Spindle and kinetochore-associated complex subunit 3 (SKA3) promotes stem cell-like properties of hepatocellular carcinoma cells through activating Notch signaling pathway.

BACKGROUND: Cancer stemness contributes to hepatocellular carcinoma (HCC) initiation, metastasis, drug resistance, and recurrence. The spindle and kinetochore-associated (SKA) complex has been shown to be involved in tumor progression; however, its effects on cancer stem cell-like properties have not yet been examined. This research sought to study each subunit of the SKA complex in HCC systematically.
METHODS: Bioinformatic analyses were carried out to examine the expression and clinical data of the SKA complex's each subunit in HCC. The expression of the target genes was detected by quantitative reverse transcription-polymerase chain reaction and Western blot assays. Clone formation and Transwell assays were performed to assess the proliferation and migration abilities of the SKA complex's each subunit. Sphere formation assays and subcutaneous xenograft experiments were performed to investigate the effects of SKA complex subunit 3 (SKA3) on the self-renewal and tumorigenic abilities of HCC.
RESULTS: Each subunit of the SKA complex was highly expressed in HCC, but only SKA complex subunit 1 (SKA1) and SKA3 were associated with the poor overall survival of HCC patients. Additionally, the HCC cells overexpressing SKA3 exhibited increased migration, invasion, proliferation, self-renewal, Sorafenib resistance and tumorigenic abilities. Notch signaling played a vital role in the process by which SKA3 promoted HCC stemness.
CONCLUSIONS: SKA3 promotes HCC stem cell-like properties via the Notch signaling pathway. As SKA3 appears to act as a regulator of stemness in HCC, it might be a potential molecular target for HCC. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

Hepatocellular carcinoma (HCC) is a highly lethal tumor. With more than 841,000 new cases and 781,000 deaths annually, HCC accounts for 5.7% of all cancers and 8.2% of all cancer-related deaths, and constitutes a major health problem worldwide (1). As the vast majority of HCC patients are diagnosed at an advanced stage, surgical treatment is difficult, and many patients have to receive systemic treatment; however, the overall survival rate of patients is poor (2). Additionally, even after surgical treatment, HCC still has a high recurrence and metastasis rate (3). There is increasing evidence that a subpopulation of cancer cells possesses stem cell-like properties, such as self-renewal, treatment resistance, and permanent proliferation abilities, which lead to complex tumor heterogeneity, cancer initiation and development (4,5). However, the mechanism of HCC stemness has not yet been fully elucidated. Clarifying the influencing factors and molecular mechanisms involved in HCC stem cell-like properties, could help develop new therapeutic targets.

Spindle and kinetochore-associated (SKA) complex subunit 1 (SKA1), SKA complex subunit 2 (SKA2), and SKA complex subunit 3 (SKA3) compose the SKA complex, which plays a vital role in the process of precise chromosome separation in cell mitosis (6,7). This key complex stabilizes the kinetochore-microtubule interface through interactions with the Ndc80 complex, which is comprised of Nuf2, Spc24, and Spc25 (8,9). Welburn showed that the SKA complex binds directly to microtubules (10). Recently, numerous studies have been conducted on the relationship between the SKA complex and cancer. These researches have shown that the SKA complex is related to tumor metastasis and proliferation in various kinds of cancers (11-15). Further, the altered expression of the SKA complex has been shown to mediate various cellular processes in cancer and be associated with poor prognosis of cancer (16,17); however, its role in cancer stem cell-like properties has not yet been explored.

We systematically investigated the role of the whole SKA complex in HCC stemness and its potential mechanism. After examining The Cancer Genome Atlas (TCGA) data sets, we found that each subunit of the SKA complex was up-regulated in HCC, and that SKA1 and SKA3 (but not SKA2) were associated with the poor overall survival of HCC patients. We further examined the effects of SKA1 and SKA3 on the proliferation and migration in HCC. Our results revealed that SKA3 promoted HCC stem cell-like properties through the activation of the Notch signaling pathway.

This article is presented following the ARRIVE reporting checklist (available at https://dx.doi.org/10.21037/atm-21-1572).

Methods

Clinical HCC samples

All of the samples (primary HCC and their corresponding adjacent non-HCC tissues) were collected from patients undergoing surgical resection at the Tongji Hospital, Huazhong University of Science and Technology (HUST, Wuhan, China). None of the patients received radiotherapy or chemotherapy before surgery. All the samples were histopathologically diagnosed as primary HCC. A small part of the tumorous and adjacent tissues was fixed with 4% paraformaldehyde for each pair of samples and then embedded in paraffin for immunohistochemistry (IHC) analysis. The rest were stored in an ultra-low temperature refrigerator for later use. Informed consent was obtained from the patients before surgery. The study was conducted following the Declaration of Helsinki (as revised in 2013). The collection and use of human samples were approved by the Ethics Committee of Tongji Hospital, HUST, Wuhan, China (IRB ID: TJ-IRB20210729).

IHC

After fixing the HCC patients’ tumorous and adjacent tissues with 4% paraformaldehyde, IHC staining with antibodies against SKA3 was performed to detect the protein expression levels. For the mice tumor tissues, IHC staining with antibodies against Ki67 was performed to detect the cell proliferation.

Cell lines and cell culture

Human-hepatic L02 cell line (KCB Cat# KCB 200511YJ, RRID: CVCL_6926), Chang-liver cell line [CLS Cat# 300139/p774_Chang-Liver_(HeLa), RRID: CVCL_0238], and HCC cell lines MHCC-97h (RRID: CVCL_4972) were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). SNU-398 (KCLB Cat# 00398, RRID: CVCL_0077) was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA), and cultured in RPMI 1640 Medium (GIBCO). The other cell lines were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; GIBCO). All medium was supplemented with fetal bovine serum (FBS; 10%, GIBCO). The cells were cultured at 37 °C [5% carbon dioxide (CO2) condition].

Western blot and antibodies

Western blot assay was performed as described previously (18). The primary antibodies were as follows: SKA3 (rabbit polyclonal antibody, Abcam Cat# ab91559, RRID: AB_2049220), Notch1 (Cell Signaling Technology Cat# 3608, RRID: AB_2153354), cleaved Notch1 (Cell Signaling Technology Cat# 4147, RRID: AB_2153348), GLI1 (Cell Signaling Technology Cat# 3538, RRID: AB_1903989), β-catenin (Cell Signaling Technology Cat# 8480, RRID: AB_11127855), Hes1 (Cell Signaling Technology Cat# 11988, RRID: AB_2728766) and Oct4 (Cell Signaling Technology Cat# 2750, RRID:AB_823583). Anti-β-actin (Proteintech, 66009-1-ig, 1:5,000) was used as an internal control. The blot density of the protein bands was quantified by the ImageJ software (NIH, America).

RNA extraction and quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Following the manufacturer’s instructions, the ribonucleic acid (RNA) was extracted using trizol reagent (Invitrogen, Carlsbad, CA, USA). The complementary deoxyribonucleic acid (cDNA) was synthesized by reverse transcription using the Prime Script RT Reagent Kit (Takara, Tokyo, Japan). The qRT-PCR assay was carried out as previously described (19). The differences the between different samples were measured using the 2−∆∆Ct method. The SKA3 primer sequence was as follows: sense strand: 5'-TACACGAGCAAGAAGCCATTAAC-3'; and antisense strand: 5'-GGATACGATGTACCGCTCAAGT-3'. The CD44 primer sequence was as follows: sense strand: 5'-CTGCCGCTTTGCAGGTGTA-3'; and antisense strand: 5'-CATTGTGGGCAAGGTGCTATT-3'.

siRNA, plasmids and lentivirus

The transfection of the small-inhibitory RNA (siRNA) and plasmid was performed using the reagent of Lipofectamine 3000 (Invitrogen, Carlsbad, America) following the instructions of the manufacturer. The sequences of siRNA for SKA3 were as follows: SKA3 siRNA1, 5'-GACCCAGAGTTGTCTAATT-3'; SKA3 siRNA2, 5'-GGTACATCGTATCCCAAGT-3'; SKA3 siRNA3, 5'-GAATCCAGGCTCAATGATA-3'. The human full-length SKA3 gene was acquired and inserted into the pGV plasmid to achieve the over-expression of SKA3. The sequence for lentivirus-based RNA interference (RNAi) targeted SKA3 was as follows: SKA3 shRNA, 5'-GACCCAGAGTTGTCTAATT-3'. The sequence for lentivirus-based RNAi targeted Notch1 was as follows: 5'-GGAGCATGTGTAACATCAA-3'. The knockdown and over-expression of SKA3 were verified by qRT-PCR and Western blot assays.

Colony formation assay

One thousand cells per well were seeded in 6-well plates. About 2 weeks later, clones were fixed with 4% methanol for 20 minutes and dyed with crystal violet solution. A light microscope was used to count the clone (>50 cells) numbers.

Cell counting kit-8 (CCK8) and CCK8 toxic assay

HCC cell proliferation was analyzed CCK8 (Promoter, China). Cells were planted in 96-well plates (1,000 cells/plate) and incubated at 37 °C. Absorption value at 450 nm was measured at 24, 48, 72, and 96 h, respectively. CCK8 was used to detect the sensitivity of the cells to Sorafenib. After cell adhesion in the 96-well plates, they were treated by Sorafenib (Sigma-Aldrich) at different concentrations (2.5, 5, 10 or 20 µM). After incubating for 24 h, CCK8 was used to detect the absorbance value.

Transwell migration and invasion assays

Transwell chambers (8 µm pore size; Millipore, Billerica, MA, USA) were used to detect the migration and invasion abilities of HCC cells. For the invasion assay, 50 mg/L Matrigel (1:8 diluent) was added to coat the Transwell chamber surface. Two hundred µL of the cell suspension without FBS was added to the upper chamber, and 10% FBS culture medium was added as a chemoattractant. The data are shown as mean ± standard deviation (SD).

Sphere formation assay

A single-cell suspension was prepared in serum-free DMEM/F12 medium (cat# 12400–024; GIBCO, Grand Island, NY, USA) with 100 µg/mL streptomycin, 100 IU/mL penicillin, 20 ng/mL human recombinant epidermal growth factor (EGF, cat# PHG0311; GIBCO), 10 ng/mL human recombinant basic fibroblast growth factor (bFGF, cat# PHG0266; GIBCO), 2% B27 supplement (cat# 17504–044; GIBCO, Grand Island, NY, USA), 1% N-2 supplement (cat# 17502–048; GIBCO, Carlsbad, CA, USA), and 1% methyl cellulose (cat# M0262; Sigma-Aldrich). The cells were then seeded in ultra-low attachment 24-well plates (Corning, NY, USA) at a density of 2×103 cells/mL. Spheres more than 100 µm in diameter were counted by a light microscope.

Mouse xenograft tumor assay

The experiments were approved by the Institutional Animal Care and Use Committee of the HUST (ID: TJH-201912001), in compliance with the guidelines on the care and use of animals. We obtained 4-week-old BALB/c nude mice from Beijing Huafukang Biotechnology Company and fed them in pathogen-free conditions. In the subcutaneous xenograft tumor model, 1×106, 1×105 cells were injected subcutaneously. Mice were randomly divided into 4 groups (with 3 mice per group). Mice were monitored for 4 weeks in specific pathogen free (SPF) animal house and examined for the growth of subcutaneous tumors. After being anesthesia, the mice were killed, and the tumors were dissected. Tumor volume was calculated using the following formula: V = LW2/2, where L is the length and W is the width of the tumor. The data are shown as mean ± SD.

Statistical analysis

Results were analyzed by the GraphPad Prism 6.0 statistical software. We performed the logistic regression and Wilcoxon signed-rank test to estimate the correlation between SKA3 and HCC clinical-pathological characteristics. To analyze whether SKA3 expression can be regarded as an independent risk factor, we performed univariate and multivariate Cox regression analysis. For comparisons between two groups, Student’s t-test was used. One-way analysis of variance (ANOVA) test was used for comparisons between more than two groups. All of the analyses were performed using R (http://www.R-project.org, version 3.5.2) or Statistical Product and Service Solutions (SPSS, IBM Corp, Armonk, NY, USA), version 25. The data are expressed as mean ± SD. P<0.05 was regarded as significant difference.

Results

The SKA family was highly expressed in HCC

To learn about the contribution of the SKA family to HCC, we downloaded TCGA transcriptome data sets, including 50 normal liver samples and 374 HCC samples. The clinical data of the 374 HCC samples are showed in Figure S1. According to the analysis, the expression of SKA1, SKA2, and SKA3 in HCC were higher in HCC tumour tissues than normal liver samples (see Figure S2A). Among them, the SKA3 expression level in HCC tumour tissues was significantly higher than in normal tissues, and the log2[fold change (FC)] is 17.55. Additionally, Kaplan-Meier analysis revealed that high levels of SKA1 and SKA3 were associated with the poor overall survival of patients with HCC, but the association between SKA2 and overall survival was not significant (see Figure S2B). By conducting a pretest study, we found that SKA3 had a greater ability to promote the proliferation and metastasis of HCC cells than SKA1 (see Figure S3); thus, we chose SKA3 for further study.

SKA3 was upregulated in HCC and predicted poor prognosis

According to the analysis of TCGA LIHC dataset, compare with normal liver samples, the SKA3 expression was significantly increased in HCC samples (see ). To examine the expression of SKA3 in HCC tissues, we randomly selected 32 pairs of samples and quantified the protein levels for SKA3 in HCC tissues and adjacent normal liver tissues. β-actin was used as the housekeeping gene for confirmation. The protein level of SKA3 in HCC tissues (T) was higher than that in non-tumor tissues (N) (see ). The qRT-PCR revealed that SKA3 was more highly expressed in HCC cell lines than normal liver cell lines (see ). We examine the expression of SKA3 by IHC staining in HCC tissues and adjacent normal liver tissues. The results showed that the expression of SKA3 was overexpressed in HCC tissues compared with adjacent normal liver tissues (see ).

Figure 1

SKA3 was upregulated in HCC and predicted poor prognosis in HCC patients. (A) Bioinformatic analysis of the SKA3 expression level in normal liver samples and HCC samples, including 50 normal liver samples and 374 HCC samples. (B,C) Western blot analysis of SKA3 protein levels in HCC tissues and adjacent non-tumor tissues selected randomly (n=32; P<0.0001, Student’s t-test). β-actin was used as a normalized control. (D) Relative expression levels of SKA3 in HCC cell lines (MHCC-97h, and SNU-398), L02 cells, and Chang-liver cells were measured by using qRT-PCR. (E) Representative images of IHC staining of SKA3 in tumor and adjacent non-tumor tissues (scale bar: 50 µm). (F-H) The relationship between SKA3 expression and grade, stage, tumor size (T) in TCGA. For the statistical analysis (Student’s t-test), ****, P<0.0001. HCC, hepatocellular carcinoma; IHC, immunohistochemistry; qRT-PCR, quantitative reverse transcription-polymerase chain reaction; SKA3, spindle and kinetochore-associated complex subunit 3; TCGA, The Cancer Genome Atlas.

Next, we investigated the correlation between SKA3 expression and clinical-pathological characteristics. The results suggested that a higher SKA3 level was correlated with a higher grade (P=1.723e−10), higher stage (P=0.002), and larger tumor size (T) (P=0.002) (see ). Using the logistic regression analysis, we found that SKA3 level was one of the categorical dependent variables (according to the median value of expression) related to pathological features (see ). The results indicated that HCC with a higher SKA3 level tended to have a worse grade, stage, and T classification. The multivariate and univariate Cox regression analyses revealed that among patients with HCC, SKA3 level was an independent risk factor of overall survival [hazard ratio (HR) =2.07, P=4.38e−06; see ].

Table 1

The expression† of SKA3 related to clinicopathological features (logistic regression)

Clinical features	Total (n)	OR in the SKA3 expression (95% CI)	P value
Age (age >50 vs. ≤50 years)	370	0.52 (0.31–0.86)	0.012
Gender (male vs. female)	371	0.84 (0.54–1.29)	0.420
Grade
(G2 vs. G1)	232	1.64 (0.87–3.19)	0.131
(G3 vs. G1)	177	4.42 (2.26–8.93)	2.095772e−05*
(G4 vs. G1)	67	11.18 (2.60–78.06)	0.003*
Stage (III vs. I)	256	1.97 (1.16–3.36)	0.012*
T
(T2 vs. T1)	275	1.79 (1.08–2.97)	0.024*
(T3 vs. T1)	261	1.97 (1.16–3.38)	0.013*
N (N1 vs. N0)	256	1.00 (0.12–8.44)	1.000
M (M1 vs. M0)	270	0.33 (0.02–2.60)	0.338

*, P<0.05; †, categorical dependent variable, greater or less than the median expression level. CI, confidence interval; OR, odds ratio; SKA3, spindle and kinetochore-associated complex subunit 3.

Table 2

Associations of overall survival with clinicopathological features in TCGA patients (Cox regression)

Clinicopathological features	Univariate analysis			Multivariate analysis
	HR (95% CI)	P value		HR (95% CI)	P value
Age	1.01 (0.99–1.02)	0.591		1.01 (0.99–1.04)	0.15
Gender (male vs. female)	0.78 (0.49–1.25)	0.301		1.03 (0.61–1.74)	0.91
Grade	1.02 (0.75–1.39)	0.914		1.00 (0.72–1.39)	0.99
Stage	1.86 (1.46–2.39)	8.07e–07*		0.91 (0.35–2.37)	0.85
T classification	1.80 (1.43–2.27)	4.73e–07*		1.81 (0.77–4.27)	0.18
M classification	3.85 (1.21–12.18)	0.023*		2.54 (0.64–10.03)	0.18
N classification	2.02 (0.49–8.28)	0.328		2.35 (0.37–14.90)	0.36
SKA3 expression (high vs. low)	2.05 (1.53–2.74)	1.24e−06*		2.07 (1.52–2.82)	4.38e−06*

*, P<0.05. CI, confidence interval; HR, hazard ratio; SKA3, spindle and kinetochore-associated complex subunit 3; TCGA, The Cancer Genome Atlas.

Together, these results indicated that SKA3 expression was significantly up-regulated in HCC samples and was related to poor clinical outcomes in patients with HCC.

SKA3 promoted HCC cell migration and invasion

SKA3 expression was downregulated by siRNA in MHCC-97h and SNU-398 cell lines, and the transfection efficiency was detected by qRT-PCR and Western blot. The results showed that siSKA3-1 had the highest downregulation efficiency (see ). In the following article, we will use siSKA3 instead of siSKA3-1. pGV plasmids were used to upregulate SKA3 expression in MHCC-97h and SNU-398 cell lines, and the transfection efficiency was also detected by qRT-PCR and Western blot (see ).

Figure 2

SKA3 promoted HCC cell migration and invasion. (A,B) The selection of siRNA sequences and the regulation of SKA3 in HCC cell lines. qRT-PCR and Western blot assays were used to detect transfection efficiency. (C,D) Representative images of Transwell migration and invasion in siSKA3 MHCC-97h and SNU-398 cells staining with crystal violet solution (scale bar: 200 µm). (E,F) Representative images of Transwell migration and invasion in SKA3 MHCC-97h and SNU-398 cells staining with crystal violet solution (scale bar: 200 µm). For the statistical analysis (Student’s t-test), **, P<0.01; ***, P<0.001; ****, P<0.0001. HCC, hepatocellular carcinoma; qRT-PCR, quantitative reverse transcription-polymerase chain reaction; SKA3, spindle and kinetochore-associated complex subunit 3.

We carried out Transwell migration and invasion assays to investigate the effects of SKA3 on the migration and invasion ability of HCC cells. The number of migratory cells and invasion cells decreased when SKA3 was downregulated (see ). Conversely, the number of migratory cells and invasion cells increased when SKA3 was upregulated (see ).

SKA3 promoted HCC cell proliferation and protected HCC cells from apoptosis

The colony-forming assay showed that the down-regulation of SKA3 weakened cell proliferation and colony-forming ability (see ). The CCK8 assay showed that suppressing SKA3 decreased cell viability and proliferation ability (see ). Conversely, over-expression of SKA3 had the opposite effects (see ). Thus, the downregulation of SKA3 was found to decrease HCC cell proliferation capacity, while the upregulation of SKA3 was found to increase HCC proliferation ability. Next, we conducted an apoptosis assay to test the effect of SKA3 in vitro. Using annexin V and propidium iodide (PI) staining, we found that the cells downregulating SKA3 had a higher apoptosis rate (see ) and that the upregulation of SKA3 inhibited cell apoptosis (see ). Together, these data indicated that SKA3 promoted HCC cell proliferation and protected HCC cells from apoptosis.

Figure 3

SKA3 promoted HCC cell proliferation and protected HCC cells from apoptosis. (A) Colony formation assay was performed using siSKA3 MHCC-97h and SNU-398 cells. The crystal violet was used for dyeing. (B) CCK8 assay was performed using siSKA3 MHCC-97h and SNU-398 cells. (C) Colony formation assay was performed using SKA3 MHCC-97h and SNU-398 cells. The crystal violet was used for dyeing. (D) CCK8 assay was performed using SKA3 MHCC-97h and SNU-398 cells. (E,F) Apoptosis was assessed by flow cytometry in indicated cells. Cells were treated with annexin V (an indicator of apoptosis) and propidium iodide. For the statistical analysis (Student’s t-test), **, P<0.01; ***, P<0.001; ****, P<0.0001. CCK8, cell counting kit-8; HCC, hepatocellular carcinoma; SKA3, spindle and kinetochore-associated complex subunit 3.

SKA3 promoted stem cell-like properties of HCC cells

Previous research has suggested that SKA3 promotes HCC cell metastasis, invasion, proliferation and inhibit apoptosis (15). To study the relationship between SKA3 and HCC stemness, we analyzed the correlations among SKA3 and stemness markers in TCGA at the GEPIA website (http://gepia.cancer-pku.cn/). The results showed that SKA3 had positive correlations with CD44 and Oct4 in HCC (see ). As CD44 and Oct4 are well-accepted stemness markers (20,21), the above data suggested that SKA3 plays a role in regulating HCC stemness.

Figure 4

SKA3 promoted HCC stem cell-like properties. (A) Correlations among SKA3 and stemness markers in TCGA HCC tumor are denoted with Pearson’s correlation coefficients. (B) Representative images of spheres and histogram analysis in siSKA3 and SKA3 SNU-398 cells (scale bar: 200 µm). (C) siSKA3 cells and SKA3 cells were treated with Sorafenib for 24 h and evaluated by CCK8 toxic assay. (D) The expression of Oct4 were tested by Western blot in indicated cells. β-actin was used as the loading control. (E) MHCC-97h cells were used to establish stable cell lines. (F) The efficiency of tumor formation of LV-shCtl and LV-shSKA3 MHCC-97h cells (number of injected cells: 1×106, 1×105; n=3). (G) Efficiency of tumor formation of LV-Ctl and LV-SKA3 MHCC-97h cells (number of injected cells: 1×106, 1×105; n=3). (H) The expression of Ki67 determined by IHC in tumors from subcutaneous xenografts in the above groups. For the statistical analysis (Student’s t-test), *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. CCK8, cell counting kit-8; HCC, hepatocellular carcinoma; IHC, immunohistochemistry; SKA3, spindle and kinetochore-associated complex subunit 3; TCGA, The Cancer Genome Atlas; TPM, transcript per million.

We then investigated the effects of SKA3 on the stem cell-like properties of HCC. Sphere formation assays were performed to evaluate the self-renewal ability of HCC cells. The over-expression of SKA3 in SNU-398 cells led to more and larger spheres, while the suppression of SKA3 in SNU-398 cells led to the formation of fewer and smaller spheres (see ). Next, different concentrations of Sorafenib were used to treat MHCC-97h and SNU-398 cells for 24 h. The results showed that cells down-regulating SKA3 were less resistant to Sorafenib, while cells up-regulating SKA3 were more resistant to Sorafenib (see ). Western blot and qRT-PCR suggested that inhibiting SKA3 expression suppressed the expression of Oct4 and CD44 in MHCC-97h and SNU-398, while overexpressing SKA3 increased the expression of Oct4 and CD44 (see and Figure S4). Finally, the tumor-initiating capacity of HCC cells was evaluated using a subcutaneous xenograft tumor model in the BALB/c nude mice. MHCC-97h cells were used to establish stable cell lines, shSKA3-MHCC-97h cells and SKA3-MHCC-97h cells with lentivirus infection (see ). The grouping and treatment of animal experiments are shown in . An analysis of 3 mice in each group showed that the knockdown of SKA3 decreased the tumor growth of MHCC-97h cells, while the overexpression of SKA3 increased the tumor growth of MHCC-97h cells, which was further confirmed by IHC staining for Ki67 in the xenograft tumors of different groups (see ). Thus, these results confirmed that SKA3 enhanced the tumorigenic capacity of HCC cells.

Table 3

Characteristics and health status of animals and the treatment to them

Mouse	Weight (g) (4 weeks old)	Microbiological status	Injected cells
1/2/3	12.3/13.1/15.2	Sterile	LV-shCtl (1×106)
4/5/6	13.5/13.9/15.7	Sterile	LV-shSKA3 (1×106)
7/8/9	12.7/15.3/16.1	Sterile	LV-shCtl (1×105)
10/11/12	14.1/14.7/16.9	Sterile	LV-shSKA3 (1×105)
13/14/15	13.9/15.5/16.0	Sterile	LV-Ctl (1×106)
16/17/18	12.9/16.7/17.1	Sterile	LV-SKA3 (1×106)
19/20/21	13.8/15.6/17.7	Sterile	LV-Ctl (1×105)
22/23/24	14.5/17.2/18.0	Sterile	LV-SKA3 (1×105)

SKA3, spindle and kinetochore-associated complex subunit 3.

SKA3 promoted stem cell-like properties through the Notch signaling pathway

Through the previous experiments, we demonstrated that SKA3 displayed marked tumor promoting effects in vitro and in vivo, and promoted stem cell-like properties, such as metastasis, invasion, proliferation, self-renewal, resistance to Sorafenib, and tumorigenic abilities.

To decipher the underlying mechanism of SKA3 in regulating stem cell-like properties, we evaluated 3 major signaling pathways critical for stem cell-like properties; that is, the Hedgehog (22), Notch (23), and Wnt signaling pathways (24). First, we tested the expressions of GLI1, NICD, and β-catenin, which are the key factors in the Hedgehog, Notch and Wnt signaling pathways, respectively. As shows, after the down-regulation of SKA3, the expression of NICD (but not GLI1 and β-catenin) decreased. In TCGA, the regression analysis about HCC clinical-pathology showed that SKA3 was positively related to Notch1 at GEPIA website (see ). As stated above, we examined whether SKA3 promoted HCC stem cell-like properties via the Notch signaling pathway. Western blot suggested that inhibiting SKA3 expression significantly suppressed Notch signaling in MHCC-97h and SNU-398, while overexpressing SKA3 activated Notch signaling (see ). Notably, after we knocked down the expression of Notch1 in SKA3 cells via lentivirus transfection, SKA3-mediated HCC migration, proliferation, self-renewal, and resistance to Sorafenib capacities were remarkably inhibited (see ). Thus, SKA3 was found to promote HCC stem cell-like properties by activating the Notch signaling pathway.

Figure 5

SKA3 activated the Notch signaling pathway in HCC cells. (A) Western blot analysis of featured protein levels of Notch, Hedgehog and Wnt pathways in siSKA3 cells. β-actin was used as the loading control. (B) Correlations among Notch1 and SKA3 in TCGA HCC tumor are denoted with Pearson’s correlation coefficients. (C) The core components of Notch signaling, including Notch1 receptor, NICD and Hes1 were tested in siSKA3 and SKA3 MHCC-97h and SNU-398 cells by Western blot. β-actin was used as a normalized control. HCC, hepatocellular carcinoma; SKA3, spindle and kinetochore-associated complex subunit 3; TCGA, The Cancer Genome Atlas; TPM, transcript per million.

Figure 6

SKA3 promoted the stem cell-like properties through the Notch signaling pathway. We knocked down the expression of Notch1 in SKA3 cells via lentivirus transfection. (A) Transwell migration and invasion assays were performed in indicated cells. The crystal violet was used for dyeing (scale bar: 200 µm). (B) Representative images of colony formation assay in indicated cells. The crystal violet was used for dyeing. (C) Representative images of spheres and histogram analyses in indicated cells (scale bar: 200 µm). (D) CCK8 toxic assay showed that the ability of cell resistance in indicated cells. For the statistical analysis (Student’s t-test), *, P<0.05; **, P<0.01; ***, P<0.001. CCK8, cell counting kit-8; SKA3, spindle and kinetochore-associated complex subunit 3.

Discussion

The stemness of cancer is considered as the “source power” of tumorigenesis and development. There is increasing evidence that due to the existence of liver cancer stem cells, the risk of tumorigenic, metastasis, drug resistance, and recurrence of HCC is increased (25). For example, cancer stem cells expressing high epithelial cell adhesion molecule (EpCAM) are thought to be responsible for the initiation of tumor progression, metastasis, drug resistance, and recurrence (26). Thus, it is necessary to explore the factors and molecular mechanisms that regulate the stem cell-like properties of HCC. In this study, we found that SKA3, one subunit of the SKA complex, was highly expressed in HCC and associated with poor overall survival of HCC patients. We also demonstrated that SKA3 promoted HCC stem cell-like properties of self-renewal, proliferation, migration, invasion, resistance to Sorafenib, and tumorigenic capacities. Finally, our results showed that SKA3 promoted HCC stem cell-like characteristics via the Notch signaling pathway.

The SKA complex, which comprises 3 subunits (i.e., SKA1, SKA2, and SKA3), plays a vital role in cell mitosis. Previous study has shown that the complex is related to various tumors (12,13,16). In this study, we systematically examined the relationship between each subunit of the SKA complex, and the clinical outcomes, and the malignant biological behavior in HCC. Our results indicated that compared to SKA1 and SKA2, SKA3 was more closely related to clinical outcomes and the stem cell-like properties of HCC.

SKA3 plays an important role in cell mitosis and tumour development. Abad found that SKA3 could interact directly with microtubules to ensure timely mitotic progression (27). The current results highlighted the importance of SKA3, as one subunit of the SKA complex, playing a role in spindle checkpoint silencing and the maintenance of chromosome cohesion in mitosis (28-31). SKA3 has been reported in various cancers, and appears to promote cancers’ progression. For example, previous studies have shown that SKA3 promotes cell proliferation and invasion in breast cancer and cervical cancer (32,33). In lung adenocarcinoma, SKA3 was found to promote cell migration through the EGFR-PI3K-Akt axis (34). In addition, Hou et al. showed that SKA3 promoted tumor growth in HCC (15). In this study, we demonstrated that HCC cells overexpressing SKA3 exhibited increased stem cell-like properties of self-renewal, migration, invasion, proliferation, resistance to Sorafenib and tumorigenic capacities.

To decipher the underlying mechanisms of SKA3 in regulating stem cell-like properties, we evaluated 3 major signaling pathways critical for cancer stemness: Hedgehog, Notch, and Wnt signaling pathways. The results showed that the expression of Notch signaling pathway was correlated with SKA3. Notch signaling pathway is a highly conserved signaling pathway, that plays a key role in the proliferation, self-renewal, differentiation, and apoptosis of cancer cells (35). It is considered as a classical stem-cell pathway, which can maintain and promote the stemness of a variety of cancers (36-39). Our previous studies have demonstrated that the Notch signaling pathway plays a crucial role in promoting the stemness properties of liver cancer stem cells (40-42). In this study, we found that the deletion of Notch1 remarkably inhibited the increasing stem cell-like properties by over-expressing SKA3. Thus, the Notch signaling pathway was shown to play a vital role in promoting the HCC stemness, which was consistent with our previous experimental results. However, the specific sites mediating the association between SKA3 and the Notch signaling pathway are not known. Thus, we intend to investigate the mechanism further and identify the molecular cross-talk. Recombination signal binding protein for immunoglobulin kappa J region (RBPJ) is a transcription factor that can activate human Notch1 (43). We found that SKA3 was positively correlated with RBPJ (see Figure S5), which may be meaningful in our mechanism research in the future.

The cancer stem cells may provide a therapeutic opportunity to cure and prevent relapse of cancer. Increasingly studies indicated that stem cell specific markers or signaling pathways contribute to maintain and promote the stemness of cancer. Thus, the selective targeting of specific markers and/or signaling pathways is now thought to be an effective therapeutic strategy. Wang et al. demonstrated that CD44 antibody-targeted liposomal nanoparticles, reduced tumor growth and promoted apoptosis by specifically targeting CD44 (44). γ-secretase inhibitors, a type of Notch inhibitors, have been shown to have antitumor effects and have been subject to clinical trials in cancers (45). In the present study, we showed that SKA3 was positively correlated with CD44 and Notch1. Thus, SKA3, which acted as a regulator of stemness in HCC, might be a potential molecular target for HCC.

Conclusions

In the present study, we systematically examined the relationship between each subunit of the SKA complex, and the clinical outcomes, and the stem cell-like properties of HCC. The results showed that SKA3 promoted HCC stem cell-like properties via the Notch signaling pathway.

The article’s supplementary files as