Upregulated MELK Leads to Doxorubicin Chemoresistance and M2 Macrophage Polarization via the miR-34a/JAK2/STAT3 Pathway in Uterine Leiomyosarcoma

Abstract

Uterine leiomyosarcoma (ULMS) is the most lethal gynecologic malignancy with few therapeutic options. Chemoresistance prevails as a major hurdle in treating this malignancy, yet the mechanism of chemoresistance remains largely unclear. In this study, we certified MELK as a poor prognostic marker through bioinformatic analysis of the GEO database. Cellular experiments in vitro revealed that MELK played an essential role in ULMS cells' chemoresistance and that a high expression of MELK could lead to doxorubicin resistance. mRNA profiling uncovered the pathways that MELK was involved in which led to doxorubicin resistance. MELK was found to affect ULMS cells' chemoresistance through an anti-apoptotic mechanism via the JAK2/STAT3 pathway. miRNA profiling also revealed that upregulated MELK could induce the decrease of miRNA-34a (regulated by JAK2/STAT3 pathway). We detected that MELK overexpression could induce M2 macrophage polarization via the miR-34a/JAK2/STAT3 pathway, contributing to doxorubicin chemoresistance in the tumor microenvironment. OTSSP167, a MELK inhibitor, may increase ULMS sensitivity to doxorubicin. Our investigation could propose novel targets for early diagnosis and precision therapy in ULMS patients.

Keywords: JAK2; M2 macrophage; MELK; STAT3; apoptosis; chemoresistance; uterine leiomyosarcoma.

Figure 1

MELK is overexpressed in uterine leiomyosarcoma (ULMS), and its high expression predicts poor prognosis. (A) Differential expression of MELK in 15 ULMS samples and seven uterine leiomyoma (ULM) and four myometrium (MM) from GEO data; ^*P < 0.05. (B) The positive rate of MELK expression in ULMS is higher than that in MM and ULM by immunohistochemistry staining; ^**P < 0.01. (C) Representative staining in MM, ULM, and ULMS. (D) Overall survival analysis based on MELK expression (high-expression group vs. low-expression group) in our cohort.

Figure 2

MELK can lead to doxorubicin chemoresistance in uterine leiomyosarcoma (ULMS) cells. (A) Cytotoxic assay of SK-UT-1 and SK-UT-1B cells with MELK knockdown or overexpression incubated with different concentrations of doxorubicin (0, 16, 32, 64, 128, or 256 nM) for 72 h; ^*P < 0.05, ^**P < 0.01. (B) When cells were incubated with doxorubicin at the concentration of 8 nM, the relative viabilities of SK-UT-1 and SK-UT-1B cells overexpressing MELK were significantly higher than that of the negative control group; ^*P < 0.05, ^**P < 0.01. (C) MELK protein expression was assessed by western blot after 0, 30, or 60 nM doxorubicin was added to the SK-UT-1 cells for 72 h. (D) Approximately 1,000 cells/well of SK-UT-1 and 1,500 cells/well of SK-UT-1B were seeded in six-well plates, treated with 0, 30, and 60 nM of doxorubicin, respectively, and cultured for 14 days; ^**P < 0.01. The results showed that MELK remarkably increased the colony formation in SK-UT-1 and SK-UT-1B cell lines. (E) Clonogenic assay of MELK suppressing the ULMS cells showed that MELK suppression apparently decreased the colony formation; ^**P < 0.01.

Figure 3

The effect of MELK on uterine leiomyosarcoma (ULMS) cells' chemoresistance was anti-apoptosis via the JAK2/STAT3 pathway and decline of miR-34a. (A) Detailed GO enrichment analysis of the pathways in mRNA profiling of SK-UT-1 PCMV-MELK (MELK) vs. SK-UT-1 PCMV-NC (PCMV) and SK-UT-1 PCMV-MELK treated with doxorubicin (MELK-ADR) vs. SK-UT-1 PCMV-MELK untreated (MELK). (B) Heatmap of miRNA profiling of SK-UT-1 PCMV-MELK (MELK) vs. SK-UT-1 PCMV-NC (PCMV) and SK-UT-1 PCMV-MELK treated with doxorubicin (MELK-ADR) vs. SK-UT-1 PCMV-NC treated with doxorubicin (PCMV-ADR). (C) miRNA-34a expression levels in MELK overexpressing SK-UT-1 and SK-UT-1B cells, MELK suppressed SK-UT-1 and SK-UT-1B cells, and negative control (NC) cells were each assessed by RT-qPCR; ^*P < 0.05, ^**P < 0.01. (D) JAK/STAT3 pathway markers and BCL2 were evaluated by western blot in SK-UT-1 and SK-UT-1B cells with MELK knockdown or overexpression. [The amino acid positions of phosphorylation were as follows: p-JAK2 (Y1007 + Y1008) and p-STAT3 (Y705)]. (E) Statistical analysis of the gray values of the western blot images of each detected protein; ^*P < 0.05, ^**P < 0.01. (F) Flow cytometric apoptosis assay was evaluated in ULMS cells with MELK knockdown or overexpression treated with doxorubicin at the concentration of 20 nM for 48 h. MELK overexpression inhibited doxorubicin-induced cell apoptosis and MELK knockdown led to more sensitivity to doxorubicin treatment in ULMS cells compared with NC cells; ^*P < 0.05, ^**P < 0.01.

Figure 4

MELK-induced M2 macrophage polarization via the miR-34a/JAK2/STAT3 pathway, and M2 macrophages can secrete IL-6, promoting uterine leiomyosarcoma cells' chemoresistance to doxorubicin. (A) SK-UT-1-suppressing MELK and negative control cells were transfected with miRNA-34a mimics with GFP and miRNA-NC. Then, the medium of each cell was used to incubate macrophages. MiR-34a (red dots) was elevated in these macrophages. (B) Number of red dots in macrophages that were treated with the conditioned medium of SK-UT-1-suppressing MELK and negative control SK-UT-1 cells; ^**P < 0.01. (C) RT-qPCR of miRNA-34a in exosomes showed that miRNA-34a was significantly higher in the exosome of MELK suppressing SK-UT-1 cells than that of negative control cells; ^**P < 0.01. (D) THP-1 cells were treated with PMA for 24 h and transfected with miR-NC and miR-34a mimics. M2 marker CD206 was significantly increased in miR-34a-transfected macrophages. (E) Putative miR-34a binding sequence in the IL6R 3′UTR. The binding sequence of miR-34a in the IL6R 3′UTR sequence was deleted as mutated type. (F) Analysis of the luciferase activity of the luciferase reporter plasmid containing either wild-type or mutant IL6R 3′UTR in the HEK 293T cell line which proved that IL6R is the downstream target of miR-34a; ^**P < 0.01. (G) Macrophages were incubated with the conditional media of MELK knockdown, overexpressing or suppressing SK-UT-1 and SK-UT-1B cells and NC cells for 48 h each. Then, macrophage polarization markers and JAK/STAT3 pathway markers were evaluated by western blot in these macrophages. [The amino acid positions of phosphorylation were as follows: p-JAK2 (Y1007 + Y1008) and p-STAT3 (Y705)]. (H) Statistical analysis of gray values of western blot images of each detected protein; ^*P < 0.05, ^**P < 0.01. (I) The macrophages were cultured with IL-4 to promote M2 polarization. Enzyme-linked immunosorbent assay (ELISA) showed that the secreted IL-6 was significantly higher in M2 macrophages (treated with IL-4) than in the normal untreated macrophages; ^**P < 0.01. (J) SK-UT-1 cells were incubated with IL6 (20 ng/mL) and then the treated and untreated cells were measured in cytotoxic assay. The relative cell viability of SK-UT-1 cells treated with IL-6 declined much more slowly than that of the untreated cells; ^*P < 0.05. (K) The macrophages were cultured with a conditioned medium of MELK-overexpressing SK-UT-1 cells and negative control cells for 48 h, respectively. ELISA showed that secreted IL-6 was significantly higher in macrophages treated with a conditioned medium of MELK-overexpressing SK-UT-1 cells than in macrophages treated with a conditioned medium of NC cells; ^*P < 0.05, ^**P < 0.01. (L) SK-UT-1 cells were cultured with conditioned media of M2 macrophages (treated with IL-4) and normal macrophages. The relative cell viability of SK-UT-1 cells treated with a conditioned medium of M2 macrophages declined much more slowly than that of cells treated with a conditioned medium of normal macrophages; ^**P < 0.01.

Figure 5

MELK inhibitor OTSSP167 could contribute to the therapeutic effect of doxorubicin. (A) The SK-UT-1B cells were injected subcutaneously. Tumor-bearing mice were separated randomly into three groups. One group was treated with doxorubicin only (3 mg/kg/day) intraperitoneally and the other treated group was given both orally fed OTSSP167 (10 mg/kg/day) and doxorubicin (3 mg/kg/day) injected intraperitoneally. The subcutaneous implanted uterine leiomyosarcoma (ULMS) tumors of each group of NCG mice. (B) The graphical representation of the tumor weights of each group of NCG mice; ^**P < 0.01. (C) The graphical representation of the tumor size of each group of NCG mice; ^*P < 0.05, ^**P < 0.01. (D) H&E and MELK immunohistochemical staining were performed in each group of tumors. (E) MELK not only affects the chemoresistance of ULMS cells through an anti-apoptotic mechanism via the JAK2/STAT3 pathway but also promotes M2 macrophages' polarization via the same pathway, ultimately contributing to the ULMS cells' chemoresistant ability.