Inhibition of MELK Protooncogene as an Innovative Treatment for Intrahepatic Cholangiocarcinoma

Abstract

Background and Objectives: Intrahepatic cholangiocarcinoma (iCCA) is a pernicious tumor characterized by a dismal outcome and scarce therapeutic options. To substantially improve the prognosis of iCCA patients, a better understanding of the molecular mechanisms responsible for development and progression of this disease is imperative. In the present study, we aimed at elucidating the role of the maternal embryonic leucine zipper kinase (MELK) protooncogene in iCCA. Materials and Methods: We analyzed the expression of MELK and two putative targets, Forkhead Box M1 (FOXM1) and Enhancer of Zeste Homolog 2 (EZH2), in a collection of human iCCA by real-time RT-PCR and immunohistochemistry (IHC). The effects on iCCA growth of both the multi-kinase inhibitor OTSSP167 and specific small-interfering RNA (siRNA) against MELK were investigated in iCCA cell lines. Results: Expression of MELK was significantly higher in tumors than in corresponding non-neoplastic liver counterparts, with highest levels of MELK being associated with patients' shorter survival length. In vitro, OTSSP167 suppressed the growth of iCCA cell lines in a dose-dependent manner by reducing proliferation and inducing apoptosis. These effects were amplified when OTSSP167 administration was coupled to the DNA-damaging agent doxorubicin. Similar results, but less remarkable, were obtained when MELK was silenced by specific siRNA in the same cells. At the molecular level, siRNA against MELK triggered downregulation of MELK and its targets. Finally, we found that MELK is a downstream target of the E2F1 transcription factor. Conclusion: Our results indicate that MELK is ubiquitously overexpressed in iCCA, where it may represent a prognostic indicator and a therapeutic target. In particular, the combination of OTSSP167 (or other, more specific MELK inhibitors) with DNA-damaging agents might be a potentially effective therapy for human iCCA.

Keywords: EZH2; FOXM1; MELK; intrahepatic cholangiocarcinoma; targeted therapies.

Figure 1

MELK and its downstream effectors are overexpressed in human intrahepatic cholangiocarcinoma (iCCA) specimens. (A–C) Quantitative real-time RT-PCR analysis of MELK, FOXM1, and EZH2 mRNA levels in normal livers (n = 5), iCCA (n = 52), and corresponding non-tumorous surrounding liver tissues (ST; n = 52). Quantitative values were calculated by using the PE Biosystems Analysis software and expressed as number target (N Target). N Target = 2^−ΔCt, wherein the ΔCt value of each sample was calculated by subtracting the average Ct value of the gene of interest from the average Ct value of the β-Actin gene. p-value was calculated using Mann–Whitney U test. * p < 0.01 when compared to Normal Liver (NL); ** p < 0.001 when compared to Normal liver (NL) and Surrounding Tissue (ST). Abbreviation: NS, not significant. (D–F) Expression levels of MELK, FOXM1, and EZH2 correlate with each other in human iCCA samples.

Figure 2

Representative immunohistochemical patterns of MELK and EZH2 proteins in human non-tumorous surrounding tumor tissues. A non-neoplastic portion of liver tissue in the proximity of one intrahepatic cholangiocarcinoma is displayed at two magnifications (100×, upper panels; 200×, lower panels) and shows weak cytoplasmic and nuclear staining of hepatocytes for MELK and EZH2 proteins, respectively. Low to moderate immunoreactivity for MELK and EZH2 in biliary epithelial cell is indicated by arrows. An equivalent staining pattern for MELK and EZH2 characterized normal healthy liver tissues (not shown). The CK19 staining was used as a marker of biliary differentiation of the tumors. Scale bar: 100 μm. Abbreviation: H&E, hematoxylin and eosin staining.

Figure 3

Representative immunohistochemical patterns of MELK and EZH2 proteins in human intrahepatic cholangiocarcinoma specimens (iCCA). A cholangiocellular tumor (T) is shown at two magnifications (100×, upper panels; 200×, lower panels) and exhibits concomitantly strong cytoplasmic and/or nuclear staining for MELK and EZH2 proteins, whereas weak/absent immunoreactivity is displayed by the non-tumorous surrounding counterpart (ST). The CK19 staining was used as a marker of biliary differentiation of the tumors. Scale bar: 100 μm. Abbreviation: H&E, hematoxylin and eosin staining.

Figure 4

MELK, FOXM1, and EZH2 levels correlate with an adverse outcome in human intrahepatic cholangiocarcinoma (iCCA). (A–C) Kaplan–Meier survival curves of human iCCA with high and low MELK, FOXM1, and EZH2 mRNA levels, showing the unfavorable outcome of patients with elevated expression of the three genes. (D–F) Linear regression analysis showing a statistically significant, inverse correlation between the levels of MELK, FOXM1, and EZH2, and the length (months) of patients’ survival.

Figure 5

Effect of inhibition of MELK by the non-selective inhibitor OTSSP167, either alone or in association with doxorubicin administration (DOXO) on proliferation (A; OD 450 nm), apoptosis (B; OD 405 nm), and DNA damage (C; absolute number of apurinic sites per 10⁵ base pairs) of HUCCT1 and HUH28 intrahepatic cholangiocarcinoma (iCCA) cell lines. Data are means ± standard deviation (SD) of three experiments conducted in triplicate. Tukey–Kramer test: At least p < 0.01; a, vs. control (untreated cells); b, vs. DMSO (solvent); c, vs. OTSSP167 40 nM; d, vs. OTSSP167 100 nM; e, vs. DOXO.

Figure 6

Effect of inhibition of MELK activity by the non-selective inhibitor OTSSP167, either alone or in association with doxorubicin administration (DOXO) on the levels of MELK, FOXM1, and EZH2 mRNA in HUCCT1 (A) and HUH28 (B) human intrahepatic cholangiocarcinoma (iCCA) cell lines. Data are means ± SD of three experiments conducted in triplicate. Tukey–Kramer test: At least p < 0.01; a, vs. control (untreated cells); b, vs. DMSO (solvent); c, vs. OTSSP167 40 nM; d, vs. OTSSP167 100 nM; e, vs. DOXO.

Figure 7

Effect of inhibition of MELK expression by the specific siRNA against MELK, either alone or association with doxorubicin administration (DOXO) on proliferation (A; OD 450 nm), apoptosis (B; OD 405 nm), and DNA damage (C; absolute number of apurinic sites per 10⁵ base pairs) of HUCCT1 and HUH28 intrahepatic cholangiocarcinoma (iCCA) cell lines. Data are means ± SD of three experiments conducted in triplicate. Tukey–Kramer test: At least p < 0.01; a, vs. control (untreated cells); b, vs. scramble siRNA; c, vs. MELK siRNA; d, vs. DOXO.

Figure 8

Effect of inhibition of MELK expression by the specific siRNA against MELK, either alone or association with doxorubicin administration (DOXO) on the levels of MELK, FOXM1, and EZH2 mRNA in HUCCT1 (A) and HUH28 (B) human intrahepatic cholangiocarcinoma (iCCA) cell lines. Tukey–Kramer test: At least p < 0.001; a, vs. control (untreated cells); b, vs. scramble siRNA; c, vs. MELK siRNA; d, vs. DOXO.

Figure 9

E2F1 is an upstream inducer of MELK in human intrahepatic cholangiocarcinoma (iCCA) cell lines. (A) In silico prediction of E2F1 binding sites on the MELK gene promoter using a cut-off of p-value of 0.001. Binding sites are represented in red squares. The EPDnew software was used for this scope. (B) Inhibition of E2F1 expression (left panel) by the specific siRNA against E2F1 induces the downregulation of MELK mRNA levels (right panel) in the HUCCT1 cell line. (C) Inhibition of E2F1 expression (left panel) by the specific siRNA against E2F1 triggers the downregulation of MELK mRNA levels (right panel) in the HUH28 cell line. Data are means ± SD of three experiments conducted in triplicate. Tukey–Kramer test: At least p < 0.001; a, vs. control (untreated cells); b, vs. scramble siRNA. (D) Quantitative real-time RT-PCR analysis of E2F1 mRNA levels in normal livers (n = 5), iCCA (n = 52), and corresponding non-tumorous surrounding liver tissues (ST; n = 52). Quantitative values were calculated by using the PE Biosystems Analysis software and expressed as number target (N Target). N Target = 2^−ΔCt, wherein the ΔCt value of each sample was calculated by subtracting the average Ct value of E2F1 gene from the average Ct value of the β-Actin gene. p-value was calculated using Mann–Whitney U test. ** p < 0.001 when compared to ST; * p < 0.01 when compared with NL. Abbreviation: NS, not significant. (E) Expression levels of E2F1 and MELK correlate with each other in human iCCA samples.