HK2 promotes cell migration and invasion of human ovarian cancer cells in vitro

To further investigate the function of HK2 on regulating migratory capacity in human ovarian cancer cells, exogenous HK2 was stably overexpressed in SKOV3 (SKOV3-GFP and SKOV3-HK2, Fig. 1A) cells; conversely, endogenous expression of HK2 was knocked down by stably transfecting shRNA plasmids in A2780 (A2780-shCtr and A2780-shHK2, Fig. 1D) cells. Firstly, transwell assays was performed to evaluate the capacity for cell motility in HK2-modified human ovarian cancer cells. As shown in Fig. 1B, the transwell migration analysis revealed that the number of SKOV3-HK2 cells that migrated across the membrane was much more than SKOV3-GFP cell (112.40 ± 13.18 vs 62.80 ± 7.39, p < 0.05, Fig. 1B). Conversely, the number of A2780-shHK2 cells that migrated across the membrane was much less than A2780-shCtr cells (20.40 ± 3.61 vs 57.80 ± 8.93, p < 0.05, Fig. 1E). Furthermore, the transwell invasive analysis revealed that the number of SKOV3-HK2 cells that migrated across the membrane was much more than SKOV3-GFP cell (56.00 ± 8.42 vs 23.60 ± 6.31, p < 0.05, Fig. 1C). Conversely, the number of A2780-shHK2 cells that migrated across the membrane was much less than A2780-shCtr cells (13.40 ± 2.07 vs 28.00 ± 5.62, p < 0.05, Fig. 1F).

Fig. 1

HK2 enhances the migration and invasion ability in human ovarian cancer cells in vitro. Stably transfected cell lines were identified by western blotting: (A) SKOV3-GFP and SKOV3-HK2; (D) A2780-shCtr and A2780-shHK2. The migratory capacities were analyzed by the transwell assay, and the number of migratory cells is shown (scale bar, 100 μm): (B) SKOV3-GFP and SKOV3-HK2; (E) A2780-shCtr and A2780-shHK2. The invasive capacities were analyzed by the transwell assay, and the number of migratory cells is shown (scale bar, 100 μm): (C) SKOV3-GFP and SKOV3-HK2; (F) A2780-shCtr and A2780-shHK2. The migratory potential was analyzed by wound-healing assays performed for 0 and 48 h: (G) SKOV3-GFP and SKOV3-HK2; (H) A2780-shCtr and A2780-shHK2. The data are shown as the mean ± SD of three independent experiments. * p < 0.05, ** p < 0.01 vs. control using one-way ANOVA

Additionally, wound-healing assays revealed that a significant increase in wound closure was observed in SKOV3-HK2 cells, comparing with that observed in the SKOV3-GFP cells (p < 0.05, Fig. 1G). Conversely, a significant decrease in wound closure was found in A2780-shHK2 cells, comparing with that observed in the A2780-shCtr cells (p < 0.05, Fig. 1H). All of these results demonstrate that exogenously expressed HK2 in human ovarian cancer cells significantly enhances cell migratory and invasive capacity in vitro.

HK2 altered the expression of EMT-related proteins in human ovarian cancer cells

To explore the potential EMT-related proteins that are probably mediated by HK2 in human ovarian cancer cells to enhance cell migratory and invasive, real-time PCR and western blotting were applied to verify the expression of key EMT-related proteins in SKOV3-GFP, SKOV3-HK2, A2780-shCtr and A2780-shHK2 cells.

As shown in Fig. 2A, the mRNA levels of Fibronectin, MMP9, CHD2, vimentin, ZEB1 and ZEB2 were much higher in SKOV3-HK2 cells than that in SKOV3-GFP cells. Conversely, the decreased mRNA levels of Fibronectin, MMP9, CHD2, vimentin, ZEB1 and ZEB2 were observed in A2780-shHK2 cells, comparing with A2780-shCtr cells (Fig. 2B, p < 0.05). Consistently, the protein levels of fibronectin, MMP9, CHD2, vimentin, ZEB1 and ZEB2 were much higher in SKOV3-HK2 cells than that in SKOV3-GFP cells (Fig. 2C, p < 0.05). And the decreased protein levels of Fibronectin, MMP9, CHD2, vimentin, ZEB1 and ZEB2 were observed in A2780-shHK2 cells, comparing with A2780-shCtr cells (Fig. 2D, p < 0.05).

Fig. 2

HK2 elevated the expression of EMT-related proteins in human ovarian cancer cells. The mRNA expression of Fibronectin, MMP9, CDH2, Vemintin, ZEB1, ZEB2 was detected by real-time quantitative PCR, and the quantitative analysis is shown: (A) SKOV3-GFP and SKOV3-HK2; (B) A2780-shCtr and A2780-shHK2. The protein level of Fibronectin, MMP9, CDH2, Vemintin, ZEB1, ZEB2 was detected by western blot: (C) SKOV3-GFP and SKOV3-HK2; (D) A2780-shCtr and A2780-shHK2, and the quantitative analysis is shown: (E) SKOV3-GFP and SKOV3-HK2; (F) A2780-shCtr and A2780-shHK2. G The positive correlation between HK2 and Fibronectin, MMP9, CDH2, Vemintin, ZEB1, ZEB2 expression in ovarian serous cystadenocarcinoma were confirmed from the GEPIA online database. The data are shown as the mean ± SD of three independent experiments. * p < 0.05, ** p < 0.01 vs. control using one-way ANOVA

Additionally, the positive correlation between HK2 and fibronectin, MMP9, CHD2, Vimentin, ZEB1 and ZEB2 in human ovarian cancer were confirmed from the GEPIA online database (Fig. 4G, p < 0.05). These results suggested that the enhanced cell migratory and invasive capacity that mediated by HK2 likely depends on the up-regulated EMT-related proteins expression in human ovarian cancer.

HK2 promotes cell growth by reducing p21/p27 expression in human ovarian cancer cells

In this study, the clonogenic formation assay was used to detected the potential function of HK2 on regulating cell proliferation in human ovarian cancer cells. As shown in Fig. 3, the number of cell clones in SKOV3-HK2 cells was much more than SKOV3-GFP cell (117.70 ± 14.14 vs 63.67 ± 6.80, p < 0.05, Fig. 3A). Conversely, the number of cell clones in A2780-shHK2 cells was much less than A2780-shCtr cells (19.33 ± 4.46 vs 91.00 ± 8.75, p < 0.05, Fig. 3D). Moreover, the cell growth curves and MTT assays revealed that the SKOV3-HK2 cells grew much faster than SKOV3-GFP cells (Fig. 3B and C, p < 0.05). Conversely, when endogenous expression of HK2 was knocked down in A2870-shHK2 cells, cells grew much slower than A2780-shCtr cells (Fig. 3E and F, p < 0.05). These results demonstrated that HK2 enhanced cell growth in human ovarian cancer cells.

Fig. 3

HK2 promoted cell proliferation of human ovarian cancer cell lines in vitro. The colony formation assay was used to detected the long-term cell survival and growth in HK2-modified cells and control cells and the quantitative analysis is shown: (A) SKOV3-GFP and SKOV3-HK2; (D) A2780-shCtr and A2780-shHK2. The growth curves were used to detected the cell proliferation in HK2 modified cells: (B) SKOV3-GFP and SKOV3-HK2; (E) A2780-shCtr and A2780-shHK2. The MTT assay were used to detected the cell viability in HK2 modified cells: (C) SKOV3-GFP and SKOV3-HK2; (F) A2780-shCtr and A2780-shHK2. The cell cycle was analyzed in HK2 modified cells by using flow cytometry: (G) SKOV3-GFP and SKOV3-HK2; (I) A2780-shCtr and A2780-shHK2. The mRNA expression of cyclin A1, cyclin D1, cyclin E1, p21, p27 and HK2was detected by real-time quantitative PCR, and the quantitative analysis is shown: (H) SKOV3-GFP and SKOV3-HK2; (J) A2780-shCtr and A2780-shHK2. The protein level of p21 and p27 was detected by western blot and the quantitative analysis is shown: (K) SKOV3-GFP and SKOV3-HK2; (L) A2780-shCtr and A2780-shHK2

Moreover, the flow cytometry was used to detect the differences in the distribution of cells in the cell cycle between the HK2-modified cells and their control cells. As shown in Fig. 3G, the decreased proportion of cells in the G0/G1 phase (49.11 ± 3.34) and increased proportion of cells in the S phase (32.02 ± 2.68) was observed in in SKOV3-HK2 cells, comparing with SKOV3-GFP cells (G0/G1 phase: 61.91 ± 2.49; S phase: 23.83 ± 2.45). Conversely, the increased proportion of cells in the G0/G1 phase (71.39 ± 2.72) and decreased proportion of cells in the S phase (22.75 ± 3.19) was observed in in A2870-shHK2 cells, comparing with A2780-shCtr cells (G0/G1 phase: 62.58 ± 2.23; S phase: 28.14 ± 2.23, Fig. 3I). These results suggested that HK2 could accelerate cell cycle progression in human ovarian cancer cells.

Additionally, real-time PCR was used to verify the expression of key cell cycle-related proteins in SKOV3-GFP, SKOV3-HK2, A2780-shCtr and A2780-shHK2 cells. As shown in Figs. 3H, the mRNA levels of cyclin A1, cyclin D1 and cyclin E1 were much higher, p21 and p27 were much lower in SKOV3-HK2 cells than that in SKOV3-GFP cells. Conversely, the mRNA levels of cyclin A1, cyclin D1 and cyclin E1 were much lower, p21 and p27 were much higher in A2780-shHK2 cells, comparing with A2780-shCtr cells (Fig. 3J, p < 0.05). Consistently, the decreased protein levels of p21 and p27 were observed in HK2 overexpressed SKOV3 cells (Fig. 3K, p < 0.05), and the increased protein levels of p21 and p27 were observed in HK2 knocked down A2780 cells (Fig. 3L, p < 0.05).

HK2 elevated Akt1 and p-Akt1 expression in human ovarian cancer cells

Previous study demonstrated that HK2 participated in regulating Akt1 and p-Akt1 expression in various cancer. In order to determine whether the Akt1 and p-Akt1 expression could be regulated under HK2 expression in human ovarian cancer cell lines, western blot was used to detected the protein levels of Akt1 and p-Akt1 in HK2- modified human ovarian cancer cells. As shown in Fig. 4, both of the protein level of Akt1 and p-Akt1 were increased in HK2 overexpressed SKOV3 cells (Fig. 4A, p < 0.05). Conversely, the reduction of protein level of Akt1 and p-Akt1 were observed in HK2 knocked down A2780 cells (Fig. 4B, p < 0.05).

Fig. 4

HK2 elevated Akt1 and p-Akt1 expression in human ovarian cancer cells. The protein level of Akt1 and p-Akt1 was detected by western blot and the quantitative analysis is shown: (A) SKOV3-GFP and SKOV3-HK2; (B) A2780-shCtr and A2780-shHK2. C The protein level of Akt1, p-Akt1, Fibronectin, MMP9, p21 and p27 was detected by western blot in MK2206 treated SKOV3-HK2 cells and the quantitative analysis is shown. D The protein level of Akt1, p-Akt1, Fibronectin, MMP9, p21 and p27 was detected by western blot in A2780-shHK2 cells that transiently transfected with an Akt1 recombinant plasmid, and quantitative analysis is shown. E The expression of HK2, p-Akt1, MMP9 and FN1 was detected in serial sections of human ovarian cancer tissues by using immunocytochemistry analysis (scale bar, 50 and 10 μm). The correlation between HK2 and p-Akt1(F), Fibronectin (G), MMP9 (H), in human ovarian cancer tissues was confirmed by using Pearson correlation analysis, n = 25

Furthermore, in order to determine whether the alteration of Akt1 and p-Akt1 expression under HK2-modified was responsible for the changing of cell growth, migratory and invasive capacity in human ovarian cancer cells, MK2206, an inhibitor of Akt1/p-Akt1, was used to block the elevated expression of Akt1 and p-Akt1 in HK2 overexpressed SKOV3 cells. As shown in Fig. 4C, when the elevated expression of Akt1 and p-Akt1 in SKOV3-HK2 cells were blocked by MK2206, reduced Akt1, p-Akt1, fibronectin, MMP9 and induced p21, p27 expression was observed in MK2206-treated SKOV3-HK2 (Fig. 4C, p < 0.05). Moreover, the protein level of Akt1 and p-Akt1 expression was rescued in HK2-knockdown A2780-shHK2 cells via transient transfection of an Akt1 recombinant plasmid (pIRES2-AcGFP-Akt1). As shown in Fig. 4D, an induced Akt1, p-Akt1, fibronectin, MMP9 and reduced p21, p27 expression was observed in A2780-shHK2-Akt1 cells. These results demonstrated that HK2 could elevate Akt1 and p-Akt1 expression in human ovarian cancer cells, subsequently enhancing cell motility by inducing Fibronectin and MMP9 expression, promoting cell growth by reducing p21 and p27 expression.

Moreover, in order to confirm the positive correlation between the expression of HK2 and p-Akt1, Fibronectin and MMP9 in human ovarian cancer tissues, serial sections of human ovarian cancer tissues (n = 25) were immunostained with antibodies specific for HK2, p-Akt1, fibronectin and MMP9 (Fig. 4E). As shown in Fig. 4, the positive correlation between HK2 and p-Akt1 (Fig. 4F, r = 0.4979, p = 0.0105), fibronectin (Fig. 4G, r = 0.1850, p = 0.0320) and MMP9 (Fig. 4H, r = 0.3821, p = 0.0020) in these human ovarian cancer tissues was confirmed by using Pearson correlation analysis.

Additionally, as shown in Fig. 5, the positive correlation between HK2 and CDH2, fibronectin, MMP9, ZEB1, ZEB2 and vimentin in OV (ovarian serous cystadenocarcinoma) was confirmed by using TIMER 2.0 (http://timer.cistrome.org/) [19]. Although the expression of HK2 in OV (ovarian serous cystadenocarcinoma) was not so correlation between the prognosis of patient, however, the expression of HK2 was associated with a poorly prognosis of patient in CESC (cervical squamous cell carcinoma and endocervical adenocarcinoma), KICH (kidney Chromophobe), SARC (sarcoma), PCPG (pheochromocytoma and paraganglioma), LIHC (liver hepatocellular carcinoma), LUAD (lung adenocarcinoma) and LGG (brain lower grade glioma, Fig. 6A and B, TIMER 2.0) [19].

Fig. 5

The correlation between HK2 and Akt1, CDH2, fibronectin, MMP9, ZEB1, ZEB2 and vimentin in various type of cancer. A The correlation between HK2 and Akt1, CDH2, fibronectin, MMP9, ZEB1, ZEB2 and vimentin in various type of cancer was confirmed by using TIMER 2.0 (http://timer.cistrome.org/). B The positive correlation between HK2 and CDH2, fibronectin, MMP9, ZEB1, ZEB2 and vimentin in OV (ovarian serous cystadenocarcinoma) was confirmed by using TIMER 2.0 (http://timer.cistrome.org/)

Fig. 6

The correlation between HK2 and patient outcomes in various type of cancer. A The correlation between HK2 and patient outcomes in various type of cancer was confirmed by using TIMER 2.0 (http://timer.cistrome.org/). B The positive correlation between HK2 and patient outcomes in CESC, KICH, SARC, PCPG, LIHC, LUAD and LGG was confirmed by using TIMER 2.0 (http://timer.cistrome.org/). C Proposed model of the mechanisms by which HK2 promotes cell motility and growth in human ovarian cancer