Regulation of UHRF1 by dual-strand tumor-suppressor microRNA-145 (miR-145-5p and miR-145-3p): inhibition of bladder cancer cell aggressiveness

Abstract

In microRNA (miRNA) biogenesis, the guide-strand of miRNA integrates into the RNA induced silencing complex (RISC), whereas the passenger-strand is inactivated through degradation. Analysis of our miRNA expression signature of bladder cancer (BC) by deep-sequencing revealed that microRNA (miR)-145-5p (guide-strand) and miR-145-3p (passenger-strand) were significantly downregulated in BC tissues. It is well known that miR-145-5p functions as a tumor suppressor in several types of cancer. However, the impact of miR-145-3p on cancer cells is still ambiguous. The aim of the present study was to investigate the functional significance of miR-145-3p and BC oncogenic pathways and targets regulated by miR-145-5p/miR-145-3p. Ectopic expression of either miR-145-5p or miR-145-3p in BC cells significantly suppressed cancer cell growth, migration and invasion and it also induced apoptosis. The gene encoding ubiquitin-like with PHD and ring finger domains 1 (UHRF1) was a direct target of these miRNAs. Silencing of UHRF1 induced apoptosis and inhibited cancer cell proliferation, migration, and invasion in BC cells. In addition, overexpressed UHRF1 was confirmed in BC clinical specimens, and the high UHRF1 expression group showed a significantly poorer cause specific survival rate in comparison with the low expression group. Taken together, our present data demonstrated that both strands of miR-145 (miR-145-5p: guide-strand and miR-145-3p: passenger-strand) play pivotal roles in BC cells by regulating UHRF1. The identification of the molecular target of a tumor suppressive miRNAs provides novel insights into the potential mechanisms of BC oncogenesis and suggests novel therapeutic strategies.

INTRODUCTION

In 2012, more than 400,000 new cases of bladder cancer (BC) were diagnosed and 165,000 patients died worldwide [1]. As for the prevalence of BC, men are three times more frequently diagnosed with BC than women [2]. The reasons for this disparity between sexes are not fully understood. BC is pathologically classified into two groups: non-muscle-invasive BC (NMIBC) and muscle-invasive BC (MIBC). Most BC patients (approximately 50%–80%) are diagnosed with NMIBC and this disease can be treated by removing the tumor by transurethral approaches [3]. In NMIBC, disease may recur, and some patients (approximately 25%) progress to MIBC [3]. Patients with advanced BC are generally treated with combination chemotherapy (gemcitabine and cisplatin), but progression-free survival is of limited duration [4]. Therefore, it is important to elucidate the molecular mechanisms of recurrence and invasiveness of BC cells to develop new treatment strategies.

The discovery of non-coding RNA in the human genome changed approaches in cancer research [5, 6]. Molecular mechanisms of post transcriptional gene regulation by protein-coding RNA/non-coding RNA networks are being studied on a genome-wide scale. MicroRNA (miRNA) is a class of small non-coding RNAs, and they are known to be involved in the repression or degradation of target RNA transcripts in a sequence-dependent manner [7]. A single miRNA can regulate thousands of target transcripts, and more than 60% of protein-coding genes may be influenced by miRNAs [8, 9]. Accumulating evidence indicates that aberrantly expressed miRNAs disturb normally regulated RNA networks, leading to pathologic responses in cancer cells [6]. Strategies to identify aberrant expression of miRNA-mediated cancer pathways are being developed as a new direction in cancer research in the post genome sequencing era.

To seek out differentially expressed miRNAs in BC cells, we used BC clinical specimens to establish deep sequencing-based miRNA expression signatures [10]. In general, the guide-strand RNA from duplex miRNA is retained to direct recruitment of the RNA induced silencing complex (RISC) to target messenger RNAs, whereas the passenger-strand RNA is degraded [11–13]. Recently, we revealed that both strands of microRNA (miR)-144-5p and miR-144-3p derived from pre-miR-144 acted as tumor suppressors in BC cells [14]. Moreover, miR-144-5p (passenger-strand) directly targeted cyclin E1 and E2 in BC cells, suggesting that the passenger-strand of miRNA has a physiological role in cells [14].

In this study, we focused on miR-145-5p and miR-145-3p because these miRNAs were significantly downregulated in BC cells as determined in our deep sequencing signature [10]. It is well known that miR- 145- 5p functions as a tumor suppressor in several types of cancer, including BC [15]. However, the role of miR-145-3p on cancer cells is still ambiguous. The aims of the present study were to investigate the anti-tumor effects of miR-145-3p as well as miR-145-5p, and to determine the BC oncogenic pathways and target genes regulated by these miRNAs. The discovery that miR- 145- 5p and miR-145-3p coordinately regulate pathways and targets provides new insight into the mechanisms of BC progression and metastasis.

RESULTS

The expression levels of miR-145-5p and miR-145-3p in BC specimens and cell lines

We evaluated the expression levels of miR-145-5p and miR-145-3p in BC tissues (n = 69), normal bladder epithelia (NBE) (n = 12), and two BC cell lines (T24 and BOY). The expression levels of miR-145-5p and miR- 145- 3p were significantly lower in tumor tissues and BC cell lines compared with NBE (Figure 1A). Spearman’s rank test showed a positive correlation between the expression of these miRNAs (r = 0.986 and P < 0.0001) (Figure 1B). On the other hand, there were no significant relationships between any of the clinicopathological parameters (i.e., tumor grade, stage, metastasis, or survival rate) and the expression levels of miR-145-5p and miR-145-3p (data not shown).

Figure 1: The expression levels of miR-145-5p and miR-145-3p, and their effects in BC cells. (A) Expression levels of miR- 145- 5p and miR-145-3p in clinical specimens and BC cell lines were determined by qRT-PCR. Data were normalized to RNU48 expression. (B) Correlation of miR-145-5p and miR-145-3p expression. (C) Cell growth was determined by XTT assays 72 hours after transfection with 10 nM miR-145-5p or miR-145-3p. *P < 0.0001. (D) Cell migration activity was determined by the wound-healing assays. *P < 0.0001. (E) Cell invasion activity was determined using Matrigel invasion assays. *P < 0.0001.

Effect of restoring miR-145-5p or miR-145- 3p expression on cell growth, migration, and invasion in BC cell lines

We performed gain-of-function studies using transfection of these miRNAs to investigate their functional roles. XTT, cell migration, and invasion assays demonstrated that cell proliferation, cell migration, and cell invasion were significantly inhibited in miR-145-5p and miR-145-3p transfectants in comparison with mock or miR-control transfectants (each P < 0.0001, Figure 1C, 1D, and 1E). These results suggested that miR-145-3p as well as miR-145-5p could have a tumor suppressive function in BC cells.

To investigate the synergistic effects of miR- 145- 5p and miR-145-3p, we performed proliferation, migration, and invasion assays with co-transfection of miR- 145-5p and miR-145-3p in BC cells (T24 and BOY), but they did not show synergistic effects of these miRNAs transfection (Supplementary Figure 1).

Effects of miR-145-5p and miR-145-3p transfection on apoptosis and cell cycle in BC cell lines

Because miR-145-5p and miR-145-3p transfection strongly inhibited cell proliferation in BC cell lines, we hypothesized that these miRNAs may induce apoptosis. Hence, we performed flow cytometric analyses to determine the number of apoptotic cells following restoration of miR- 145-5p or miR-145-3p expression.

The apoptotic cell numbers (apoptotic and early apoptotic cells) were significantly larger in miR-145-5p or miR-145-3p transfectants than in mock or miR-control transfectants (Figure 2A and 2C). Western blot analyses showed that cleaved PARP expression was significantly increased in miR-145-5p or miR-145-3p transfectants compared with mock or miR-control transfectants (Figure 2B and 2D).

Figure 2: Effects of miR-145-5p and miR-145-3p on apoptosis. (A, C) Apoptosis assays were carried out using flow cytometry. Early apoptotic cells are in area R4 and apoptotic cells are in area R2. The normalized ratios of apoptotic cells are shown in the histograms. Cycloheximide (2 μg/mL) was used as positive control. *P = 0.0266 and **P < 0.0001. (B, D) Western blot analyses for apoptotic markers (cleaved PARP) in BC cell lines. GAPDH was used as a loading control.

We also investigated the cell cycle assays using miR-145-5p and miR-145-3p transfectants. The fraction of cells in the G2/M phase was significantly larger in miR-145-5p and miR-145-3p transfectants in T24 cells in comparison with mock or miR-control transfectants (Supplementary Figure 2). In contrast, miR-145-5p and miR-145-3p transfection induced cell cycle arrest at the G1 phase in BOY cells (Supplementary Figure 2). The reason why the cell cycle arrest (G2 arrest in T24 and G1 arrest in BOY) varies according to a cell types is a future problem.

Identification of common target genes regulated by miR-145-5p and miR-145-3p in BC cells

To gain further insight into the molecular mechanisms and pathways regulated by tumor suppressive miR-145-5p and miR-145-3p in BC cells, we used a combination of in silico analyses and gene expression analyses. Figure 3 shows our strategy to narrow down the common target genes of miR-145-5p and miR-145-3p.

Figure 3: Flow chart illustrates the strategy for analysis of miR-145-5p and miR-145-3p target genes. A total of 4,555 and 6,295 downregulated genes in expression analysis of miR-145-5p and miR-145-3p transfected BC cell lines, respectively, (T24 and BOY) were selected as putative target genes. Next we merged the data of those selected genes and the microRNA.org database. The analyses showed 398 common putative target genes between miR-145-5p and miR-145-3p. We then analyzed gene expression with available GEO data sets (GSE11783 + GSE31684). The analyses showed that 79 genes were significantly upregulated in BC specimens compared with NBE.

In gene expression analyses, a total of 4,555 and 6,295 genes were downregulated in miR-145-5p and miR- 145-3p transfectants, respectively, in comparison with control transfectants (Gene Expression Omnibus (GEO), accession number: GSE66498). Of those downregulated genes, 1,735 and 1,680 genes, respectively, had putative binding sites for miR-145-5p and miR-145- 3p in their 3′ untranslated regions (UTRs) according to the microRNA.org database. We found that there were 398 common genes targeted by both miRNAs, and among them, we ultimately identified 79 genes that were upregulated in the clinical BC samples from the GEO (accession numbers: GSE11783, GSE31684) (Table 1). We subsequently focused on the ubiquitin-like with PHD and ring finger domains 1 (UHRF1) gene because it was the top ranked gene in the list.

Table 1: Highly expressed genes putatively regulated by miR-145-5p and miR-145-3p

UHRF1 was a direct target of miR-145-5p and miR-145-3p in BC cells

We performed quantitative real-time RT-PCR (qRT-PCR) to validate that miR-145-5p and miR-145-3p repressed UHRF1 mRNA expression in BC cell lines, and we did indeed observe that it was significantly reduced in transfectants of these miRNAs in comparison with mock or miR-control transfectants (P < 0.0001 and P = 0.0036, Figure 4A). The protein expression levels of UHRF1 were also repressed in the miRNAs transfectants (Figure 4B).

Figure 4: Direct regulation of UHRF1 by miR-145-5p and miR-145-3p. (A) UHRF1 mRNA expression was evaluated by qRT-PCR in T24 and BOY 72 hours after transfection with miR-145-5p and miR-145-3p. GUSB was used as an internal control. *P = 0.0036 and **P < 0.0001. (B) UHRF1 protein expression was evaluated by Western blot analyses in T24 and BOY 72–96 hours after transfection with miR-145-5p or miR-145-3p. GAPDH was used as a loading control. (C) miR-145-5p and miR-145-3p binding sites in the 3′ UTR of UHRF1 mRNA. Dual Luciferase reporter assays using vectors encoding putative miR-145-5p and miR-145-3p target sites of the UHRF 3′ UTR (positions 1,179–1,198 and 287–292, respectively) for both wild-type and deleted regions. Normalized data were calculated as ratios of Renilla/firefly luciferase activities. *P < 0.0001.

We carried out dual luciferase reporter assays in T24 and BOY cells to determine whether the UHRF1 gene was directly regulated by miR-145-5p/3p. The microRNA.org database predicted that there was one binding site for miR- 145-5p in the 3′ UTR of UHRF1 (position 1,179– 1,198); for miR-145-3p, there was a binding site in the 3′ UTR at position 287–292. We used vectors encoding the partial wild-type sequence of the 3′ UTR of the mRNA, including the predicted miR-145-5p or miR-145- 3p target sites. We found that the luminescence intensity was significantly reduced by co-transfection with these miRNAs and the vector carrying the wild-type 3′ UTR, whereas no reduction of luminescence was observed by transfection with the deletion vector (binding site had been removed) (P < 0.0001, Figure 4C). These suggested that either of miR-145-5p and miR-145-3p were directly bounded to specific sites in the 3′ UTR of UHRF1 mRNA.

Effects of silencing UHRF1 in BC cell lines

To investigate the functional role of UHRF1 in BC cells, we carried out loss-of-function studies by using si-UHRF1 transfectants. First, we evaluated the knockdown efficiency of si-UHRF1 transfection in BC cell lines. In the present study, we used two types of si- UHRF1 (si- UHRF1-1 and si-UHRF1-2). The qRT- PCR and Western blot analyses showed that both siRNAs effectively downregulated UHRF1 expression in both cell lines (Figure 5A and 5B).

Figure 5: UHRF1 mRNA and protein expression after si-UHRF1 transfection and effects of UHRF1 silencing in BC cell lines. (A) UHRF1 mRNA expression was evaluated by qRT-PCR in T24 and BOY 72 hours after transfection with si-UHRF1-1 and si-UHRF1-2. GUSB was used as an internal control. (B) UHRF1 protein expression was evaluated by Western blot analysis in T24 and BOY 72 - 96 hours after transfection with miR-145-5p or miR-145-3p. GAPDH was used as a loading control. (C) Cell proliferation was determined with the XTT assays 72 hours after transfection with 10 nM si-UHRF1-1 or si-UHRF1-2. *P < 0.0001. (D) Cell migration activity was determined by wound-healing assays. *P < 0.0001. (E) Cell invasion activity was determined using Matrigel invasion assays. *P < 0.0001.

XTT, cell migration, and invasion assays demonstrated that cell proliferation, cell migration, and cell invasion were inhibited in si-UHRF1 transfectants in comparison with the mock or siRNA-control transfectant cells (each P < 0.0001, Figure 5C, 5D, and 5E).

In the apoptosis assays, the apoptotic cell numbers were significantly greater in si-UHRF1 transfectants than in mock or siRNA-control transfectants (Figure 6A and 6C). Western blot analyses showed that cleaved PARP expression was significantly increased in si-UHRF1 transfectants compared with mock or siRNA-control transfectants (Figure 6B and 6D).

Figure 6: Effects of silencing UHRF1 on apoptosis in BC cell lines. (A, C) Apoptosis assays were carried out using flow cytometry. Early apoptotic cells are in area R4 and apoptotic cells are in area R2. The normalized ratios of the apoptotic cells are shown in the histogram. Cycloheximide (2 μg/mL) was used as a positive control. *P < 0.0001 (B, D) Western blot analyses for apoptotic markers (cleaved PARP) in BC cell lines. GAPDH was used as a loading control.

Expression of UHRF1 in BC clinical specimens

The qRT-PCR analyses showed that the expression level of UHRF1 mRNA was significantly upregulated in 69 BC specimens and 2 BC cell lines compared with 12 NBE (P < 0.0001, Figure 7A). Spearman’s rank test showed negative correlations between miR-145-5p/miR-145-3p expression and UHRF1 mRNA expression (r = –0.324 and –0.298, P = 0.0024 and 0.0051, Figure 7B). As shown in Figure 7C, the expression level of UHRF1 was significantly greater in high grade clinical BCs (P = 0.0135), MIBCs (T2 ≤) (P = 0.0379), BCs with positive lymph node invasion (N1) (P = 0.00182), and in BCs with positive distant metastasis (M1) (P = 0.0307) than in their counterparts. Kaplan-Meier analysis showed that the high UHRF1 expression group had significantly lower cause specific survival probabilities compared to the low UHRF1 expression group (P = 0.0259, Figure 8).

Figure 7: The expression level of UHRF1 mRNA in BC clinical specimens and cell lines, and association of UHRF1 expression with clinicopathological parameters. (A) Expression levels of UHRF1 in clinical specimens and BC cell lines were determined by qRT-PCR. Data were normalized to GUSB expression. (B) The correlated expression among miR-145-5p, miR-145-3p, and UHRF1. (C) Association of UHRF1 expression with clinicopathological parameters. Relationships between two variables were analyzed using the Mann-Whitney U test.

Figure 8: The association between the expression level of UHRF1 and cause specific survival rate. Kaplan-Meier survival curves for cause specific survival rates based on UHRF1 expression in 57 BC patients. P-values were calculated using the log-rank test.

We validated the expression status of UHRF1 in BC clinical specimens using immunohistochemical staining. UHRF1 was expressed moderately or strongly in several cancer lesions, and normal bladder tissues stained weakly (Figure 9).

Figure 9: Immunohistochemical staining of UHRF1 in BC clinical specimens. UHRF1 was expressed more strongly in several cancer lesions than in noncancerous tissues. Left panel, original magnification ×40; Right panel, original magnification ×200. (A) Positively stained tumor lesion (High grade, T2bN0M0), (B) Positively stained tumor lesion (High grade, T1N0M0), (C) Positively stained tumor lesion (Low grade, T3N0M0), (D) Negative staining in normal bladder tissue.

Investigation of downstream genes regulated by UHRF1 in BC cells

To identify the downstream genes regulated by UHRF1, genome-wide gene expression analyses and in silico analyses were performed in two BC cell lines transfected with si-UHRF1. A total of 533 genes were downregulated (log₂ FC < –1.5) by si-UHRF1 transfection, and a total of 704 genes were upregulated (log₂ FC > 1.0) by si-UHRF1 transfection compared with negative control cells (GEO, accession number: GSE77790). Among the downregulated genes in the si-UHRF1 transfectants, 104 genes were upregulated in the BC clinical samples from GEO database (accession numbers: GSE11783, GSE31684), whereas among the upregulated genes, 62 genes were downregulated in the clinical BCs. These results imply that the 104 upregulated genes may act as oncogenes, and the 62 downregulated genes may act as tumor suppressors downstream from UHRF1 in BC (Tables 2 and 3).

Table 2: Significantly downregulated genes by si-UHRF1 in BC cell lines

Entrez Gene ID	Gene Symbol	Description	Genomic location	Gene Expression Omnibus (GSE11783 + GSE31684)			Expression in si-UHRF1 transfectant (Log2 FC)
				Expression	Log2FC	P-value	T24	BOY
7153	TOP2A	topoisomerase (DNA) II alpha 170kDa	17q21.2	up	6.312	1.049E-03	–1.880	–1.681
29128	UHRF1	ubiquitin-like with PHD and ring finger domains 1	19p13.3	up	4.984	1.049E-03	–3.213	–2.907
259266	ASPM	asp (abnormal spindle) homolog, microcephaly associated (Drosophila)	1q31.3	up	4.299	1.049E-03	–3.431	–3.444
332	BIRC5	baculoviral IAP repeat containing 5	17q25.3	up	4.110	1.049E-03	–2.258	–1.777
9928	KIF14	kinesin family member 14	1q32.1	up	3.866	1.049E-03	–3.294	–1.544
1063	CENPF	centromere protein F, 350/400kDa	1q41	up	3.576	1.049E-03	–2.613	–3.307
1894	ECT2	epithelial cell transforming 2	3q26.31	up	3.469	1.049E-03	–1.928	–1.813
55247	NEIL3	nei endonuclease VIII-like 3 (E. coli)	4q34.3	up	3.428	1.049E-03	–1.728	–2.065
9401	RECQL4	RecQ protein-like 4	8q24.3	up	3.414	1.049E-03	–1.751	–2.102
3832	KIF11	kinesin family member 11	10q23.33	up	3.356	1.049E-03	–2.299	–1.657
57082	CASC5	cancer susceptibility candidate 5	15q15.1	up	3.230	1.049E-03	–2.470	–2.188
151176	FAM132B	family with sequence similarity 132, member B	2q37.3	up	3.100	1.058E-03	–2.420	–2.184
151246	SGOL2	shugoshin-like 2 (S. pombe)	2q33.1	up	2.694	1.049E-03	–3.124	–2.407
1062	CENPE	centromere protein E, 312kDa	4q24	up	2.689	1.058E-03	–3.676	–3.218
23529	CLCF1	cardiotrophin-like cytokine factor 1	11q13.2	up	2.646	1.049E-03	–1.905	–2.363
81930	KIF18A	kinesin family member 18A	11p14.1	up	2.553	1.049E-03	–3.246	–2.128
7130	TNFAIP6	tumor necrosis factor, alpha-induced protein 6	2q23.3	up	2.531	2.835E-03	–1.795	–2.735
55502	HES6	hes family bHLH transcription factor 6	2q37.3	up	2.506	6.688E-03	–1.572	–1.508
5328	PLAU	plasminogen activator, urokinase	10q22.2	up	2.244	1.740E-03	–2.417	–1.791
9824	ARHGAP11A	Rho GTPase activating protein 11A	15q13.3	up	2.051	2.348E-03	–1.675	–1.613
23057	NMNAT2	nicotinamide nucleotide adenylyltransferase 2	1q25.3	up	2.050	1.247E-03	–1.707	–1.863
59285	CACNG6	calcium channel, voltage-dependent, gamma subunit 6	19q13.42	up	2.016	1.049E-03	–1.502	–1.763
675	BRCA2	breast cancer 2, early onset	13q13.1	up	2.015	1.049E-03	–1.764	–2.356
6524	SLC5A2	solute carrier family 5 (sodium/glucose cotransporter), member 2	16p11.2	up	1.900	1.214E-03	–1.855	–1.569
79412	KREMEN2	kringle containing transmembrane protein 2	16p13.3	up	1.893	1.348E-03	–2.309	–1.796
6274	S100A3	S100 calcium binding protein A3	1q21.3	up	1.825	8.102E-03	–2.215	–1.848
5331	PLCB3	phospholipase C, beta 3 (phosphatidylinositol-specific)	11q13.1	up	1.790	1.049E-03	–2.219	–1.735
55349	CHDH	choline dehydrogenase	3p21.1	up	1.743	1.049E-03	–1.926	–2.008
811	CALR	calreticulin	19p13.2	up	1.652	1.049E-03	–1.554	–1.500
4987	OPRL1	opiate receptor-like 1	20q13.33	up	1.627	2.626E-03	–1.927	–1.766
375248	ANKRD36	ankyrin repeat domain 36	2q11.2	up	1.530	8.102E-03	–3.873	–1.791
441054	C4orf47	chromosome 4 open reading frame 47	4q35.1	up	1.485	2.151E-02	–2.229	–2.522
201475	RAB12	RAB12, member RAS oncogene family	18p11.22	up	1.468	1.058E-03	–2.353	–2.947
286151	FBXO43	F-box protein 43	8q22.2	up	1.463	2.396E-02	–1.528	–2.082
9091	PIGQ	phosphatidylinositol glycan anchor biosynthesis, class Q	16p13.3	up	1.434	3.574E-03	–1.594	–1.693
81575	APOLD1	apolipoprotein L domain containing 1	12p13.1	up	1.354	1.808E-03	–2.237	–2.383
132320	SCLT1	sodium channel and clathrin linker 1	4q28.2	up	1.340	1.049E-03	–3.140	–3.098
100131211	TMEM194B	transmembrane protein 194B	2q32.2	up	1.325	1.049E-03	–1.573	–1.967
153642	ARSK	arylsulfatase family, member K	5q15	up	1.252	1.049E-03	–2.052	–1.875
21	ABCA3	ATP-binding cassette, sub-family A (ABC1), member 3	16p13.3	up	1.170	4.892E-02	–1.879	–1.831
55036	CCDC40	coiled-coil domain containing 40	17q25.3	up	1.160	1.049E-03	–1.562	–1.531
84259	DCUN1D5	DCN1, defective in cullin neddylation 1, domain containing 5	11q22.3	up	1.151	1.247E-03	–1.591	–1.993
80381	CD276	CD276 molecule	15q24.1	up	1.146	1.072E-03	–2.656	–2.096
6487	ST3GAL3	ST3 beta-galactoside alpha-2,3-sialyltransferase 3	1p34.1	up	1.139	1.049E-03	–1.828	–2.380
5351	PLOD1	procollagen-lysine, 2-oxoglutarate 5-dioxygenase 1	1p36.22	up	1.104	2.942E-03	–1.650	–1.570
343099	CCDC18	coiled-coil domain containing 18	1p22.1	up	1.075	1.578E-03	–3.521	–2.428
30818	KCNIP3	Kv channel interacting protein 3, calsenilin	2q11.1	up	1.069	2.723E-03	–3.678	–2.733
10051	SMC4	structural maintenance of chromosomes 4	3q25.33	up	1.066	1.578E-03	–2.612	–1.745
51427	ZNF107	zinc finger protein 107	7q11.21	up	1.040	1.316E-03	–2.527	–2.104
10592	SMC2	structural maintenance of chromosomes 2	9q31.1	up	1.032	6.688E-03	–3.520	–2.180
20	ABCA2	ATP-binding cassette, sub-family A (ABC1), member 2	9q34.3	up	0.965	1.372E-02	–1.511	–2.291
55183	RIF1	replication timing regulatory factor 1	2q23.3	up	0.960	1.058E-03	–1.712	–1.605
9898	UBAP2L	ubiquitin associated protein 2-like	1q21.3	up	0.952	1.049E-03	–1.587	–2.301
29780	PARVB	parvin, beta	22q13.31	up	0.952	1.096E-02	–3.288	–1.888
9585	KIF20B	kinesin family member 20B	10q23.31	up	0.933	5.720E-03	–2.282	–3.122
9534	ZNF254	zinc finger protein 254	19p12	up	0.920	3.863E-03	–2.072	–2.662
57520	HECW2	HECT, C2 and WW domain containing E3 ubiquitin protein ligase 2	2q32.3	up	0.884	3.179E-03	–1.838	–1.958
84083	ZRANB3	zinc finger, RAN-binding domain containing 3	2q21.3	up	0.873	1.578E-03	–1.987	–1.915
6498	SKIL	SKI-like proto-oncogene	3q26.2	up	0.859	1.808E-03	–2.709	–1.845
64770	CCDC14	coiled-coil domain containing 14	3q21.1	up	0.842	6.943E-03	–2.453	–1.711
254065	BRWD3	bromodomain and WD repeat domain containing 3	Xq21.1	up	0.808	1.393E-03	–1.852	–2.546
22973	LAMB2P1	laminin, beta 2 pseudogene 1	3p21.31	up	0.804	7.521E-03	–2.336	–2.311
7525	YES1	YES proto-oncogene 1, Src family tyrosine kinase	18p11.32	up	0.794	2.526E-03	–3.127	–2.099
1984	EIF5A	eukaryotic translation initiation factor 5A	17p13.1	up	0.793	5.486E-03	–2.297	–2.018
22852	ANKRD26	ankyrin repeat domain 26	10p12.1	up	0.787	3.303E-03	–2.798	–2.663
23322	RPGRIP1L	RPGRIP1-like	16q12.2	up	0.778	1.182E-02	–1.517	–1.806
79677	SMC6	structural maintenance of chromosomes 6	2p24.2	up	0.764	8.401E-03	–1.909	–2.083
84920	ALG10	ALG10, alpha-1,2-glucosyltransferase	12p11.1	up	0.763	6.688E-03	–1.828	–2.360
8570	KHSRP	KH-type splicing regulatory protein	19p13.3	up	0.762	3.303E-03	–1.767	–1.820
5819	PVRL2	poliovirus receptor-related 2 (herpesvirus entry mediator B)	19q13.32	up	0.757	9.078E-03	–3.014	–2.465
51575	ESF1	ESF1, nucleolar pre-rRNA processing protein, homolog (S. cerevisiae)	20p12.1	up	0.755	9.430E-03	–1.786	–1.732
51361	HOOK1	hook microtubule-tethering protein 1	1p32.1	up	0.689	3.067E-02	–2.156	–2.000
10198	MPHOSPH9	M-phase phosphoprotein 9	12q24.31	up	0.667	1.947E-03	–2.113	–1.502
4983	OPHN1	oligophrenin 1	Xq12	up	0.632	5.277E-03	–2.278	–1.747
4976	OPA1	optic atrophy 1 (autosomal dominant)	3q29	up	0.619	2.169E-03	–2.190	–1.526
168850	ZNF800	zinc finger protein 800	7q31.33	up	0.611	1.227E-02	–1.807	–1.867
26272	FBXO4	F-box protein 4	5p13.1	up	0.611	3.512E-02	–2.224	–2.445
7390	UROS	uroporphyrinogen III synthase	10q26.13	up	0.605	6.433E-03	–3.120	–2.062
4683	NBN	nibrin	8q21.3	up	0.590	5.720E-03	–2.986	–1.966
79670	ZCCHC6	zinc finger, CCHC domain containing 6	9q21.33	up	0.587	5.486E-03	–2.353	–1.839
79573	TTC13	tetratricopeptide repeat domain 13	1q42.2	up	0.587	6.943E-03	–1.740	–2.064
50840	TAS2R14	taste receptor, type 2, member 14	12p13.2	up	0.574	1.598E-02	–1.947	–1.509
79042	TSEN34	TSEN34 tRNA splicing endonuclease subunit	19q13.42	up	0.570	1.138E-02	–2.455	–1.761
6801	STRN	striatin, calmodulin binding protein	2p22.2	up	0.563	2.723E-03	–1.964	–2.434
3597	IL13RA1	interleukin 13 receptor, alpha 1	Xq24	up	0.552	2.075E-02	–2.460	–2.403
147657	ZNF480	zinc finger protein 480	19q13.41	up	0.547	3.893E-02	–3.434	–3.276
8683	SRSF9	serine/arginine-rich splicing factor 9	12q24.31	up	0.534	1.227E-02	–1.523	–2.098
252983	STXBP4	syntaxin binding protein 4	17q22	up	0.516	2.151E-02	–1.776	–1.599
284325	C19orf54	chromosome 19 open reading frame 54	19q13.2	up	0.510	4.734E-02	–1.614	–2.171
91147	TMEM67	transmembrane protein 67	8q22.1	up	0.509	9.799E-03	–1.647	–2.069
114799	ESCO1	establishment of sister chromatid cohesion N-acetyltransferase 1	18q11.2	up	0.495	4.873E-03	–2.173	–2.401
57670	KIAA1549	KIAA1549	7q34	up	0.480	4.582E-02	–2.127	–1.789
6103	RPGR	retinitis pigmentosa GTPase regulator	Xp11.4	up	0.467	3.290E-02	–1.583	–2.025
5700	PSMC1	proteasome (prosome, macropain) 26S subunit, ATPase, 1	14q32.11	up	0.449	1.274E-02	–1.639	–1.711
253260	RICTOR	RPTOR independent companion of MTOR, complex 2	5p13.1	up	0.442	2.666E-02	–2.458	–1.683
23241	PACS2	phosphofurin acidic cluster sorting protein 2	14q32.33	up	0.442	3.179E-03	–3.416	–2.028
27154	BRPF3	bromodomain and PHD finger containing, 3	6p21.31	up	0.440	5.720E-03	–1.772	–2.598
7703	PCGF2	polycomb group ring finger 2	17q12	up	0.439	2.865E-02	–1.828	–1.974
51105	PHF20L1	PHD finger protein 20-like 1	8q24.22	up	0.383	9.078E-03	–3.492	–2.007
57697	FANCM	Fanconi anemia, complementation group M	14q21.2	up	0.364	3.067E-02	–1.648	–1.627
9730	VPRBP	Vpr (HIV-1) binding protein	3p21.2	up	0.363	2.075E-02	–2.342	–1.568
5378	PMS1	PMS1 postmeiotic segregation increased 1 (S. cerevisiae)	2q32.2	up	0.350	4.734E-02	–2.701	–1.616
255520	ELMOD2	ELMO/CED-12 domain containing 2	4q31.1	up	0.334	4.582E-02	–2.360	–1.637
80124	VCPIP1	valosin containing protein (p97)/p47 complex interacting protein 1	8q13.1	up	0.304	3.893E-02	–3.107	–2.286

Table 3: Significantly upregulated genes by si-UHRF1 in BC cell lines

Entrez Gene ID	Gene Symbol	Description	Genomic location	Gene Expression Omnibus (GSE11783 + GSE31684)			Expression in si-UHRF1 transfectant (Log2 FC)
				Expression	Log2FC	P-value	T24	BOY
3043	HBB	hemoglobin, beta	11p15.4	down	–3.263	1.214E-03	1.204	2.109
137835	TMEM71	transmembrane protein 71	8q24.22	down	–2.428	4.873E-03	2.813	3.920
8639	AOC3	amine oxidase, copper containing 3	17q21.31	down	–2.188	1.434E-03	1.907	3.140
1408	CRY2	cryptochrome circadian clock 2	11p11.2	down	–2.141	1.058E-03	2.134	2.108
7644	ZNF91	zinc finger protein 91	19p12	down	–2.058	1.155E-03	1.435	2.063
197257	LDHD	lactate dehydrogenase D	16q23.1	down	–1.626	2.965E-02	1.844	1.362
316	AOX1	aldehyde oxidase 1	2q33.1	down	–1.601	2.169E-03	1.841	1.049
26051	PPP1R16B	protein phosphatase 1, regulatory subunit 16B	20q11.23	down	–1.547	6.688E-03	1.076	1.198
63976	PRDM16	PR domain containing 16	1p36.32	down	–1.439	2.075E-02	2.639	3.846
254827	NAALADL2	N-acetylated alpha-linked acidic dipeptidase-like 2	3q26.31	down	–1.313	4.873E-03	1.621	3.168
154	ADRB2	adrenoceptor beta 2, surface	5q32	down	–1.242	9.799E-03	2.384	2.302
10477	UBE2E3	ubiquitin-conjugating enzyme E2E 3	2q31.3	down	–1.117	1.135E-03	1.053	2.755
7099	TLR4	toll-like receptor 4	9q33.1	down	–1.053	6.943E-03	1.402	2.356
57478	USP31	ubiquitin specific peptidase 31	16p12.2	down	–1.037	4.169E-03	1.570	1.234
57185	NIPAL3	NIPA-like domain containing 3	1p36.11	down	–0.986	1.316E-03	1.329	1.189
30815	ST6GALNAC6	ST6 (alpha-N-acetyl-neuraminyl-2,3-beta-galactosyl-1,3)-N-acetylgalactosaminide alpha-2,6-sialyltransferase 6	9q34.11	down	–0.936	1.660E-02	1.093	2.348
29915	HCFC2	host cell factor C2	12q23.3	down	–0.928	1.393E-03	1.304	1.296
54741	LEPROT	leptin receptor overlapping transcript	1p31.3	down	–0.893	1.049E-03	1.280	2.248
7779	SLC30A1	solute carrier family 30 (zinc transporter), member 1	1q32.3	down	–0.879	8.736E-03	1.267	1.262
79027	ZNF655	zinc finger protein 655	7q22.1	down	–0.863	1.393E-03	1.570	1.589
64344	HIF3A	hypoxia inducible factor 3, alpha subunit	19q13.32	down	–0.845	1.016E-02	1.284	2.411
79844	ZDHHC11	zinc finger, DHHC-type containing 11	5p15.33	down	–0.834	3.176E-02	1.505	1.890
79815	NIPAL2	NIPA-like domain containing 2	8q22.2	down	–0.825	6.688E-03	1.929	1.259
7923	HSD17B8	hydroxysteroid (17-beta) dehydrogenase 8	6p21.32	down	–0.821	3.512E-02	2.657	3.759
8629	JRK	Jrk homolog (mouse)	8q24.3	down	–0.820	1.740E-03	1.358	2.076
79591	C10orf76	chromosome 10 open reading frame 76	10q24.32	down	–0.812	1.808E-03	1.099	1.917
599	BCL2L2	BCL2-like 2	14q11.2	down	–0.775	2.835E-03	1.384	1.730
412	STS	steroid sulfatase (microsomal), isozyme S	Xp22.31	down	–0.770	1.372E-02	1.440	1.471
56900	TMEM167B	transmembrane protein 167B	1p13.3	down	–0.755	2.626E-03	2.282	2.366
23509	POFUT1	protein O-fucosyltransferase 1	20q11.21	down	–0.747	1.274E-02	1.400	2.132
25923	ATL3	atlastin GTPase 3	11q12.3	down	–0.727	3.290E-02	1.179	1.907
79669	C3orf52	chromosome 3 open reading frame 52	3q13.2	down	–0.708	4.021E-02	1.200	1.482
55844	PPP2R2D	protein phosphatase 2, regulatory subunit B, delta	10q26.3	down	–0.691	2.666E-02	1.422	1.303
5939	RBMS2	RNA binding motif, single stranded interacting protein 2	12q13.3	down	–0.626	5.943E-03	1.193	1.438
6158	RPL28	ribosomal protein L28	19q13.42	down	–0.618	1.808E-03	2.026	3.427
2145	EZH1	enhancer of zeste 1 polycomb repressive complex 2 subunit	17q21.2	down	–0.618	1.393E-03	1.391	1.171
388969	C2orf68	chromosome 2 open reading frame 68	2p11.2	down	–0.611	3.435E-03	1.309	1.192
55422	ZNF331	zinc finger protein 331	19q13.42	down	–0.594	1.725E-02	2.855	2.230
92400	RBM18	RNA binding motif protein 18	9q33.2	down	–0.594	8.401E-03	1.172	2.001
80017	C14orf159	chromosome 14 open reading frame 159	14q32.11	down	–0.590	1.182E-02	1.072	1.748
7556	ZNF10	zinc finger protein 10	12q24.33	down	–0.563	1.480E-02	1.592	1.127
55957	LIN37	lin-37 DREAM MuvB core complex component	19q13.12	down	–0.543	1.857E-02	1.002	1.205
84267	C9orf64	chromosome 9 open reading frame 64	9q21.32	down	–0.543	5.720E-03	1.215	1.299
8799	PEX11B	peroxisomal biogenesis factor 11 beta	1q21.1	down	–0.535	4.679E-03	1.083	1.163
8790	FPGT	fucose-1-phosphate guanylyltransferase	1p31.1	down	–0.524	2.075E-02	1.680	1.222
6992	PPP1R11	protein phosphatase 1, regulatory (inhibitor) subunit 11	6p22.1	down	–0.517	6.433E-03	1.104	1.329
116224	FAM122A	family with sequence similarity 122A	9q21.11	down	–0.507	2.169E-03	1.231	1.549
51710	ZNF44	zinc finger protein 44	19p13.2	down	–0.499	1.372E-02	2.385	1.001
7265	TTC1	tetratricopeptide repeat domain 1	5q33.3	down	–0.487	1.182E-02	1.109	1.112
80213	TM2D3	TM2 domain containing 3	15q26.3	down	–0.485	1.182E-02	1.342	1.742
81631	MAP1LC3B	microtubule-associated protein 1 light chain 3 beta	16q24.2	down	–0.480	1.725E-02	1.210	2.109
6016	RIT1	Ras-like without CAAX 1	1q22	down	–0.473	2.666E-02	1.556	1.432
7247	TSN	translin	2q14.3	down	–0.467	4.582E-02	1.101	1.496
167227	DCP2	decapping mRNA 2	5q22.2	down	–0.447	1.016E-02	1.284	1.104
11046	SLC35D2	solute carrier family 35 (UDP-GlcNAc/UDP-glucose transporter), member D2	9q22.32	down	–0.431	1.227E-02	1.318	1.340
54946	SLC41A3	solute carrier family 41, member 3	3q21.2	down	–0.402	4.294E-02	1.526	1.988
7799	PRDM2	PR domain containing 2, with ZNF domain	1p36.21	down	–0.384	7.805E-03	1.438	1.294
6651	SON	SON DNA binding protein	21q22.11	down	–0.374	5.486E-03	1.126	1.155
80255	SLC35F5	solute carrier family 35, member F5	2q14.1	down	–0.369	4.441E-02	1.143	1.619
55197	RPRD1A	regulation of nuclear pre-mRNA domain containing 1A	18q12.2	down	–0.364	3.893E-02	1.480	1.761
91603	ZNF830	zinc finger protein 830	17q12	down	–0.358	2.075E-02	1.040	1.085
5094	PCBP2	poly(rC) binding protein 2	12q13.13	down	–0.286	4.734E-02	1.454	1.158

To further investigate the UHRF1 downstream genes, we performed the classification of these candidate genes to known molecular pathways by using DAVID program (https://david.ncifcrf.gov/). Classification strategy of downstream genes by si-UHRF1 transfectants is shown in Figure 10A and 10B. Significantly upregulated and downregulated pathways and their involved genes are indicated in Tables 4 and 5. Several genes were classified into biological process categories and a variety of biological pathways, “M phase”, “cell cycle”, and “cell cycle phase” were significantly downregulated by si- UHRF1 transfectants (Table 4).

Figure 10: Flow chart demonstrating the strategy for analysis of genes regulated by UHRF1. (A) A total of 2,222 and 1,512 downregulated genes in expression analyses of si-UHRF1 transfectants of BC cell lines (T24 and BOY, respectively) were selected. We then analyzed 533 common downregulated genes by using available GEO data sets (GSE11783 + GSE31684). The analyses showed that 104 genes were significantly upregulated in BC specimens compared with NBE. (B) A total of 2,665 and 2,434 upregulated genes in expression analysis of si-UHRF1 transfectants of BC cell lines (T24 and BOY, respectively) were selected. We then analyzed 704 common upregulated genes by using GEO data sets. The analyses showed that 62 genes were significantly downregulated in BC specimens compared with NBE.

Table 4: Downregulated genes by si-UHRF1 were classified by DAVID program

Biological process	Number of genes	P-Value	Genes
M phase	15	8.10E-09	ASPM, BIRC5, BRCA2, CENPE, CENPF, FBXO43, KIF11, KIF18A, KIF20B, MPHOSPH9, NBN, SGOL2, SMC2, SMC4, VCPIP1
cell cycle	20	1.10E-07	ASPM, BIRC5, BRCA2, CALR, CENPE, CENPF, ESCO1, FBXO43, KIF11, KIF18A, KIF20B, MPHOSPH9, NBN, PSMC1, RIF1, SGOL2, SMC2, SMC4, UHRF1, VCPIP1
cell cycle phase	15	1.40E-07	ASPM, BIRC5, BRCA2, CENPE, CENPF, FBXO43, KIF11, KIF18A, KIF20B, MPHOSPH9, NBN, SGOL2, SMC2, SMC4, VCPIP1
cell cycle process	17	1.90E-07	ASPM, BIRC5, BRCA2, CALR, CENPE, CENPF, FBXO43, KIF11, KIF18A, KIF20B, MPHOSPH9, NBN, PSMC1, SGOL2, SMC2, SMC4, VCPIP1
chromosome segregation	8	5.20E-07	BIRC5, CENPE, CENPF, KIF18A, SGOL2, SMC2, SMC4, TOP2A
M phase of mitotic cell cycle	11	8.50E-07	ASPM, BIRC5, CENPE, CENPF, KIF11, KIF18A, KIF20B, MPHOSPH9, SMC2, SMC4, VCPIP1
organelle fission	11	1.00E-06	ASPM, BIRC5, CENPE, CENPF, KIF11, KIF18A, KIF20B, OPA1, SMC2, SMC4, VCPIP1
mitosis	10	6.40E-06	ASPM, BIRC5, CENPE, CENPF, KIF11, KIF18A, KIF20B, SMC2, SMC4, VCPIP1
nuclear division	10	6.40E-06	ASPM, BIRC5, CENPE, CENPF, KIF11, KIF18A, KIF20B, SMC2, SMC4, VCPIP1
mitotic cell cycle	12	1.20E-05	ASPM, BIRC5, CENPE, CENPF, KIF11, KIF18A, KIF20B, MPHOSPH9, PSMC1, SMC2, SMC4, VCPIP1
DNA repair	10	4.90E-05	BRCA2, ESCO1, FANCM, NBN, NEIL3, PMS1, RECQL4, SMC6, TOP2A, UHRF1
cell division	10	6.50E-05	ASPM, BIRC5, BRCA2, CENPE, CENPF, KIF11, KIF20B, SGOL2, SMC2, SMC4
response to DNA damage stimulus	11	7.40E-05	BRCA2, ESCO1, FANCM, NBN, NEIL3, PMS1, RECQL4, RIF1, SMC6, TOP2A, UHRF1
establishment of chromosome localization	4	8.90E-05	BIRC5, CENPE, CENPF, KIF18A
chromosome localization	4	8.90E-05	BIRC5, CENPE, CENPF, KIF18A
chromosome organization	12	1.40E-04	BRCA2, BRPF3, CENPE, CENPF, FBXO4, KIF18A, NBN, PCGF2, SGOL2, SMC2, SMC4, TOP2A
DNA metabolic process	12	2.00E-04	BRCA2, CENPF, ESCO1, FANCM, FBXO4, NBN, NEIL3, PMS1, RECQL4, SMC6, TOP2A, UHRF1
microtubule-based movement	6	5.80E-04	CENPE, KIF11, KIF14, KIF18A, KIF20B, OPA1
regulation of cell cycle process	6	6.00E-04	BIRC5, BRCA2, CALR, CENPE, CENPF, KIF20B
microtubule-based process	8	7.90E-04	BRCA2, CENPE, HOOK1, KIF11, KIF14, KIF18A, KIF20B, OPA1
mitotic sister chromatid segregation	4	1.30E-03	CENPE, KIF18A, SMC2, SMC4
sister chromatid segregation	4	1.40E-03	CENPE, KIF18A, SMC2, SMC4
metaphase plate congression	3	1.90E-03	CENPE, CENPF, KIF18A
cellular response to stress	11	2.00E-03	BRCA2, ESCO1, FANCM, NBN, NEIL3, PMS1, RECQL4, RIF1, SMC6, TOP2A, UHRF1
regulation of mitotic cell cycle	6	2.20E-03	BIRC5, BRCA2, CENPE, CENPF, KIF20B, NBN
organelle localization	5	2.20E-03	ASPM, BIRC5, CENPE, CENPF, KIF18A
spindle checkpoint	3	2.20E-03	BIRC5, CENPE, CENPF
positive regulation of cell cycle	4	4.80E-03	BIRC5, BRCA2, CALR, CENPE
establishment of organelle localization	4	8.20E-03	BIRC5, CENPE, CENPF, KIF18A
chromosome condensation	3	9.70E-03	SMC2, SMC4, TOP2A
glucose transport	3	1.30E-02	SLC5A2, STXBP4, YES1
hexose transport	3	1.40E-02	SLC5A2, STXBP4, YES1
regulation of cell cycle	7	1.40E-02	BIRC5, BRCA2, CALR, CENPE, CENPF, KIF20B, NBN
monosaccharide transport	3	1.50E-02	SLC5A2, STXBP4, YES1
negative regulation of neuron differentiation	3	1.70E-02	ASPM, CALR, NBN
cell cycle checkpoint	4	1.70E-02	BIRC5, CENPE, CENPF, NBN
kinetochore assembly	2	1.80E-02	CENPE, CENPF
meiosis	4	2.10E-02	BRCA2, FBXO43, NBN, SGOL2
M phase of meiotic cell cycle	4	2.10E-02	BRCA2, FBXO43, NBN, SGOL2
meiotic cell cycle	4	2.20E-02	BRCA2, FBXO43, NBN, SGOL2
germ cell development	4	2.30E-02	BRCA2, CASC5, HOOK1, PVRL2
kinetochore organization	2	2.40E-02	CENPE, CENPF
DNA recombination	4	2.50E-02	BRCA2, NBN, RECQL4, SMC6
mitotic cell cycle checkpoint	3	2.70E-02	CENPE, CENPF, NBN
centromere complex assembly	2	3.50E-02	CENPE, CENPF
spermatid development	3	4.10E-02	CASC5, HOOK1, PVRL2
regulation of nuclear division	3	4.40E-02	CENPE, CENPF, KIF20B
regulation of mitosis	3	4.40E-02	CENPE, CENPF, KIF20B
negative regulation of macromolecule biosynthetic process	8	4.50E-02	BRCA2, CALR, CD276, CENPF, KCNIP3, PCGF2, SKIL, ZNF254
spermatid differentiation	3	4.60E-02	CASC5, HOOK1, PVRL2
cytoskeleton organization	7	4.60E-02	BRCA2, CALR, HOOK1, KIF11, KIF18A, OPHN1, RICTOR
negative regulation of cellular biosynthetic process	8	5.10E-02	BRCA2, CALR, CD276, CENPF, KCNIP3, PCGF2, SKIL, ZNF254
positive regulation of cellular protein metabolic process	5	5.10E-02	CLCF1, EIF5A, FBXO4, PSMC1, RICTOR
carbohydrate transport	3	5.20E-02	SLC5A2, STXBP4, YES1
mitotic metaphase plate congression	2	5.30E-02	CENPE, KIF18A
regulation of DNA replication	3	5.30E-02	BRCA2, CALR, NBN
double-strand break repair	3	5.30E-02	BRCA2, NBN, RECQL4
negative regulation of biosynthetic process	8	5.50E-02	BRCA2, CALR, CD276, CENPF, KCNIP3, PCGF2, SKIL, ZNF254
positive regulation of protein metabolic process	5	5.80E-02	CLCF1, EIF5A, FBXO4, PSMC1, RICTOR
microtubule cytoskeleton organization	4	5.80E-02	BRCA2, HOOK1, KIF11, KIF18A
negative regulation of mitotic metaphase/anaphase transition	2	6.40E-02	CENPE, CENPF
blastocyst growth	2	6.40E-02	BRCA2, NBN
mitotic cell cycle spindle assembly checkpoint	2	6.40E-02	CENPE, CENPF
positive regulation of mitotic cell cycle	2	7.00E-02	BIRC5, BRCA2
negative regulation of mitosis	2	7.00E-02	CENPE, CENPF
negative regulation of nuclear division	2	7.00E-02	CENPE, CENPF
negative regulation of macromolecule metabolic process	9	7.20E-02	BRCA2, CALR, CD276, CENPF, KCNIP3, PCGF2, PSMC1, SKIL, ZNF254
reproductive cellular process	4	7.30E-02	BRCA2, CASC5, HOOK1, PVRL2
mitotic chromosome condensation	2	7.50E-02	SMC2, SMC4
negative regulation of transcription from RNA polymerase II promoter	5	7.50E-02	CALR, KCNIP3, PCGF2, SKIL, ZNF254
protein localization	10	8.00E-02	CALR, CENPE, CENPF, EIF5A, HOOK1, KIF18A, RAB12, RPGR, SGOL2, STXBP4
negative regulation of nucleobase, nucleoside, nucleotide and nucleic acid metabolic process	7	8.60E-02	BRCA2, CALR, CENPF, KCNIP3, PCGF2, SKIL, ZNF254
establishment of protein localization	9	8.90E-02	CALR, CENPE, CENPF, EIF5A, HOOK1, KIF18A, RAB12, RPGR, STXBP4
in utero embryonic development	4	8.90E-02	BRCA2, NBN, PCGF2, RPGRIP1L
negative regulation of nitrogen compound metabolic process	7	9.10E-02	BRCA2, CALR, CENPF, KCNIP3, PCGF2, SKIL, ZNF254
positive regulation of cellular component organization	4	9.50E-02	CALR, CENPE, EIF5A, RICTOR
developmental growth	3	9.60E-02	BRCA2, NBN, PLAU

Table 5: Upregulated genes by si-UHRF1 were classified by DAVID program

Biological process	Number of genes	P-Value	Genes
regulation of transcription	15	1.40E-02	CRY2, ADRB2, EZH1, HCFC2, HIF3A, JRK, POFUT1, PRDM16, PRDM2, TLR4, ZNF10, ZNF331, ZNF44, ZNF655, ZNF91
regulation of transcription, DNA-dependent	10	7.00E-02	ADRB2, HCFC2, HIF3A, PRDM16, PRDM2, ZNF10, ZNF331, ZNF44, ZNF655, ZNF91
regulation of RNA metabolic process	10	7.90E-02	ADRB2, HCFC2, HIF3A, PRDM16, PRDM2, ZNF10, ZNF331, ZNF44, ZNF655, ZNF91
negative regulation of myeloid leukocyte differentiation	2	4.90E-02	PRDM16, TLR4
fucose metabolic process	2	5.20E-02	POFUT1, FPGT
brown fat cell differentiation	2	6.90E-02	ADRB2, PRDM16
negative regulation of myeloid cell differentiation	2	8.50E-02	PRDM16, TLR4

DISCUSSION

miRNAs are critical regulators of gene expression and they control many physiologic processes in mammalian cells [5–7]. There are abundant evidences that aberrantly expressed miRNAs can dysregulate otherwise well-controlled cellular RNA networks, thereby enhancing cancer cell development, progression, and metastasis [6–9]. The discovery of aberrantly expressed miRNAs and the resultant changes in RNA networks in cancer cells provide novel molecular explanations for cancer cell progression and metastasis. It is now apparent that dysregulated miRNAs play important roles in BC cell development [16]. Our past miRNA studies of BC cells showed that clustered miRNAs (including miR-1/133a (targeting TAGLN2), miR-23b/27b/24-1 (targeting EGFR, MET, and FOXM1), and miR-195/497 (targeting BIRC5 and WNT7A)) act as tumor-suppressive miRNAs through their regulation of several oncogenic genes and pathways [10, 17–19].

Improved technological developments (next generation sequencing) have illuminated the role of miRNA networks in cancer cells. In this study, we examined the expression of miR-145-5p and miR-145 3p in BC cells because these miRNAs were significantly reduced in cancer cells as determined by deep sequencing. Our data demonstrated that miR-145-3p (the passenger-strand from pre-miR-145) had anti-tumor effects through targeting of UHRF1 in BC cells.

Downregulation of miR-145-5p (the guide-strand) is frequently observed in many types of cancer, and past studies have established the anti-tumor function of miR-145-5p through its regulation of several types of oncogenes in cancer cells [15]. Our group also identified the anti-tumor function of miR-145-5p in prostate cancer, renal cell carcinoma, bladder cancer, and esophageal squamous cell carcinoma [20–23]. Importantly, p53 appears to transcriptionally regulate miR-145-5p by interaction with a potential p53 response element at the pre-miR-145 promoter region [24]. Moreover, c-MYC is directly repressed by miR-145-5p, indicating that it acts as a new member of the p53 regulatory network and contributes to the direct linkage between p53 and c-MYC in human cancer pathways [24]. In contrast to miR-145-5p, the functional significance of miR-145-3p in cancer cells has been obscure. This is the first report to evaluate the anti-tumor function of miR-145-3p in BC cells by gain-of-function assays.

miRNAs are often associated in clusters in the genome, and several studies have focused on the functional role of clustered miRNAs in human cancers [17, 18, 20–23, 25]. In the human genome, 429 human miRNAs have been found to be clustered at 144 sites, with inter-miRNA distances of less than 5,000 base pair (miRBase, release 21). Both miR-143 and miR-145-5p are known to be located close together on human chromosome 5q32, where they form a cluster [26]. Based on our miRNA signatures, miR-143 and miR-145-5p are the most frequently downregulated miRNAs in various types of human cancers [26]. These two miRNAs have been reported as tumor suppressors and studied extensively for their role in oncogenic pathways in several cancers [15]. Our past studies demonstrated that hexokinase-2 (HK2) and Golgi membrane protein 1 (GOLM1) were directly regulated by miR-143 and miR-145-5p in renal carcinoma and prostate cancer, respectively [22, 23].

In this study, we speculated that miR-145-5p and miR-145-3p worked together to regulate pathways in BC cell progression and metastasis. Our present data showed that UHRF1 was directly regulated by both miR-145-5p and miR-145-3p in BC cells. In previous studies of miRNA regulation of UHRF1 in cancers, UHRF1 was regulated by miR-146a/146b in gastric cancer [27], miR-9 in colorectal cancer [28], and miR-124 in BC [29]. However, there have been no previous reports about the effects of miR-145-5p and miR-145-3p on UHRF1.

The UHRF1 gene was first cloned as a transcription factor that binds to the promoter region of the topoisomerase IIα (TOP2A) gene and controls its expression levels [30]. UHRF1 is involved in a wide range of physiological and pathological phenomena, including cancer development and metastasis [31]. UHRF1 plays a pivotal role in controlling gene expression through regulating epigenetic mechanisms, including DNA methylation, histone deacetylation, histone methylation, and histone ubiquitination [31]. Overexpression of UHRF1 occurs in many types of cancer, and aberrantly expressed UHRF1 causes cancer cell activation through hyper-methylation of tumor-suppressor genes such as BRCA1, CDKN2A, p73, and RASSF1 [32]. Expression of UHRF1 might be used as a progression marker in cancer [32]. For example, the expression of UHRF1 in MIBC was greater than in NMIBC, and upregulation was associated with an increased risk of progression after transurethral resection [33]. Our present data showed that knockdown of UHRF1 significantly induced apoptosis in BC cells and expression levels of the gene correlated with cause specific survival. Our data support the past studies of UHRF1 in cancer research, suggesting UHRF1 plays essential roles in BC cell progression and might be a molecular target for BC treatment.

In this study, we identified UHRF1-regulated BC pathways by using genome-wide gene expression analysis of si-UHRF1-transfected cells. Our expression data showed that UHRF1 and TOP2A were significantly reduced by si-UHRF1 transfection, indicating the usefulness of the present analytic approach. Our data showed that several anti-apoptosis genes and pro-proliferation genes were involved in pathways downstream of UHRF1, such as BIRC5 and CENPF. BIRC5 is a member of the inhibitor of apoptosis (IAP) family preferentially expressed by many cancers, including BC [10], and its mediated cellular networks are essential for cancer cell proliferation and viability [34]. CENPF is a master regulator of prostate cancer malignancy. Together, FOXM1 and CENPF regulate target gene expression and activation in cancer cells [35, 36]. The identification of these novel molecular pathways and targets mediated by the miR-145-5p/145-3p/UHRF1 axis may lead to a better understanding of BC cell progression and metastasis.

In conclusion, downregulation of dual-strand miR- 145-5p and miR-145-3p was validated in BC clinical specimens, and these miRNAs were shown to function as tumor suppressors in BC cells. To the best of our knowledge, this is the first report demonstrating that tumor suppressive miR-145-5p and miR-145-3p directly targeted UHRF1. Moreover, UHRF1 was upregulated in BC clinical specimens and contributed to anti-apoptotic effects through its regulation of several oncogenic genes. Expression of UHRF1 might be a useful prognostic marker for survival of BC patients. The identification of novel molecular pathways and targets regulated by the miR-145-5p/miR-145-3p/UHRF1 axis may lead to a better understanding of BC progression and aggressiveness.

MATERIALS AND METHODS

Clinical specimens and cell lines

Clinical tissue specimens were collected from BC patients (n = 69) who had undergone transurethral resection of their bladder tumors (TURBT, n = 59) or cystectomy (n = 10) at Kagoshima University Hospital between 2003 and 2013. NBE (n = 12) were derived from patients with noncancerous disease. The specimens were staged according to the American Joint Committee on Cancer-Union Internationale Contre le Cancer tumor-node-metastasis (TNM) classification and histologically graded [37]. Our study was approved by the Bioethics Committee of Kagoshima University; written prior informed consent and approval were obtained from all patients. Patient details and clinicopathological characteristics are listed in Table 6.

Table 6: Characteristic of patients

Bladder cancer (BC)
Total number	69
Median age (range)	73	(40–94)	years
Gender
Male	53	76.8%
Female	16	23.2%
Tumor grade
Low grade	45	65.2%
High grade	22	31.9%
Unknown	2	2.9%
T stage
Tis	2	2.9%
Ta	7	10.1%
T1	25	36.2%
T2	27	39.1%
T3	4	5.8%
T4	4	5.8%
N stage
N0	40	58.0%
N1	8	11.6%
Unknown	21	30.4%
M stage
M0	58	84.1%
M1	5	7.2%
Unknown	6	8.7%
Operation method
TURBT	59	85.5%
Cystectomy	10	14.5%
Normal bladder epithelium
Total number	12
Median age (range)	61	(47–72)	years

Abbreviation: TURBT = transurethral resection of bladder tumor

We used two human BC cell lines: T24, which was invasive and obtained from the American Type Culture Collection; and BOY, which was established in our laboratory from an Asian male patient, 66 years old, who was diagnosed with stage III BC and lung metastasis [38, 39]. These cell lines were maintained in minimum essential medium supplemented with 10% fetal bovine serum in a humidified atmosphere of 5% CO₂ and 95% air at 37°C.

Tissue collection and RNA extraction

Tissues were immersed in RNAlater (Thermo Fisher Scientific; Waltham, MA, USA) and stored at –20°C until RNA extraction was conducted. Total RNA, including miRNA, was extracted using the mirVana^™ miRNA isolation kit (Thermo Fisher Scientific) following the manufacturer’s protocol. The integrity of the RNA was checked with an RNA 6000 Nano Assay kit and a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) following the manufacturer’s protocol.

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

The procedure for qRT-PCR quantification was described previously [40, 41]. Stem-loop RT-PCR (TaqMan MicroRNA Assays; product ID: 002278 for miR-145-5p and product ID: 002149 for miR-145-3p; Thermo Fisher Scientific) was used to quantify miRNAs according to previously published conditions [40–42]. TaqMan probes and primers for UHRF1 (product ID: Hs 01086727_m1; Thermo Fisher Scientific) were assay-on-demand gene expression products. We used human GUSB (product ID: Hs99999908_m1; Thermo Fisher Scientific) and RNU48 (product ID: 001006; Thermo Fisher Scientific), respectively, as internal controls.

Transfections with miRNA mimic and small interfering RNA (siRNA) into BC cell lines

Mature miRNA molecules, Pre-miR^™ miRNA precursors (hsa-miR-145-5p; product ID: PM11480, hsa-miR-145-3p; product ID: PM13036, and negative control miRNA; product ID: AM 17111; Thermo Fisher Scientific) were used in the gain-of-function experiments, whereas UHRF1 siRNA (product ID: HSS120939 and HSS179006; Thermo Fisher Scientific) and negative control siRNA (product ID: D-001810-10; Thermo Fisher Scientific) were used in the loss-of-function experiments. The transfection procedures and transfection efficiencies of miRNA in T24 and BOY cells were reported previously [40–42].

Cell proliferation, migration, and invasion assays

To investigate the functional significance of the miR-145-5p, miR-145-3p, and UHRF1, we performed cell proliferation, migration, and invasion assays using T24 and BOY cells. The experimental procedures were performed as described in our previous studies [40-42].

Apoptosis assays

BC cell lines were transiently transfected with reagent only (mock), miR-control, miR-145-5p, miR- 145- 3p, siRNA-control, or si-UHRF1 at 10 nM in 6 well tissue culture plates, as described previously [14, 17–19]. Cells were harvested by trypsinization 72 hours after transfection and washed in cold phosphate-buffered saline. For apoptosis assays, double staining with FITC-Annexin V and propidium iodide was carried out using a FITC Annexin V Apoptosis Detection Kit (BD Biosciences, Bedford, MA, USA) according to the manufacturer’s recommendations and analysed within 1 hour by flow cytometry (CyAn ADP analyzer; Beckman Coulter, Brea, CA, USA). Cells were identified as viable cells, dead cells, early apoptotic cells, and apoptotic cells using Summit 4.3 software (Beckman Coulter), and the percentages of early apoptotic and apoptotic cells from each experiment were then compared. As a positive control, we used 2 μg/mL cycloheximide.

Cell cycle assays

For the cell cycle analyses, cells were stained with PI using the Cycletest PLUS DNA Reagent Kit (BD Biosciences) following the protocol and analyzed by CyAn ADP analyzer (Beckman Coulter). The percentages of the cells in the G0/G1, S, and G2/M phases were determined and compared. Experiments were performed in triplicate.

Western blot analyses

Immunoblotting was performed with rabbit anti-UHRF1 antibodies (1:500, PA5-29884; Thermo Fisher Scientific), anti-PARP antibodies (1:500 #9542; Cell Signaling Technology; Danvers, MA, USA), anti-cleaved PARP antibodies (1:500 #5625; Cell Signaling Technology), and anti-GAPDH antibodies (1:10000 MAB374; Chemicon, Temecula, CA, USA). Specific complexes were visualized with an echochemiluminescence detection system (GE Healthcare, Little Chalfont, UK).

Immunohistochemistry

A tissue microarray of 68 urothelial cancers and 20 normal bladder tissues was obtained from US Biomax, Inc. (Rockville, MD, USA) (product ID: BL1002). Detailed information on all tumor specimens can be found at http://www.biomax.us/index.php. The tissue microarray was immunostained following the manufacturer’s protocol with an Ultra Vision Detection System (Thermo Scientific). The primary rabbit polyclonal antibodies against UHRF1 (PA5-29884; Thermo Fisher Scientific) were diluted 1:300. Immunostaining was evaluated according to a scoring method as described previously [17].

Genome-wide gene expression and in silico analyses for the identification of genes regulated by miR-145-5p and miR-145-3p

To further investigate the specific genes affected by miR-145-5p and miR-145-3p, we performed a combination of in silico and genome-wide gene expression analyses. We attempted to identify target genes using a BC cell line transfected with these miRNAs. A Sure Print G3 Human GE 8 × 60K Microarray (Agilent Technologies) was used for expression profiling of miR-145-5p and miR-145-3p transfectants. The microarray data were deposited into GEO (http://www.ncbi.nlm.nih.gov/geo/) and were assigned GEO accession number GSE66498. Next, we selected putative miRNA target genes using the microRNA.org database (August, 2010 release, http://www.microrna.org). Finally, to identify upregulated genes in BC, we analyzed publicly available gene expression data sets in GEO (accession numbers: GSE11783, GSE31684). The data were normalized and analyzed with Gene Spring software (Agilent Technologies) as described previously [22, 23, 40–42]. The strategy for investigation of the target genes is shown in Figure 3.

Plasmid construction and dual luciferase reporter assays

Partial wild-type sequences of the 3′ UTR of UHRF1 or those with a deleted miR-145-5p and miR- 145- 3p target site (positions 1,179–1,198 of UHRF1 3′ UTR for miR- 145-5p, and positions 287–292 of UHRF1 3′ UTR for miR-145-3p) were inserted between the XhoI and PmeI restriction sites in the 3′ UTR of the hRluc gene in the psiCHECK-2 vector (C8021; Promega, Madison, WI, USA). T24 and BOY cell lines were transfected with 50 ng of the vector and 10 nM miR-145-5p or miR-145-3p using Lipofectamine 2000 (Thermo Fisher Scientific) and Opti-MEM (Thermo Fisher Scientific). The activities of firefly and Renilla luciferases in cell lysates were determined with a dual luciferase reporter assay system according to the manufacturer’s protocol (E1960; Promega). Normalized data were calculated as the ratio of Renilla/firefly luciferase activities.

Identification of downstream targets regulated by UHRF1 in BC

To investigate molecular targets regulated by UHRF1 in BC cells, we carried out gene expression analyses using si-UHRF1-transfected BC cell lines. Microarray data were used for expression profiling of si-UHRF1 transfectants. The microarray data were deposited into GEO (accession number: GSE77790). We analyzed common down or upregulated genes using the GEO dataset. The flow chart outlining the investigation of UHRF1 downstream genes is shown in Figure 10A and 10B.

Statistical analysis

Relationships among two or three variables and numerical values were analysed using the Mann-Whitney U test or Bonferroni-adjusted Mann-Whitney U test. Spearman’s rank test was used to evaluate the correlation among the expressions of miR-145-5p, miR-145-3p, and UHRF1. We estimated cause specific survival of 57 BC patients by using the Kaplan-Meier method. Among the 69 BC patients, 12 died of other causes. Therefore, we analyzed cause specific survival of 57 BC patients. Patients were divided into two groups according to the median value of UHRF1 expression, and the differences between the two groups were evaluated by the log-rank tests. We used Expert Stat View software, version 5.0 (SAS Institute Inc., Cary, NC, USA), for these analyses.

ACKNOWLEDGMENTS AND FUNDING

This study was supported by JSPS KAKENHI Grant Numbers 26293354, 25462490, and 26462416.

CONFLICTS OF INTEREST

The authors indicated no potential conflicts of interest.