal melanoma: a new player enters the game. Oncotarget. 2015;6: 4562–8. Bader AG, Brown D, Winkler M. The promise of microRNA replacement therapy. Cancer Res. 2010;70:7027–30. Pencheva N, Tavazoie SF. Control of metastatic progression by microRNA regulatory networks. Nat Cell Biol. 2013;15:546–54. Sivadas VP, Kannan S. The microRNA networks of TGFbeta signaling in cancer. Tumour Biol. 2014;35:2857–69. Li Z, Yu X, Shen J, Law PT, Chan MT, Wu WK. MicroRNA expression and its implications for diagnosis and therapy of gallbladder cancer. Oncotarget. 2015;6:13914–21. Zhou H, Guo W, Zhao Y, Wang Y, Zha R, Ding J, et al. MicroRNA26a acts as a tumor suppressor inhibiting gallbladder cancer cell proliferation by directly targeting HMGA2. Int J Oncol. 2014;44: 2050–8. Jin K, Xiang Y, Tang J, Wu G, Li J, Xiao H, et al. miR-34 is associated with poor prognosis of patients with gallbladder cancer

through regulating telomere length in tumor stem cells. Tumour Biol. 2014;35:1503–10. 26. Letelier P, Garcia P, Leal P, Alvarez H, Ili C, Lopez J, et al. miR-1 and miR-145 act as tumor suppressor microRNAs in gallbladder cancer. Int J Clin Exp Pathol. 2014;7:1849–67. 27. Zhang X, Wu J. Prognostic role of microRNA-145 in prostate cancer: a systems review and meta-analysis. Prostate Int. 2015;3:71–4. 28. Ichimi T, Enokida H, Okuno Y, Kunimoto R, Chiyomaru T, Kawamoto K, et al. Identification of novel microRNA targets based on microRNA signatures in bladder cancer. Int J Cancer. 2009;125: 345–52. 29. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-delta delta C(T)) method. Methods. 2001;25:402–8. 30. Tomayko MM, Reynolds CP. Determination of subcutaneous tumor size in athymic (nude) mice. Cancer Chemother Pharmacol. 1989;24:148–54. 31. Kozomara A, Griffiths-Jones S. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 2014;42:D68–73. 32. Li JH, Luo N, Zhong MZ, Xiao ZQ, Wang JX, Yao XY, et al. Inhibition of microrna-196a might reverse cisplatin resistance of a549/ddp non-small-cell lung cancer cell line. Tumour Biol. 2015. 33. Liu RL, Dong Y, Deng YZ, Wang WJ, Li WD. Tumor suppressor miR-145 reverses drug resistance by directly targeting DNA damage-related gene RAD18 in colorectal cancer. Tumour Biol. 2015;36:5011–9. 34. Stordal B, Hamon M, McEneaney V, Roche S, Gillet JP, O’Leary JJ, et al. Resistance to paclitaxel in a cisplatin-resistant ovarian cancer cell line is mediated by P-glycoprotein. PLoS One. 2012;7: e40717. 35. Dimitrova N, Gocheva V, Bhutkar A, Resnick R, Jong RM, Miller KM, et al. Stromal expression of mir-143/145 promotes neoangiogenesis in lung cancer development. Cancer Discov. 2015.