Midkine as a potential diagnostic marker in epithelial ovarian cancer for cisplatin/paclitaxel combination clinical therapy

Abstract

The paclitaxel/cisplatin combination therapy commonly is used as the first-line treatment for advanced ovarian cancer patients. Midkine (MK), known as a novel tumor biomarker, has been elevated in the serum of patients with epithelial ovarian cancer (EOC). In this study, we aimed to detect the expression of MK in EOC tissues and evaluate clinical value of MK in diagnosis and therapy of EOC. We perform immunohistochemistry analysis to detect MK in EOC sample with postoperative platinum/paclitaxel combination therapy, we found that 71.4% (85 in 119 samples) of these samples were MK positive (> 10% of the cells were stained), and the expression of MK was significantly associated with disease histology (P = 0.038) as well as differentiation grade (P < 0.001). Moreover, MK positive samples show much more sensitive to cisplatin/paclitaxel combination therapy, compared with MK negative samples (P = 0.029). Those results indicated that MK expression might correlate with paclitaxel and/or cisplatin cytotoxicity in clinical therapy of EOC. Then, we evaluated the sensitivity to cisplatin and paclitaxel in 5 ovarian cancer cell lines (ES2, A2870, HO-8910, SKOV3 and SW626), and ES2, the highest MK expression among those cell lines, show the most sensitive to paclitaxel and cisplatin. Further, we confirmed this correlation between MK and paclitaxel and/or cisplatin cytotoxicity with the gain- and lost- of function. Finally, we demonstrated that MK enhanced the cytotoxicity of paclitaxel and/or cisplatin by accumulated cisplatin and paclitaxel through inhibited the expression of multidrug resistance-associated protein 3 (MRP3). In conclusion, MK could be an effective biomarker in diagnosis and therapy of EOC, especially for the drug selection at the time of initial diagnosis.

Introduction

Epithelial ovarian cancer (EOC) is the fourth most common type of female cancer worldwide [1,2] and has the highest mortality rate of all gynecologic cancers in China [3]. Debulking surgery is the initial line of therapy for advanced EOC [4], followed with chemotherapy, for which the current standard of care is platinum plus taxane therapy [5]. However, the 5-year survival rate for patients with stage III and IV tumors is less than 40% [6]. This may be attributed to the lack of early detection and appropriate individualized treatment [7-9]. Even patients with similar clinical characteristics, as cancer stage, histological type and grade, exhibit different disease progression and outcomes [10,11]. Therefore, it is necessary to discover new prognostic factors to support the diagnosis of EOC in patients at a high risk of developing the disease.

In clinical practice, the useful of immunohistochemistry (IHC) panels have been proven in diagnosing ovarian pathology [12,13]. The distinctive IHC stains of various markers may be used to confirm ovarian tumor categories and distinguish primary or metastatic EOC from colorectal cancer [12,14]. Single marker or combined markers, as the estrogen receptor (ER) [14], progesterone receptor (PR), antigen KI-67 (Ki67), Wilms tumor protein (WT1), and tumor protein 53 (p53) [15-17], have been reported in diagnosis of gynecologic cancers, but for EOC diagnosis and therapy, there are no single effective marker which has been reported. Although there are some reports that only demonstrate the proof-of-concept and potential to improve diagnostic efficiency by combining multiple biomarkers, but the role of those kinds of markers have not been understand.

Midkine (MK), as a novel prognostic biomarker in cancer, has been verified overexpressed in many cancers, including ovarian cancer [18-20]. Previous studies have shown that MK is a sensitive and specific tumor marker, and may be used as a novel therapeutic strategy to treat a large number of cancers [21,22]. Moreover, MK can play a role as a growth factor to support cell survival, and also has been regarded as a drug-resistance factor. Recently, some studies have shown that MK can protect different cancer cells against adriamycin [23], cannabinoid [24], and doxorubicin [22]. However, studies performed by Ota T et al. [21] and Kawai H et al. [25] show that the downregulation of MK could induce cisplatin resistance in oral squamous cell carcinomas and renal cells, respectively, and a negative correlation between MK expression and cisplatin cytotoxicity was reported in primary lung cancer cells [26]. It might indicate MK demonstrates different roles in different cancer.

Cisplatin/paclitaxel combination therapy, which is known as the first line treatment in a single institutional series of primary untreated advanced ovarian cancer patients, while lots of patients performed low sensitive to this kind therapy. In present study, we evaluated kinds of candidate biomarkers, as WT, MK, ER, PR, Ki67 and p53, and discussed the correlation of those markers with diagnostic performance in EOC. We found that the MK expression show a positive correlation with the predicted the survival time and chemosensitivity of EOC to paclitaxel/cisplatin. Moreover, MK could inhibit the expression of multidrug resistance-associated protein 3 (MRP3), and then enhanced the cytotoxicity of paclitaxel and/or cisplatin by accumulated cisplatin and paclitaxel.

Methods

Patients

To identify cases of primary EOC, we examined the surgical pathology files of the Department of Pathology, Nanjing First Hospital (Nanjing, China) and Jiangsu Cancer Hospital (Nanjing, China) between 2005 and 2012. Diagnosis in each case was based on recommended criteria listed in the classification proposed by the World Health Organization in 2004. All of the patients were Chinese. To select candidate markers, we reviewed two hospital’s IHC database results of all patients. The data for 119 primary EOC samples were investigated. The data were analyzed for WT1, P53, Ki67, PR, and ER staining, which are the routine tests used to diagnose EOC at our institution. Informed consent according to the criteria specified by the Nanjing Medical University was obtained from all patients. None of the patients had been treated with preoperative chemotherapy or radiation. The clinic pathologic characteristics of the study population are listed in Supplementary material Table S1. According to standard therapy guidelines, all patients underwent complete surgical resection, and received postoperative platinum/paclitaxel combination therapy. To define chemosensitivity, patients who relapsed 6 months or more after chemotherapy were designated “sensitive”, according to the Gynecological Cancer Intergroup (GCIG) cancer antigen 125 (CA125) criteria [27].

Serum tumor markers

Serum levels of traditional tumor markers, CA125, alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), carbohydrate antigen 15-3 (CA15-3), and carbohydrate antigen 19-9 (CA19-9) were measured by an automated chemiluminescent enzyme immunoassay system (CLEIA).

Immunohistochemistry

IHC assay was used to determine the expression of MK in EOC tissues. Paraffin-embedded tissues from primary surgery of 119 patients with EOC were used in this assay (The International Federation of Gynecology and Obstetrics [FIGO] stages I-IV). Briefly, 5-μm sections of EOC tissue from each case were prepared by extracting the paraffin in xylene. The sections were then re-hydrated and submerged in 3% H₂O₂ phosphate-buffered saline (PBS) for 15 min to inactivate endogenous peroxidase. The sections were treated with citrate solution (10 mM, pH 6.0; Maixin-Bio, Fujian, China) in a microwave oven at 95-100°C for 10 min to retrieve the antigens. The sections then were incubated with 2% bovine serum albumin for 30 min at room temperature to block nonspecific reactivity. Subsequently, the sections were incubated with antibodies against human MK (Clone ID: EP1143Y; dilution, 1:100; Epitomics, California, USA), WT1 (Clone ID: WT49; Zhongsan-Jingqiao, Beijing, China), p53 (Clone ID: DO7; Zhongsan-Jingqiao, Beijing, China), ER (Clone ID: EP1; Zhongsan-Jingqiao, Beijing, China), PR (Clone ID: EP2; Zhongsan-Jingqiao, Beijing, China), and Ki67 (Clone ID: SP6; Zhongsan-Jingqiao, Beijing, China) overnight at 4°C. After washing 3 times with PBS, an anti-rabbit secondary antibody labeled with horseradish peroxidase (Maixin-Bio, Fujian, China) was added to the sections, and incubated for 1 h at 37°C. The slides were then washed 5 times with PBS, and they were developed using a diaminobenzidine kit (Maixin-Bio, Fujian, China), according to the manufacturer’s instructions. Subsequently, the sections were counterstained, dehydrated, and mounted.

Cell culture and transfection

Human ovarian carcinoma cell lines (ES-2, A2780, SW626, HO-8910, and SKOV3) were obtained from the Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences. All the cell lines were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco-BRL, USA) supplemented with 10% fetal bovine serum (FBS; Gibco-BRL, USA) and 100 μg/mL each of streptomycin and penicillin G (Amresco, USA) at 37°C in a humidified atmosphere containing 5% CO₂.

The pcDNA3.1/MK recombinant plasmid and a specific RNAi sequence for human MK were transfected using Lipofectamine® 2000 reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions (as described previously; 30). Briefly, approximately 5 × 10⁴ cells/well were grown overnight in 24-well plates. When the cells reached 60-70% confluence, they were transfected with 0.8 μg of the recombinant plasmid, control plasmid, anti-MK-siRNA (100 nM), or negative control siRNA (100 nM) in serum-free medium using Lipofectamine® 2000. After 4 h of incubation at 37°C, 400 μL RPMI 1640 with 10% FBS was added to the cells. Stable plasmid transfectants were selected in the presence of 0.4 mg/mL G418 (Amersco) during 2 weeks of culture. The cells were then harvested for further testing. The expression of endogenous MK was analyzed by western blotting and real time polymerase chain reaction, as described in our previous study [28].

MK and MRP3 were analyzed by Q-PCR and Western blot (anti-MK IgG, Clone ID: EP1143Y; dilution, 1:1000, Epitomics, California, USA; anti-MRP3 polyclonal IgG, dilution, 1:2000, santa cruz, California, USA) as previously described (24). The primer sequences are as follows: MK, 5-AGCACCGAGGCTTCCT-3 and 5-AGGCTTGGCGTCTAGT-3; MRP3 5-CCATCGACCTGGAGACT-3 and 5-TATTTGGTGTCATTTCCTTCCT-3; GAPDH, 5-TGAAGGTCGGAGTCAACG-3 and 5-CAAAGTTGTCATGGATGA-3.

Half of the inhibition concentration (IC₅₀)

The cell counting kit (CCK)-8 was used to monitor the cytotoxicity of cisplatin and paclitaxel. Cells were plated at a density of 2000 cells/well in 96-well plates. After 24 h of culture, the cells were treated with paclitaxel or cisplatin (at final concentrations of 0, 0.625, 1.25, 2.5, 5, 10, or 20 µg/mL) for 48 h. The cells were then treated with the CCK-8 kit reagent (Beyotime, Shanghai, China) according to the manufacturer’s instructions and incubated at 37°C for 1.5 h. The absorbance at 570 nm was then estimated using an enzyme calibrator. Inhibition of cell growth was estimated by computing the percentage of viable cells compared with the control. All experiments were carried out in triplicate. The IC₅₀ was calculated by computing survival curves.

Intracellular drug accumulation assay

The intracellular paclitaxel [29] or cisplatin [21] accumulation assay was performed as described previously. Briefly, cells were cultured in 6-well plates and allowed to adhere to the plate overnight. When the cells reached approximately 80% confluence, they were treated with paclitaxel (4, 2, 1, and 0 μg/mL) or cisplatin (20, 10, 5, and 0 μg/mL) in serum-free medium. After incubation for different time periods, both the dead cells in the culture medium and the adherent viable cells were centrifuged at 1,000 × g for 15 min, harvested, and washed four times with ice-cold PBS.

The harvested samples, which treated with paclitaxel, were incubated with 0.5 mL of ZnSO₄ (4% w/v in H₂O/CH₃OH, 70/30) to precipitate proteins, and then the protein sample was dissolved with 300 μL of CH₃CN, filtered, and analyzed by high performance liquid chromatography (HPLC) to evaluate the amount of paclitaxel in the cells.

The resulting cells treated with cisplatin were solubilized by direct addition of 1 ml of 70% nitric acid for 10 min at room temperature, and then harvested, diluted to 5 ml by addition of 0.5% nitric acid, and analyzed the amount of cisplatin in the cells via inductively coupled plasma mass spectrometry (ICP-MS).

The protein concentration was estimated using the BCA Protein Assay Kit (Beyotime, Haimen, China), according to the manufacturer’s instructions.

Statistical analysis

The GraphPad Prism5 software (San Diego, CA) was used to analyze the clinical data. The correlations between IHC staining and multiple clinicopathologic characteristics were calculated using Fisher’s exact test. Survival was calculated from the day of surgery, and survival curves were obtained using the Kaplan-Meier method. P-values of less than 0.05 were considered significant.

Results

Profiles of serum tumor markers

The detection of serum tumor markers was known as an effective method to monitor the cancer recurrence during the recovery period. According to the clinical characteristics, including age, histology, quantity of ascites, FIGO, histological grade, and metastasis, we grouped 119 primary EOC patients and detected the levels of serum markers, as CA125, AFP, CEA, CA15-3, and CA19-9. As shown in the Supplementary material Table S2, we found that CA-125, a good marker of ovarian cancer [30], is associated with clinical parameters, including histology, quantity of ascites, FIGO and metastasis. AFP, CEA and CA15-3 could also represent some of the parameters. In EOC patients, AFP was associated with the tumor grade, CEA was related with the tumor histology, and CA15-3 levels tended to be positively correlated with increasing ascites quantity in serous EOC. But CA19-9 did not associate with any of the clinical characteristics.

Correlation between immunohistochemical profiles and clinicopathologic parameters

Previous study revealed that MK might significantly improve the detection efficiency of CA125 for ovarian cancer [19]. To explore the significance of MK in ovarian cancer, we evaluated IHC to analyze the MK expression in 119 EOC sample as well as other tumor biomarkers, as WT, ER, PR, Ki67 and p53. As shown in Table 1, we found that the positive samples of WT1, MK, ER, PR, p53 and Ki67 were 70.6%, 71.4%, 51.3%, 36.1%, 53.8% and 52.1% in 119 EOC patients, respectively. The representative images were shown in Supplementary material Figure S1. Then we further discussed the relation of WT, MK, ER, PR, Ki67 and p53 with clinicopathologic parameters. In serous ovarian cancer, there were significantly stronger of WT1 (P = 0.006) and MK (P = 0.038) expression, compared with in other types of ovarian cancer. Moreover, the expression of WT1, MK, and Ki67 were found to be markedly associated with a poor differentiation grade (P = 0.045, P = 0.000, and P = 0.000, respectively), and ER expression was found to be related with metastasis in EOC (P = 0.001). While no correlation was demonstrated between p53 and PR expression or any of the other clinicopathologic parameters examined.

Table 1.

Clinicopathological analysis of the expression of various tumor markers in surgically resected EOC

Variables		WT1		MK		ER		PR		Ki67		p53
		Neg	Pos	Neg	Pos	Neg	Pos	Neg	Pos	Neg	Pos	Neg	Pos
Histology	Serous (n = 75)	15	60	16	59	37	38	49	26	35	40	34	41
	Others (n = 44)	20	24	18	26	21	23	27	17	22	22	21	23
	P		0.006		0.038		0.983		0.812		0.872		0.950
FIGO stage	I/II (n = 51)	17	34	14	37	25	26	33	18	24	27	25	26
	III/IV (n = 68)	18	50	20	48	33	35	43	25	33	35	30	38
	P		0.542		0.977		0.895		0.978		0.979		0.730
Grade	G1/G2 (n = 37)	16	21	21	16	19	18	22	15	17	20	19	18
	G3 (n = 82)	19	63	13	69	39	43	54	28	40	42	36	46
	P		0.045		0.000		0.853		0.641		0.000		0.578
Metastasis	Yes (n = 75)	20	55	23	52	27	48	45	30	34	41	31	44
	No (n = 44)	15	29	11	33	31	13	31	13	23	21	24	20
	P		0.516		0.652		0.001		0.343		0.588		0.228
Chemosensitivity	No (n = 39)	13	26	18	21	17	22	23	16	20	19	20	19
	Yes (n = 57)	21	36	13	44	22	35	32	25	30	27	28	29
	P		0.892		0.029		0.781		0.948		0.938		1.000

Legend: ER, estrogen receptor; Ki67, antigen KI-67; MK, midkine; p53, tumor protein 53; PR, progesterone receptor; WT1, Wilms tumor protein; Neg, negative; Pos, positive.

Interestingly, we found that MK positivity patients show more sensitivity to platinum/paclitaxel combination therapy than to single-agent platinum therapy (P = 0.029). To investigate the role of MK in chemotherapy, we analyze the correlation between MK expression and outcomes of cisplatin/paclitaxel combination therapy (non-sensitive, n = 39; sensitive, n = 57). As shown in Figure 1, patients with positive and negative MK expression exhibit the significant differences in the clinical outcome and survival rate of cisplatin/paclitaxel-based chemotherapies (P = 0.0356). The Kaplan-Meier survival curve demonstrated that the MK positive patients preformed a good recovery with cisplatin/paclitaxel-based chemotherapies. This data indicated that the expression of MK could be an effective index in the clinical used the cisplatin/paclitaxel chemotherapy for EOC patients.

Figure 1.

The Kaplan-Meier survival curves computed for the EOC patients with cisplatin/paclitaxel-based chemotherapies. The patients with MK positive expression could significantly survival longer than those with negative expression who all received the cisplatin/paclitaxel-based chemotherapies (p = 0.0356).

MK sensitizes ovarian cancer cells to paclitaxel and cisplatin

To figure out the role of MK in cisplatin/paclitaxel chemotherapy, we compared the expression of MK and the cytotoxicity of cisplatin and/or paclitaxel in ovarian cancer cell lines, as SKOV3, ES2, SW626, A2780, and HO-8910 cells. As shown in Figure 2, those cells show the different expression of MK and exhibited different IC₅₀ values with cisplatin and/or paclitaxel. ES2 cells, which had the highest expression of MK among the 5 cell lines, had the lowest IC₅₀ values (paclitaxel, 2.02 μg/mL; cisplatin, 7.82 μg/mL). Moreover, combine of cisplatin (5 μg/mL) and paclitaxel (2 μg/mL) to treat, we found that the inhibition of ES2 cells were the highest, compared with other cell lines. Those data suggested that MK expression might contribute to the toxicity of paclitaxel and cisplatin in ovarian cancer cells.

Figure 2.

Comparison of the expression levels of MK mRNA (A) and protein (B), and the inhibition concentration at 50% (IC₅₀) values (C) and inhibition rate of paclitaxel (D) in different human ovarian cancer cell lines: ES2, A2780, HO-8910, Skov3 and SW626. Cells were treated with various concentrations of paclitaxel or cisplatin (0, 0.625, 1.25, 2.5, 5, 10 or 20 µg/mL), and the IC₅₀ values and the inhibition rates were determined via the cell counting kit (CCK-8) assay. *P < 0.05, ***P < 0.01 vs. the ES2 group.

Then, we applied gain- and lost- of function to further evaluate the effect MK effect on paclitaxel and cisplatin sensitivity of ovarian cancer cells, including treated by 200 ng/mL exogenous MK protein, overexpression of MK, or knockdown expression of MK with siRNA. After treaded with exogenous MK or overexpressing MK, ES2 cells preformed more sensitive to paclitaxel (IC₅₀, 1.43 and 0.87 μg/mL, respectively), native ES2 (IC₅₀, 2.02 μg/mL) or ES2-3.1 (IC₅₀, 2.53 μg/mL) as the control cells (Figure 3A and 3B). Next, we inhibited MK expression with siRNA. As shown in Figure 3B and 3C, the ES2-MK-siRNA cells performed lower sensitive to paclitaxel, compared with the ES2-Ne-siRNA cells, while the ES2-MK-siRNA cells didn’t show lower sensitive to cisplatin, but exogenous MK and overexpressing MK in ES2 cells could enhance the cytotoxicity of cisplatin (Figure 3B and 3C). Those result figured out that MK could strengthen the cytotoxicity of paclitaxel and cisplatin in chemotherapy.

Figure 3.

The expression levels of MK influenced paclitaxel cytotoxicity in ES2 ovarian cancer cells. Comparison of the expression levels of MK, as measured by western blotting (A) and the IC50 values of paclitaxel or cisplatin (B, C) in ES2 and ES2-related genetically modified cells. Cells were treated with various concentrations (0, 0.625, 1.25, 2.5, 5, 10, or 20 µg/mL), and the IC₅₀ values were determined via the CCK-8 assay. *P < 0.05, **P < 0.01 vs. the ES2 group.

MK enhances paclitaxel accumulation in cells

The paclitaxel plays the cytotoxicity function when it enter into the cells, whether MK regulated the accumulation of paclitaxel. To verify this hypothesis, we quantified intracellular paclitaxel during the time course, as 1 h, 2 h, 3 h and 4 h after paclitaxel treated. As shown in Figure 4, we found that overexpression of MK could increase the accumulation of paclitaxel, as in ES2-MK cells, while silence the expression of MK would decrease the paclitaxel accumulation, as in ES2-MK-siRNA cells. Moreover, the accumulation of paclitaxel was modified by MK as a dose- and time-dependent manner. ATP-binding cassette, subfamily B, member 1 (ABCB1) was known as an important transporter to help the paclitaxel into the cells. To investigate the mechanisms underlying MK-induced paclitaxel sensitivity, we compared the expression of ABCB1 and the extent of microtubule stabilization by paclitaxel in the ES2, ES2-3.1, and ES2-MK cell lines. However, there was no correlation between the ABCB1, microtubule stabilization, and MK in ES2 cells (data not shown).

Figure 4.

MK enhances paclitaxel accumulation in ES2 cells. Paclitaxel accumulation in ES2, ES2-MK, and ES2-MK-siRNA cells. Cells were treated with 2 μg/mL paclitaxel for the indicated times (A) or at the indicated paclitaxel doses for 3 h (B), and paclitaxel accumulation was measured by high performance liquid chromatography. *P < 0.05; ***P < 0.01 vs. the ES2 group.

MK regulates cisplatin accumulation in cells by MRP3

MK could not regulate the transporter of paclitaxel, whether MK exhibit the function in the efflux process. The efflux of anticancer agents was mediated by multidrug resistance proteins (MRPs) on cell surface, which is an important mechanism of drug resistance in tumor cells. MRP1 and MRP3 are the major transporters of cisplatin, which could extrude anticancer drugs out of the cell to inhibit the cytotoxicity [31,32]. After exogenous MK treatment or MK overexpression, we found that MRP3 expression was significantly increased, compared with control cells (Figure 5A and 5B), and the cisplatin accumulation was enhanced in ES2 cells as dose- and time-dependent manner (Figure 5C and 5D). While when we silenced the expression of MK in ES2 cells, the expression ofMRP3 and cisplatin accumulation did not show a significant change (data not shown). Those results might suggest that MK can promote the cisplatin accumulation through inhibited MRP3 expression in ovarian cells.

Figure 5.

MK inhibits the MPR3 expression (A, B) and enhances cisplatin accumulation in ES2 cells (C, D). Cisplatin accumulation in ES2, ES2-MK, and ES2-MK-siRNA cells. Cells were treated with 10 μg/mL paclitaxel for the indicated times (D) or at the indicated cisplatin doses for 3 h (C), and cisplatin accumulation was measured by ICP-MS. *P < 0.05; ***P < 0.01 vs. the ES2 group.

Discussion

Due to the association between tumor biomarkers and susceptibility of EOC in chemotherapy according to the type of cytotoxic drug used has not been studied in detail, ovarian cancer still has the highest mortality rates among gynecological malignancies [4,5]. Gregory ER et al. reported that the plasma MK was a good biomarker for ovarian cancer, and could significantly improve the diagnostic utility of CA125 in symptomatic women in a multi-analyte panel [19]. Therefore, we analyzed the concentrations of 5 serum biomarkers (CA125, CA19-9, CA15-3, AFP, and CEA) in ovarian cancer patients categorized by clinicopathologic features and 6 histological biomarkers (ER, P53, WT1, PR, Ki67, and MK), and we summarized the relationship between serum biomarkers and the histological biomarkers. Our results indicated that MK positivity patients had higher serum levels of CA125 and CEA than those with MK negativity patients. Next, we investigated the prognostic significance of the 6 histological biomarkers and demonstrated a significant correlation between MK expression in tumor tissues and clinicopathologic parameters (stage, grade, and response to platinum/paclitaxel chemotherapy). As reported in previous data, WT1 or Ki67 was more frequently detected in serous ovarian cancer than in other types of ovarian cancer. Taken together, these data support the hypothesis that MK not only plays an important role in tumor progression but may also be an indicator of the response to paclitaxel and/or cisplatin in the clinical treatment of ovarian cancer.

To verify this conclusion from clinical data, 5 ovarian cells with different MK expression levels were performed to detect drug cytotoxicity in vitro. The results revealed that the ES2 cell line, which had the highest expression level of MK, was most sensitive to paclitaxel and/or cisplatin. Subsequently, a similar result was observed for the MK-overexpressing ES2-MK cell line. MK could activate the Akt signal pathway, which is thought to provide intercellular cytoprotective signals to decrease the response to doxorubicin in SK-N-SH cells [33], while in our study, MK sensitized ovarian cancer cells to paclitaxel and/or cisplatin. The contradictory of these conclusions might indicated that MK regulated the function of drugs in chemotherapy through different mechanism.

The mechanisms underlying the MK-mediated sensitivity to paclitaxel and cisplatin are complex, and might include the induction of cellular transport disorders, aggravation of detoxification, decline of drug export, as well as alterations in tubulin isotype composition. In our study, ABCB1 expression (which mediates paclitaxel pharma cokinetics) and microtubules quantity were not altered by MK. Consistent with the results of the cytotoxicity assays, accumulation studies also showed that MK significantly increased the intracellular accumulation of paclitaxel in MK-overexpressing cells with a dose dependent manner. Those findings suggest that MK might play an important role in paclitaxel sensitivity, while the mechanism underlying this function of MK is still unclear and needs further research. Interestingly, we discovered that MK could promote the cisplatin accumulation in ES2 cells by inhibiting the MRP3 expression.

Our study has some limitations. Even though our samples were collected from patients with ovarian cancers, all patients had epithelial ovarian cancers. Therefore, the relationship between MK and paclitaxel/cisplatin sensitivity identified in this study might be limited to this tumor type. The association between MK and sensitivity to ovarian cancer needs to be validated in a larger sample size. Moreover, we could not completely eliminate interference caused by surgical debulking, which is a sub-optimal surgery and could lead to residual tumor cells and affect recurrence and metastasis in our study.

In conclusion, our data suggest that MK may be a useful biomarker for ovarian cancer and the expression of MK in tumor tissues may help to predict the aggressiveness of the cancer and sensitivity to paclitaxel/cisplatin chemotherapy. However, larger cohort studies are needed to confirm these findings. The association between MK expression and paclitaxel and cisplatin accumulation in tumor cells is an important finding that needs to be considered during drug selection at the time of initial diagnosis.

Disclosure of conflict of interest

The authors do not report any conflict of interest regarding this work.