Regulation of spindle and kinetochore‐associated protein 1 by antitumor miR‐10a‐5p in renal cell carcinoma

Abstract

Analysis of our original microRNA (miRNA) expression signature of patients with advanced renal cell carcinoma (RCC) showed that microRNA‐10a‐5p (miR‐10a‐5p) was significantly downregulated in RCC specimens. The aims of the present study were to investigate the antitumor roles of miR‐10a‐5p and the novel cancer networks regulated by this miRNA in RCC cells. Downregulation of miR‐10a‐5p was confirmed in RCC tissues and RCC tissues from patients treated with tyrosine kinase inhibitors (TKI). Ectopic expression of miR‐10a‐5p in RCC cell lines (786‐O and A498 cells) inhibited cancer cell migration and invasion. Spindle and kinetochore‐associated protein 1 (SKA1) was identified as an antitumor miR‐10a‐5p target by genome‐based approaches, and direct regulation was validated by luciferase reporter assays. Knockdown of SKA1 inhibited cancer cell migration and invasion in RCC cells. Overexpression of SKA1 was observed in RCC tissues and TKI‐treated RCC tissues. Moreover, analysis of The Cancer Genome Atlas database demonstrated that low expression of miR‐10a‐5p and high expression of SKA1 were significantly associated with overall survival in patients with RCC. These findings showed that downregulation of miR‐10a‐5p and overexpression of the SKA1 axis were highly involved in RCC pathogenesis and resistance to TKI treatment in RCC.

Renal cell carcinoma accounts for approximately 2–5% of adult malignancies worldwide, and approximately 350 000 new cases are diagnosed, with over 140 000 deaths, each year. Most cases of RCC (approximately 80%) are classified as ccRCC. In the majority of patients, ccRCC is associated with dysfunction of the VHL gene. Lack of VHL function causes activation of HIF and VEGF pathways in ccRCC cells. The mTOR pathway is also activated by dysregulation of HIF and VEGF pathways in patients with ccRCC.

Based on this information regarding the molecular pathogenesis of ccRCC, molecular targeted therapies for patients with advanced and metastatic RCC have been developed during the past decade. The molecular‐targeted agents sorafenib, sunitinib, pazopanib, axitinib, bevacizumab, and cabozantinib inhibit VEGF and VEGF receptor pathways, and temsirolimus and everolimus inhibit the mTOR pathway; treatment with these agents has resulted in significant benefits to patients with advanced RCC. However, the curative effects of these treatments are limited because cancer cells exhibit activation of several alternative signal cascades and acquire resistance to these treatments during therapeutic processes. Treatment strategies for drug‐resistant cancer cells are limited, and the prognosis of these patients is extremely poor. However, the molecular mechanisms of resistance to molecular‐targeted therapies in RCC cells are still unclear.

miRNAs act as pivotal players that regulate the expression control of protein‐coding/protein‐noncoding RNAs in a sequence‐dependent manner. Notably, a single miRNA can directly control many mRNAs in human cells. Therefore, aberrantly expressed miRNAs can disrupt the tight control of RNA expression in cancer cells. Moreover, dysregulation of miRNAs is deeply involved in cancer cell progression, metastasis, and drug resistance.

In RCC, miRNAs are closely related to the development of cancer, and previous studies have reported the relationships among many miRNAs and RCC. For example, the miR‐200 family, containing miR‐141, miR200a/b/c, and miR‐429, forms two clusters, miR‐200a/200b/429 and miR‐141/200c, and expression of the miR‐200 family is markedly downregulated in RCC tissues. Additionally, the miR‐200 family has been reported to be involved in the EMT in several cancers, and miR‐141 and miR‐200c function as tumor suppressors in RCC by inhibiting the EMT through targeting of ZFHX1B, a transcriptional repressor for CDH1/E‐cadherin. In this way, investigation of molecular networks based on miRNAs may help to elucidate the molecular mechanisms mediating the progression of RCC.

Analysis of our original miRNA expression signature of TKI failure in patients with RCC showed that antitumor miR‐101 directly regulated ubiquitin‐like with PHD and ring finger domains 1 (UHRF1), which acted as an oncogene in RCC cells. Based on this signature, we focused on miR‐10a‐5p because miR‐10a‐5p is significantly downregulated in TKI‐treated ccRCC compared with primary ccRCC. Moreover, The Cancer Genome Atlas database showed that the overall survival of patients in the low miR‐10a‐5p expression group was significantly shorter than that of patients in the high expression group in ccRCC (P = 0.00991, Fig. a).

Kaplan‐Meier survival curves based on miR‐10a‐5p expression in patients with clear cell renal cell carcinoma (ccRCC), and schematic representation of the chromosomal location of human miR‐10a. (a) Kaplan‐Meier survival curve for overall survival rate based on miR‐10a‐5p expression in patients with ccRCC from The Cancer Genome Atlas (TCGA) database. (b) miR‐10a is located on human chromosome 17q21.32. Mature microRNAs (miRNAs), miR‐10a‐5p (guide strand) and miR‐10a‐3p (passenger strand), are derived from pre‐miR‐10a.

The aims of the present study were to investigate the functional significance of miR‐10a‐5p and to identify the molecular targets regulated by miR‐10a‐5p in ccRCC cells. Our data showed that restoration of mature miR‐10a‐5p inhibited ccRCC cell proliferation, migration, and invasion. Furthermore, our data demonstrated that the SKA1 gene was overexpressed in primary RCC and advanced RCC specimens and was directly regulated by miR‐10a‐5p in RCC cells. These results demonstrated that SKA1 was involved in RCC pathogenesis, implying that SKA1 could be a novel diagnostic and therapeutic target for patients with advanced RCC.

Materials and Methods

ccRCC clinical specimens and cell culture

A total of 15 pairs of ccRCC specimens and adjacent noncancerous specimens were collected from patients who had undergone radical nephrectomy at Chiba University Hospital (Chiba, Japan) from 2012 to 2015. Clinicopathological characteristics of the patients are summarized in Table . Four patients who died of ccRCC after TKI treatment underwent autopsy at Teikyo University Chiba Medical Center Hospital from 2012 to 2016, as summarized in Table . These specimens were staged according to the General Rule for Clinical and Pathological Studies on Renal Cell Carcinoma based on the American Joint Committee on Cancer (AJCC)‐UICC TNM classification. Written consent for tissue donation for research purposes was obtained from each patient before tissue collection. We used two human ccRCC cell lines (786‐O and A498) obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) as previously described.

Characteristics of primary ccRCC clinical specimens

No.	Age (years)	Sex	Pathology	Grade	pT	INF	v	ly	e.g or ig	fc	im	rc	rp	s
1	71	F	Clear cell	G2	T1a	a	0	0	e.g	1	0	0	0	0
2	74	M	Clear cell	G1>G2	T1a	a	0	0	e.g	1	0	0	0	0
3	59	M	Clear cell	G3>G2	T1b	a	0	0	e.g	1	0	0	0	0
4	79	M	Clear cell	G2>G3>G1	T1a	a	0	0	e.g	1	0	0	0	0
5	52	M	Clear cell	G2>G3	T1b	a	0	0	e.g	1	1	0	0	0
6	64	M	Clear cell	G2>G3>G1	T3a	b	1	0	ig	0	1	1	0	0
7	67	M	Clear cell	G2>G3>G1	T3a	b	1	0	ig	1	0	0	0	0
8	59	M	Clear cell	G3	T3a	b	1	0	ig	0	0	0	0	0
9	73	M	Clear cell	G1>>G3	T2a	a	0	1	e.g	1	0	0	0	0
10	77	M	Clear cell	G1>G2	T1b	a	0	0	e.g	1	0	0	0	0
11	51	F	Clear cell	G2>G1>G3	T3a	b	1	0	ig	0	0	0	0	0
12	84	F	Clear cell	G2	T1a	a	0	0	e.g	0	0	0	0	0
13	78	M	Clear cell	G2>G1≫G3	T1b	b	0	0	e.g	1	0	0	0	0
14	44	M	Clear cell	G2>G1	T1a	b	0	0	e.g	1	0	0	0	0
15	57	M	Clear cell	G2	T1b	a	0	0	e.g	0	0	0	0	0

ccRCC, clear cell renal cell carcinoma; e.g, expansive growth; fc, capsular formation; ig, infiltrative growth; im, intrarenal metastasis; INF, infiltration; ly, lymph node; rc, renal capsule invasion; rp, pelvis invasion; s, sinus invasion; v, vein.

Characteristics of ccRCC autopsy specimens after TKI treatment

Patient	Specimen no.	Location	Age (years)	Stage at diagnosis	Histological type	Grade	Treatment	Treatment duration (months)	Pathological feature of autopsy	Survival from diagnosis (months)
Stage	cT	cN	cM
A	1	Kidney	69	IV	4	2	1	Clear cell carcinoma	3	Sunitinib Temsirolimus	8.5	Multiple lung metastasis Bone metastasis	9.1
2	Lymph node
3	Liver
4	Tumor emboli
B	5	Kidney	80	III	3c	0	0	Clear cell carcinoma	3	Sunitinib	0.7	IVC tumor emboli	1.8
6	Tumor emboli
C	7	Mesenterium	62	I	1b	0	0	Clear cell carcinoma with spindle cell carcinoma	3	Sunitinib Axitinib	34	Multiple bone metastasis Pleural metastasis Lung metastasis Para‐aorta lymph node metastasis	43
8	Lymph node
9	Pleura
D	10	Kidney	69	IV	X	2	1	Clear cell carcinoma	3	Pazopanib	2	Multiple lung metastasis Multiple bone metastasis Kidney metastasis Adrenal metastasis Pleural metastasis Skin metastasis	5.1
11	Lymph node
12	Lymph node
13	Adrenal gland
14	Skin
15	Pleura

ccRCC, clear cell renal cell carcinoma; TKI, tyrosine kinase inhibitor.

RNA extraction

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol, as described previously. RNA quality was confirmed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

Quantitative real‐time reverse transcription polymerase chain reaction (qRT‐PCR)

The procedure for PCR quantification was described previously. Expression levels of miR‐10a‐5p (Assay ID: 000387) were analyzed by TaqMan qRT‐PCR (TaqMan MicroRNA Assay; Applied Biosystems, Foster City, CA, USA) and normalized to the expression of RNU48 (assay ID: 001006; Applied Biosystems). TaqMan probes and primers for SKA1 (P/N: Hs00536843_m1; Applied Biosystems), GAPDH (internal control; P/N: Hs02758991_m1; Applied Biosystems), and GUSB (internal control; P/N: Hs00939627_ml; Applied Biosystems) were assay‐on‐demand gene expression products.

Transfections with mature miRNA, siRNA, or plasmid vectors

The following mature miRNA species were used in this study: mature miRNA and Pre‐miR miRNA Precursor (hsa‐miR‐10a‐5p; P/N: AM17100; Applied Biosystems). The following siRNAs were used: Stealth Select RNAi siRNA, si‐SKA1 (cat. nos. HSS137392 and HSS137393; Invitrogen), and negative control miRNA/siRNA (P/N: AM17111; Applied Biosystems). SKA1 plasmid vectors were designed and provided by ORIGENE (cat. no. RC202370; Rockville, MD, USA). miRNAs and siRNAs were incubated with Opti‐MEM (Invitrogen) and Lipofectamine RNAiMax transfection reagent (Invitrogen), as previously described. Plasmid vectors were incubated with Opti‐MEM and Lipofectamine 3000 reagent (Invitrogen) by forward transfection following the manufacturer's protocol.

Cell proliferation, migration, and invasion assays

786‐O and A498 cells were transfected with 10 nM miRNAs or siRNAs by reverse transfection. Cell proliferation was determined by XTT assays using a Cell Proliferation Kit II (Sigma‐Aldrich, St Louis, MO, USA). Cell migration was evaluated with wound healing assays. Cell invasion was analyzed using modified Boyden chambers containing Transwell‐precoated Matrigel membrane filter inserts (cat. no. 354480; BD Biosciences, Bedford, MA, USA). These assays were carried out as described previously. All experiments were carried out in triplicate.

Selection of putative target genes regulated by miR‐10a‐5p in ccRCC cells

To identify miR‐10a‐5p target genes, we used in silico analyses and genome‐wide gene expression analyses, as described previously. We used the TargetScanHuman 7.1 (June, 2016 release, http://www.targetscan.org/vert_71) database, TCGA, and OncoLnc (http://www.oncolnc.org/) to select and narrow down putative miRNA target genes. An oligo microarray (Human Ge 60K; Agilent Technologies) was used for gene expression analysis. The microarray data were deposited into the GEO database (https://www.ncbi.nlm.nih.gov/geo/; accession number: GSE93290).

Western blotting

Cells were harvested 48 h after transfection, and lysates were prepared. Immunoblotting was carried out with rabbit anti‐SKA1 antibodies (1:1,000 dilution, SAB2701430; Sigma‐Aldrich), anti‐AKT antibody (1:1000, #4691; Cell Signaling Technology, Danvers, MA, USA), anti‐p‐AKT antibody (1:1000, #4060; Cell Signaling Technology), anti‐ERK1/2 antibody (1:1000, #4695; Cell Signaling Technology), anti‐p‐ERK1/2 antibody (1:2000, #4370; Cell Signaling Technology), anti‐FAK antibody (1:1000, #3285; Cell Signaling Technology) and anti‐p‐FAK antibody (1:1000, #8556; Cell Signaling Technology), anti‐SRC antibody (1:1000, #2123; Cell Signaling Technology) and anti‐p‐SRC antibody (1:1000, #6943; Cell Signaling Technology). Anti‐glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) antibodies (1:10000, ab8245; Abcam, Cambridge, UK) were used as an internal loading control. Experimental procedures were carried out as described previously. Protein expression was quantified by using NIH‐ImageJ.

Plasmid construction and dual‐luciferase reporter assays

The partial wild‐type sequence of the SKA1 3′‐UTR or that with deletion of the miR‐10a‐5p target site was inserted between the XhoI‐PmeI restriction sites in the 3′‐UTR of the hRluc gene in the psiCHECK‐2 vector (C8021; Promega, Madison, WI, USA). The procedures were described previously.

Immunohistochemistry

Tissue specimens were incubated overnight at 4°C with anti‐SKA1 antibodies (1:500 dilution, SAB2701430; Sigma‐Aldrich). The slides were treated with biotinylated goat antibodies (Histofine SAB‐PO kit; Nichirei, Tokyo, Japan). The procedures were described previously.

TCGA database analysis of ccRCC

To explore the clinical significance of miRNAs and target genes, we used the TCGA database. Gene expression and clinical data were retrieved from cBioportal (http://www.cbioportal.org/) and OncoLnc (data downloaded on April 30, 2017). We selected high and low target genes expression groups defined by the median value, and data were analyzed by Kaplan‐Meier survival curves and log‐rank statistics.

Statistical analysis

Relationships between two groups and the numerical values obtained by qRT‐PCR were analyzed using Mann‐Whitney U‐tests and paired t‐tests. Spearman's rank test was used to evaluate the correlation between the expression levels of miR‐10a‐5p and SKA1. Relationships among more than three variables and numerical values were analyzed using Bonferroni‐adjusted Mann‐Whitney U‐tests. Survival analysis was carried out using the Kaplan–Meier method and log‐rank tests with JMP software (version 13; SAS Institute Inc., Cary, NC, USA); all other analyses were carried out using Expert StatView (version 5; SAS Institute Inc.).

Results

Expression levels of miR‐10a‐5p in ccRCC specimens and cell lines

The public miRNA database (miRbase: release 21) showed that miR‐10a‐5p was located on chromosome 17q21.32. The mature sequence of miR‐10a‐5p was found to be 5′‐UACCCUGUAGAUCCGAAUUUGUG‐3′ (Fig. b). We evaluated the expression of miR‐10a‐5p in clinical kidney specimens (noncancerous tissues, ccRCC tissues, and autopsy specimens of ccRCC) and cell lines. Expression levels of miR‐10a‐5p were significantly downregulated in primary cancer tissues and TKI‐treated tissues compared with those in noncancerous tissues (P = 0.0010, P = 0.0009, respectively; Fig. a). In 786‐O and A498 cells, expression levels of miR‐10a‐5p were relatively low compared with those of clinical specimens (Fig. a).

Expression levels of miR‐10a‐5p in clear cell renal cell carcinoma (ccRCC) clinical specimens and functional significance of miR‐10a‐5p in ccRCC cells. (a) Expression levels of miR‐10a‐5p in ccRCC clinical specimens and cell lines determined using qRT‐PCR. RNU48 was used as an internal control. TKI, tyrosine kinase inhibitor. (b) Cell proliferation was determined by XTT assays 72 h after transfection with 10 nM miR‐10a‐5p. *P < 0.0001, **P < 0.01. (c) Cell migration activity was assessed by wound‐healing assays 48 h after transfection with 10 nM miR‐10a‐5p. *P < 0.0001. (d) Cell invasion activity was characterized by invasion assays 48 h after transfection with 10 nM miR‐10a‐5p. *P < 0.0001.

Effects of restoring miR‐10a‐5p on cell proliferation, migration, and invasion activities in ccRCC cell lines

To investigate the functional efficacy of miR‐10a‐5p, we carried out gain‐of‐function studies using miRNA transfection into 786‐O and A498 cells. XTT assays showed that cell proliferation was significantly inhibited in miR‐10a‐5p transfectants compared with that in mock or miR‐control transfectants (Figs b; S1a). Migration assays showed that cell migration activity was significantly inhibited in miR‐10a‐5p transfectants in comparison with those in mock or miR‐control transfectants (Figs c; S2a). Similarly, Matrigel invasion assays showed that cell invasion activity was significantly inhibited in miR‐10a‐5p transfectants in comparison with those in mock or miR‐control transfectants (Figs d; S2b).

Identification of candidate genes regulated by miR‐10a‐5p in ccRCC cells

To further elucidate the molecular mechanisms and pathways regulated by antitumor miR‐10a‐5p in ccRCC cells, we carried out a combination of in silico analyses and oligo microarray analyses using miR‐10a‐5p transfectants. The strategy for selection of miR‐10a‐5p target genes is shown in Figure . First, we used TargetScanHuman 7.1 database and identified that 3661 genes had putative target sites for miR‐10a‐5p in their 3′‐UTR. Next, we paired down the 3661 genes based on gene expression data (GEO database accession number: GSE93290), and 31 genes were markedly downregulated after transfection of 786‐O and A498 cells with miR‐10a‐5p (fold‐change log₂ < −2.0; Table). Finally, we checked the expression status of these 31 genes and the clinical significance of ccRCC using the OncoLnc database (http://www.oncolnc.org/). Kaplan‐Meier survival curves showed that high expression of eight genes was associated with poor prognosis in ccRCC (Table; Figs a, ). Among these genes, we focused on SKA1, which was most downregulated by transfection of miR‐10a‐5p, and showed the most dramatic difference in OncoLnc prognostic analysis (P = 2.72E−08, Fig. a).

Identification of miR‐10a‐5p target genes. Flow chart of the strategy for identification of miR‐10a‐5p target genes.

Candidate target genes regulated by miR‐10a‐5p in clear cell renal cell carcinoma (ccRCC)

Gene Symbol	Gene name	Location	No. conserved sites	No. poorly conserved sites	Log2 ratio (A498)	Log2 ratio (786‐O)	Log2 ratio (average)	TCGA‐KIRC OncoLnc 25:25 P‐value
SKA1	Spindle and kinetochore‐associated protein 1	18q21.1	0	1	−5.06	−2.01	−3.54	2.72E−08
SLC7A1	Solute carrier family 7 (cationic amino acid transporter, y+ system), member 1	13q12.3	0	1	−3.38	−2.23	−2.81	2.62E−01
SUMF1	Sulfatase modifying factor 1	3p26.1	0	1	−2.35	−2.65	−2.50	9.42E−01
P4HB	Prolyl 4‐hydroxylase, beta polypeptide	17q25.3	0	1	−2.61	−2.35	−2.48	2.75E−04
USP46	Ubiquitin specific peptidase 46	4q12	1	1	−1.96	−2.9	−2.43	6.62E−02
ELOVL2	ELOVL fatty acid elongase 2	6p24.2	1	0	−2.25	−2.39	−2.32	6.39E−04
KCTD13	Potassium channel tetramerization domain containing 13	16p11.2	0	1	−2.69	−1.72	−2.21	4.00E−06
LINC00908	Long intergenic non‐protein coding RNA 908	18q23	0	1	−2.62	−1.79	−2.21	1.52E−01
SUN2	Sad1 and UNC84 domain containing 2	22q13.1	0	3	−2.10	−2.31	−2.21	1.31E−01
FCF1	FCF1 rRNA‐processing protein	14q24.3	0	1	−2.51	−1.84	−2.18	8.74E−04
ARG2	Arginase 2	14q24.1	0	1	−2.06	−2.25	−2.16	6.91E−01
SLAMF7	SLAM family member 7	1q23.3	0	1	−1.52	−2.77	−2.15	1.23E−01
PRKAA2	Protein kinase, AMP‐activated, alpha 2 catalytic subunit	1p32.2	1	3	−1.66	−2.58	−2.12	2.10E−09
FANCC	Fanconi anemia, complementation group C	9q22.32	0	3	−2.23	−2.00	−2.12	4.29E−03
PDCL	Phosducin‐like	9q33.2	0	1	−2.03	−2.03	−2.03	1.19E−03
PSIP1	PC4 and SFRS1 interacting protein 1	9p22.3	0	1	−2.51	−1.53	−2.02	4.96E−02
CD3D	CD3d molecule, delta (CD3‐TCR complex)	11q23.3	0	2	−2.20	−1.82	−2.01	9.89E−02
C1QL4	Complement component 1, q subcomponent‐like 4	12q13.12	0	1	−1.97	−2.04	−2.01	7.02E−03
SDC1	syndecan 1	2p24.1	1	0	−1.94	−2.05	−2.00	6.76E−01
DNAL4	Dynein, axonemal, light chain 4	22q13.1	0	1	−1.91	−2.06	−1.99	7.34E−01
NUP62CL	Nucleoporin 62kDa C‐terminal like	Xq22.3	0	2	−1.7	−2.03	−1.87	4.12E−02
MTUS1	Microtubule associated tumor suppressor 1	8p22	0	1	−1.93	−1.76	−1.85	4.03E−03
RTN4R	Reticulon 4 receptor	22q11.21	0	1	−2.07	−1.6	−1.84	1.24E−05
PM20D2	Peptidase M20 domain containing 2	6q15	0	1	−2.04	−1.56	−1.80	2.92E−02
PCBD1	Pterin‐4 alpha‐carbinolamine dehydratase/dimerization cofactor of hepatocyte nuclear factor 1 alpha	10q22.1	0	1	−1.55	−2.03	−1.79	4.00E−01
APAF1	Apoptotic peptidase activating factor 1	12q23.1	0	3	−1.90	−1.63	−1.77	1.18E−03
DOHH	Deoxyhypusine hydroxylase/monooxygenase	19p13.3	0	2	−1.78	−1.56	−1.67	6.16E−03
AIM1	Absent in melanoma 1	6q21	0	1	−1.71	−1.55	−1.63	2.67E−01
ANXA7	Annexin A7	10q22.2	1	0	−1.50	−1.65	−1.58	5.17E−02
HS3ST1	Heparan sulfate (glucosamine) 3‐O‐sulfotransferase 1	4p15.33	0	1	−1.54	−1.56	−1.55	2.99E−03
IMPAD1	Inositol monophosphatase domain containing 1	8q12.1	0	1	−1.52	−1.56	−1.54	5.59E−03

⁴Kaplan–Meier survival analysis P < 0.05. Poor prognosis with high expression in

ccRCC.

Kaplan‐Meier survival curves based on SKA1 expression in patients with clear cell renal cell carcinoma (ccRCC), and expression levels of SKA1 in ccRCC clinical specimens. (a) Kaplan‐Meier survival curve for overall survival rate based on SKA1 expression in patients with ccRCC. (b) Expression levels of SKA1 in ccRCC clinical specimens and cell lines. GUSB was used as an internal control. TKI, tyrosine kinase inhibitor. (c) Negative correlation between miR‐10a‐5p and SKA1.

Kaplan‐Meier survival curves for overall survival rates based on expression of seven genes, excluding SKA1, in patients with clear cell renal cell carcinoma (ccRCC).

Expression of SKA1 in ccRCC clinical specimens

A total of 15 pairs of primary ccRCC specimens, adjacent noncancerous specimens, and TKI‐treatment failure autopsy specimens from 15 sites were used for the expression analysis of SKA1 by qRT‐PCR. Expression of SKA1 was significantly upregulated in primary cancer tissues compared with that in normal tissues (P = 0.0011; Fig. b) and was significantly higher in autopsy specimens than in primary cancer tissues (P = 0.0011; Fig. b). Additionally, Spearman's rank test indicated a negative correlation between the expression levels of miR‐10a‐5p and SKA1 (Fig. c).

Furthermore, to analyze SKA1 protein expression, immunohistochemistry was carried out with a ccRCC tissue microarray (cat. no. KD806; US Biomax, Rockville, MD, USA) and autopsy specimens after TKI treatment (Patient D, Table). Patient characteristics for samples used in the tissue microarray are as described in http://www.biomax.us/tissue-arrays/Kidney/KD806. SKA1 was strongly expressed in primary ccRCC tissues and autopsy specimens after TKI treatment compared with that in normal kidney (Fig.).

Expression of SKA1 in clinical clear cell renal cell carcinoma (ccRCC) specimens using a tissue microarray and autopsy tissues. Representative immunohistochemical staining for SKA1 in a ccRCC tissue microarray (cat. no. KD806; US Biomax, Inc., Rockville, MD, USA) and autopsy specimens after tyrosine kinase inhibitor (TKI) treatment (Patient D, Table ). SKA1 was strongly expressed in ccRCC tissues. (a) Normal kidney. (b) ccRCC tissues. (c) ccRCC autopsy tissues after TKI treatment.

SKA1 was directly regulated by miR‐10a‐5p transfection in ccRCC cells

We carried out qRT‐PCR and western blotting to validate whether restoration of miR‐10a‐5p in 786‐O and A498 cells reduced the expression of SKA1. Expression of SKA1 mRNA was significantly suppressed by miR‐10a‐5p transfection compared with that in mock‐ or miR‐control‐transfected cells (Fig. a). Similarly, SKA1 protein expression was repressed in the miR‐10a‐5p transfectants (Fig. b).

Direct regulation of SKA1 by miR‐10a‐5p in clear cell renal cell carcinoma (ccRCC) cells. (a) SKA1 mRNA expression was evaluated using qRT‐PCR in 786‐O and A498 cells 48 h after transfection with miR‐10a‐5p. GAPDH was used as an internal control. *P < 0.0001. (b) SKA1 protein expression was evaluated by western blotting in 786‐O and A498 cells 72 h after transfection with miR‐10a‐5p. GAPDH was used as a loading control. (c) miR‐10a‐5p binding site in the 3′‐UTR of SKA1 mRNA. Dual luciferase reporter assays in 786‐O using vectors encoding the putative miR‐10a‐5p target site of SKA1 3′‐UTR (positions 28‐35). Data were normalized by expression ratios of Renilla/firefly luciferase activities. *P < 0.0001.

Next, we carried out luciferase reporter assays to determine whether SKA1 mRNA had a functional target site. The TargetScan database predicted that miR‐10a‐5p bound at position 28–35 in the 3′‐UTR of SKA1. We used vectors encoding a partial wild‐type sequence of the 3′‐UTR of SKA1 mRNA, including the predicted miR‐10a‐5p target site, or a vector lacking the miR‐10a‐5p target site. Luminescence intensity was significantly reduced by cotransfection with miR‐10a‐5p and the vector carrying the wild‐type 3′‐UTR of SKA1. However, luminescence intensity was not suppressed when the target site of miR‐10a‐5p was deleted from the vectors (Fig. c).

Effects of silencing SKA1 in ccRCC cell lines

To examine the functional significance of SKA1, we carried out loss‐of‐function studies using si‐SKA1 transfectants. First, we evaluated the knockdown efficiency of si‐SKA1 transfection in 786‐O and A498 cells. In this study, we used two types of si‐SKA1 (si‐SKA1‐1 and si‐SKA1‐2). qRT‐PCR and western blotting analyses showed that transfection with both siRNAs effectively downregulated SKA1 mRNA and SKA1 protein expression in 786‐O and A498 cells (Fig. a,b). Furthermore, functional assays indicated that si‐SKA1 transfection markedly inhibited cell proliferation, migration, and invasion in comparison with mock‐ or si‐control‐transfected cells (Figs c; S1b, S3a, S4a).

Effects of SKA1 silencing on clear cell renal cell carcinoma (ccRCC) cell lines. (a) SKA1 mRNA expression was evaluated using qRT‐PCR analysis of 786‐O and A498 cells 48 h after transfection with si‐SKA1‐1 or si‐SKA1‐2. GAPDH was used as an internal control. *P < 0.0001. (b) SKA1 protein expression was evaluated by western blotting analysis of 786‐O and A498 cells 72 h after transfection with miR‐10a‐5p. GAPDH was used as a loading control. (c) Cell proliferation was determined using XTT assays 72 h after transfection with 10 nM si‐SKA1‐1 or si‐SKA1‐2. *P < 0.0001. (d) Cell migration activity was assessed by wound‐healing assays 48 h after transfection with 10 nM si‐SKA1‐1 or si‐SKA1‐2. *P < 0.0001. (e) Cell invasion activity was characterized by invasion assays 48 h after transfection with 10 nM si‐SKA1‐1 or si‐SKA1‐2. *P < 0.0001.

Effects of cotransfection of SKA1/miR‐10a‐5p in 786‐O cells

To validate whether the molecular pathway of SKA1/miR‐10a‐5p was critical for the progression of ccRCC, we carried out SKA1 rescue experiments by cotransfection with SKA1 and miR‐10a‐5p in 786‐O cells. SKA1 protein expression by Western blotting analysis is shown in Figure a. Functional assays showed that the migration and invasion abilities of ccRCC cells were recovered by SKA1 and miR‐10a‐5p transfection compared with cells with restored miR‐10a‐5p only (Figs b–d; S1c, S5a,b). These results supported that SKA1 affected the aggressiveness of ccRCC cells.

Effects of cotransfection of SKA1/miR‐10a‐5p in 786‐O cells. (a) SKA1 protein expression was evaluated by western blotting analysis of 786‐O cells 72 h after reverse transfection with miR‐10a‐5p and 48 h after forward transfection with the SKA1 vector. GAPDH was used as a loading control. (b) Cell proliferation was determined using XTT assays 72 h after reverse transfection with miR‐10a‐5p and 48 h after forward transfection with the SKA1 vector. **P < 0.01. (c) Cell migration activity was assessed by wound‐healing assays 48 h after reverse transfection with miR‐10a‐5p and 24 h after forward transfection with the SKA1 vector. *P < 0.0001. (d) Cell invasion activity was characterized by invasion assays 48 h after reverse transfection with miR‐10a‐5p and 48 h after forward transfection with SKA1 vector. *P < 0.0001.

Clinical significance of SKA1 in ccRCC

To explore the clinical significance of SKA1 in ccRCC, we analyzed Kaplan‐Meier curves of DFS rates according to the expression level of SKA1, and the relationships among SKA1 expression and cancer stage, tumor stage, and histological grade in ccRCC were evaluated using the TCGA‐KIRC database. Kaplan‐Meier curves for DFS rates showed that the DFS of the high SKA1 expression group was significantly shorter than that of the low expression group in ccRCC (P < 0.0001, Fig. a). Additionally, expression levels of SKA1 were significantly increased in cases of advanced disease stage, advanced T stage, and advanced histological grade (Fig. b–d). These analyses suggested that SKA1 affected disease progression and malignancy in ccRCC. Similarly, TCGA data analysis results of clinical significance for the other seven genes are shown in Figures S6–S9.

Kaplan‐Meier survival curve based on SKA1 expression in patients with clear cell renal cell carcinoma (ccRCC), and expression levels of SKA1 according to TNM stage, T stage, and histological grade. (a) Kaplan‐Meier survival curves for disease‐free survival rate based on SKA1 expression in patients with ccRCC. (b‐d) Expression levels of SKA1 were significantly increased in cases of advanced TNM stage, advanced T stage, and advanced histological grade. *P < 0.01, **P < 0.001, ***P < 0.0001.

Effects of SKA1 downstream signaling by si‐SKA1 knockdown or miR‐10a‐5p restoration in RCC cells

We investigated the downstream signals of miR‐10a‐5p/SKA1 axis in 786‐O cells using mature miR‐10a‐5p or si‐SKA1 transfectants. To explore the downstream survival pathways of miR‐10a‐5p/SKA1 axis, phosphorylation of ERK1/2 (Thr 202/Tyr 204), AKT (Ser 473), FAK (Tyr 397) and SRC (Tyr 416) was examined. Knockdown of SKA1 or restoration of miR‐10a‐5p markedly reduced the phosphorylation of ERK1/2, AKT, FAK and SRC (Fig. ).

Effects of the gene encoding SKA1 protein on downstream signaling. Knockdown of SKA1 and restoration of miR‐10a‐5p in 786‐O cells reduced the phosphorylation of ERK1/2, AKT, FAK and SRC. GAPDH was used as a loading control.

Discussion

Based on the underlying molecular oncogenic mechanisms of RCC, several molecular targeted agents have been developed to improve the prognosis of patients with advanced RCC. However, although almost all patients with RCC respond to initial treatment with molecular targeted therapies, cancer cells ultimately become resistant to these treatments. Several molecular mechanisms of drug resistance have been reported in RCC; however, all of these mechanisms are not sufficient to explain the observed changes in cancer cells. Understanding these molecular mechanisms using current genomic approaches is the first step to overcoming drug resistance in RCC cells.

To investigate the molecular mechanisms of drug resistance in RCC cells, we constructed a miRNA expression signature using autopsy specimens from patients with ccRCC who exhibited TKI‐treatment failure. Our present data demonstrated that miR‐10a‐5p acted as an antitumor miRNA in RCC cells. TCGA database analyses indicated that downregulation of miR‐10a‐5p was associated with poor prognosis in patients with RCC. miR‐10a belongs to the miR‐10 family together with miR‐10b, and miR‐10a‐5p is a guide strand of miR‐10a. Previous studies have indicated that miR‐10a is deregulated in several types of cancers. In different types of cancer, miR‐10a has dual functions as either a cancer‐promoting or cancer‐suppressing miRNA. In RCC, a recent study showed that miR‐10a‐5p is a predictor of progression and survival for RCC; however, the functional significance and relevant molecular mechanisms of miR‐10a‐5p in cancer progression are still unknown.

One of the main challenges in miRNA studies is identification of miRNA‐target genes and RNA networks mediated by antitumor miRNAs in cancer cells. A total of 31 genes were identified as putative targets of miR‐10a‐5p regulation in ccRCC cells in this study. Among them, eight genes (SKA1, P4HB, ELOVL2, KCTD13, RTN4R, APAF1, ANXA7, and IMPAD1) were associated with poor prognosis in patients with RCC by TCGA analyses. We focused on the SKA1 gene because overexpression of this gene showed the most significant association with poor prognosis in RCC. Another seven genes may also be involved in the pathology of RCC. Analyses of these genes are important for elucidating the molecular mechanisms of RCC oncogenesis, metastasis and drug resistance.

It is well known that one mRNA was regulated by a number of miRNAs. TargetScan database searching showed that SKA1 was a putative target for miR‐10b‐5p, miR‐24‐3p and miR‐23b‐3p which were significantly downregulated in our miRNA signature of TKI treatment of RCC specimens. Our previous studies demonstrated that miR‐24‐3p and miR‐23b‐3p acted as antitumor miRNAs targeting several oncogenic genes. Also, miR‐10b‐5p functioned as an antitumor miRNA in several types of cancers including RCC. We made the following hypothesis from these findings, SKA1 is regulated by several antitumor miRNAs which contribute to RCC pathogenesis and drug resistance. From the analyses of these miRNAs, it can be expected to lead to elucidation of the mechanism of drug resistance of RCC.

Several studies have indicated that SKA1 is involved in the growth and proliferation of various types of cancer, including oral adenosquamous and hepatocellular carcinoma, bladder cancer, gastric cancer, prostate cancer, thyroid cancer, non‐small cell lung cancer, and glioblastoma. In the human genome, SKA1, SKA2, and SKA3 form the SKA complex. During mitosis, the SKA complex is localized between the outer kinetochore interface and the spindle microtubules. This complex is indispensable for stabilizing adhesion of spindle microtubules to kinetochores and maintaining the metaphase plate; therefore, this complex is critical for appropriate chromosome segregation during mitosis. Interestingly, the expression of SKA1 contributes to cisplatin resistance in lung cancer cells by protecting the cells from cisplatin‐induced apoptosis. Knockdown of SKA1 decreases the activation of extracellular signal‐regulated kinase (ERK1/2) and AKT‐mediated signaling pathways in lung cancer cells. Another study also demonstrated that knockdown of SKA1 alleviated the activation of ERK1/2 and AKT in bladder cancer cells. In adenoid cystic carcinoma, knockdown of SKA1 inhibited cell proliferation, invasion, and migration, and cell cycle arrest by regulating cell cycle‐promoting genes and the matrix metalloproteinase‐9 gene. These findings suggested that the expression of SKA1 may be induced by cancer‐promoting genes and could contribute to cancer cell aggressiveness and drug resistance. Many reports have demonstrated that acquired resistance of RCC cells to molecular targeted therapies induces cancer‐promoting genes and activates several alternative pathways. A previous study showed that sunitinib treatment significantly suppressed phosphorylation of ERK1/2 and AKT in TKI‐sensitive RCC cells, whereas inhibition of phosphorylation was not observed in TKI‐resistant RCC cells. In this study, phosphorylation of ERK1/2 and AKT was suppressed by knockdown of SKA1 and restoration of miR‐10a‐5p. This suggests that downregulation of miR‐10a‐5p and overexpression of the SKA1 axis may be involved in resistance to VEGF‐ and mTOR‐targeted treatments in RCC.

In conclusion, downregulation of miR‐10a‐5p was detected in the miRNA signature of TKI‐failure RCC and acted as an antitumor miRNA in RCC cells. To the best of our knowledge, this is the first study showing that antitumor miR‐10a‐5p directly regulated SKA1 in RCC cells. Overexpression of SKA1 was observed in primary and TKI‐failure RCC specimens. Moreover, downregulation of miR‐10a‐5p and overexpression of SKA1 were associated with poor prognosis in patients with RCC. Elucidation of miR‐10a‐5p/SKA1‐mediated molecular networks may improve our understanding of the pathogenesis of primary RCC and molecular targeted treatment failure in RCC and facilitate the development of new treatment strategies.

Disclosure Statement

Authors declare no conflicts of interest.

Abbreviations

• ccRCC
• clear cell renal cell carcinoma
• DFS
• disease‐free survival
• EMT
• epithelial‐to‐mesenchymal transition
• HIF
• hypoxia‐inducible factor
• miRNA
• microRNA
• RCC
• renal cell carcinoma
• SKA1
• spindle and kinetochore‐associated protein 1
• TKI
• tyrosine kinase inhibitor
• VEGF
• vascular endothelial growth factor
• VHL
• von Hippel‐Lindau