LINC00565 Enhances Proliferative Ability in Endometrial Carcinoma by Downregulating KLF9.

OBJECTIVE: To detect LINC00565 expression level in endometrial carcinoma (EC) samples and cell lines, and the correlations between LINC00565 and clinical features of EC patients. After intervening LINC00565, the underlying mechanism about proliferative ability in EC cell lines is observed.
METHODS: Relative levels of LINC00565 and KLF9 in 52 paired EC and paracancerous tissues were detected by quantitative real-time polymerase chain reaction (qRT-PCR). The relationship between relative level of LINC00565 or KLF9 and clinical features of EC patients was analyzed. After knockdown of LINC00565 and KLF9, potential regulations of them on biological functions of EC were examined by Cell Counting Kit (CCK-8), colony formation assay and in vivo xenograft model in nude mice, respectively. At last, dual-luciferase reporter assay and rescue experiments were conducted to illustrate the mechanisms of LINC00565 and KLF9 in mediating the development of EC.
RESULTS: LINC00565 was upregulated in EC tissues. Chi-square analysis showed that a high level of LINC00565 predicted large tumor size, advanced pathological staging and poor prognosis in EC. Silence of LINC00565 decreased proliferative ability in EC cells and tumor growth in nude mice bearing EC. KLF9 was the target gene of LINC00565. The negative interaction between LINC00565 and KLF9 was responsible for stimulating the malignant development of EC. Knockdown of KLF9 could abolish the regulatory effects of silenced LINC00565 on proliferative ability and tumorigenesis in EC.
CONCLUSION: LINC00565 is upregulated in EC tissues and closely linked to tumor size, pathological staging and poor prognosis in EC patients. LINC00565 stimulates proliferative ability in EC by downregulating KLF9.

Introduction

In recent years, there are 320,000 women diagnosed as endometrial carcinoma (EC) and 76,000 deaths of EC in the world each year.1–3 EC is considered to be one of the three malignant tumors in the female genital tract. In the Western developed countries, the prevalence of EC ranks the first in female genital tract tumors, which is also on the rise in developing countries and displays a younger onset trend.2–4 Although a great number of people are diagnosed as EC owing to screening or abnormal virginal bleeding, many patients were initially diagnosed as an advanced stage. Therapeutic efficacy of advanced EC patients is relatively unsatisfactory, leading to high recurrence and poor prognosis.5,6 Therefore, clarifying the pathogenesis of EC and searching for effective targets are beneficial to improve prognosis and life quality in affected people.6–8 Accumulating evidences have shown the vital functions of lncRNAs in the occurrence, progression, metastasis and prognosis of cancer.9,10 Researches on the relationship between lncRNAs and EC have achieved breakthroughs.

Long noncoding RNAs (LncRNAs), as one of the non-coding RNAs with over 200 bases long, are used to be thought of as byproducts of gene transcription.11,12 Recent studies have demonstrated that in the mammalian genome, lncRNAs are transcribed from thousands of loci and processed into mature ones, thus exerting biological functions in embryo development, gene expressions and tumorous disease progression.13–16 It is previously reported that LINC00565 is an oncogene triggering deterioration of tumor development.17,18 However, its potential role in EC remains unclear.

According to the competing endogenous RNAs hypothesis and bioinformatics prediction, binding sequences are identified in the 3ʹUTR of KLF9 and LINC00565.19,20 KLF9 is a member of the KLF family. Members in the KLF family all have a C2-H2 zinc-finger structure at the C-terminal, where promoters, GC/GT boxes and CACCC cis-acting elements in the enhancer sequences are recognized.21–23 In this paper, differential expressions of LINC00565 and KLF9 in EC tissues and paracancerous ones were detected. Their involvement in the development of EC was further explored.

Methods

Patients and EC Samples

EC tissues and paracancerous ones were surgically resected from 52 patients undergoing surgery or biopsy, which were pathologically confirmed and stored at −80°C. None of the included patients had preoperative anti-cancer treatment. Follow up was conducted by telephone and outpatient review. This study got approval by the Ethics Committee of Yantai Yuhuangding Hospital and it was conducted after obtaining written informed consent of each subject. This study was conducted in accordance with the Declaration of Helsinki.

Cell Lines and Reagents

EC cell lines (HEC-1A, HEC-1B, KLE and Ishikawa) and an endometrial stromal cell line (T-HESC) were purchased from ATCC (Manassas, VA, USA). Cells were cultured in Dulbecco’s modified eagle medium (DMEM) (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, Rockville, MD, USA), 100 U/mL penicillin and 100 μg/mL streptomycin in a 5% CO2 incubator at 37°C.

Transfection

To construct knockdown of LINC00565 in KLE and HEC-1B cells, respectively. Transfection plasmids were purchased from GenePharma, Shanghai, China. In addition, the small interfering RNA for KLF6 (si-KLF6) was designed, and si-NC were used as negative control. Thus, sh-NC, sh-LINC00565, sh-LINC00565+si-NC or sh-LINC00565+si-KLF9 were used. Cells were cultured to 30–50% confluence and transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Transfected cells were collected 48 hours later for the following use.

Cell Proliferation Assay

Cells were inoculated in a 96-well plate with 2×103 cells per well. At the appointed time points, absorbance value at 490 nm of each sample was recorded using the Cell Counting Kit (CCK-8) kit (Dojindo Laboratories, Kumamoto, Japan) for plotting the viability curves.24

Colony Formation Assay

Cells were inoculated in a 6-well plate with 200 cells per well and cultured for 2 weeks. Culture medium was replaced once in the first week and twice in the second week. Afterwards, visible colonies were washed in phosphate-buffered saline (PBS), fixed in methanol for 20 min and dyed in 0.1% crystal violet for 20 min. Finally, colonies were captured and calculated.

Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR)

According to the reference,25 the extracted RNAs by TRIzol reagent (Invitrogen, Carlsbad, CA, USA) were purified by DNase I treatment, and reversely transcribed into cDNAs using Primescript RT Reagent (Takara, Otsu, Japan). The obtained cDNAs underwent qRT-PCR using SYBR®Premix Ex Taq™ (Takara, Japan). Each sample was performed in triplicate, and relative level was calculated by 2−ΔΔCt and normalized to that of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). LINC00565: forward: 5ʹ-TAGACGGTCGCTCCATCAGT-3ʹ, reverse: 5ʹ-CCATCCTCAGGTTTGCATTT-3; KLF9: forward: 5ʹ-ATGTGCAGCATCTTCCAG-3ʹ, reverse: 5ʹ-CTCTAGGCAGGTCTGTTG-3ʹ; GAPDH: forward: 5ʹ-GCACCACACCTTCTACAATG-3ʹ, reverse: 5ʹ-TGCTTGCTGATCCACATCTG-3.

Western Blot

According to the reference,26 cells were lysed for isolating proteins and electrophoresed. Protein samples were loaded on polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Subsequently, non-specific antigens were blocked in 5% skim milk for 2 hours. Primary and secondary antibodies were applied for the indicated time. Band exposure and analyses were finally conducted.

In vivo Xenograft Model

Establishment of the in vivo xenograft model in nude mice bearing EC cell line was approved by the Animal Ethics and Use Committee of Yantai Yuhuangding Hospital Animal Center. All the protocols complied with UKCCCR (United Kingdom Co-ordinating Committee on Cancer Research) guidelines for the welfare of animals in experimental neoplasia. Twenty male nude mice were randomly assigned into four groups (n=5) and they were subcutaneously administrated with KLE cells transfected with sh-NC, sh-LINC00565, sh-LINC00565+si-NC or sh-LINC00565+si-KLF9, respectively. The indicated stable cell lines (2×106) were subcutaneously injected into the right flank of BALB/c (nu/nu) 4- to 6-week-old female nude mice. Tumor size was weekly recorded, and these mice were sacrificed for collecting tumor tissues 6 weeks later. Tumor volume = (width2×length)/2.

Dual-Luciferase Reporter Assay

Cells were pre-seeded in 24-well plates. Wild-type and mutant-type LINC00565 vectors were constructed based on the binding sequences in the 3ʹUTR of KLF9. Subsequently, they were co-transfected with pcDNA-NC/pcDNA-KLF9 and LINC00565-WT/LINC00565-MUT using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). The dual-luciferase reporter assay was performed to normalize the reporter luciferase activity to the control firefly luciferase activity 48 h later.

Statistical Analysis

SPSS 19.0 (SPSS IBM, Armonk, NY USA) was used for data analysis. Data were expressed as mean ± standard deviation. Differences between groups were analyzed by the t-test. The relationship between LINC00565 level and clinical features of EC patients was analyzed by the χ2 test or Fisher ’s exact probability method. Kaplan–Meier curves were depicted based on LINC00565 levels in EC patients. P < 0.05 was considered statistically significant.

Results

LINC00565 Expression Level in EC Samples and Cell Lines

LINC00565 expression level in 52 paired EC tissues and paracancerous tissues were detected. QRT-PCR data showed that the higher level of LINC00565 in EC tissues than paracancerous ones (Figure 1A). Compared with the endometrial stromal cell line, LINC00565 was identically up-regulated in EC cell lines (Figure 1B).

Figure 1

LINC00565 level in EC samples. (A) Differential expressions of LINC00565 in EC and paracancerous tissues; (B) LINC00565 levels in EC cell lines; (C) Kaplan–Meier curves depicted based on LINC00565 levels in EC patients. Data were expressed as mean±SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Relationship Between LINC00565 Level and Clinical Features of EC

The median level of LINC00565 in the collected 52 EC tissues was calculated, and thus divided EC patients into high (n=26) and low (n=26) LINC00565 expression group, respectively. The analysis results uncovered that LINC00565 level was correlated to tumor size and pathological staging of EC patients, while it was unrelated to age and rates of lymphatic metastasis and distant metastasis (Table 1). In addition, Kaplan–Meier curves demonstrated a poor prognosis in EC patients of high LINC00565 expression group (Figure 1C).

Table 1

Association of LINC00565 and KLF9 Expression with Clinicopathologic Characteristics of Endometrial Carcinoma

Parameters	Number of Cases	LINC00565 Expression		P -value	KLF9 Expression		P -value
		Low (%)	High (%)		High (%)	Low (%)
Age (years)				0.397			0.782
<60	21	12	9		8	13
≥60	31	14	17		13	18
BMI (kg/m2)				0.358			0.557
<24	37	20	17		14	23
≥24	15	6	9		7	8
Tumor size				0.027			0.011
<4 cm	26	17	9		15	11
≥4 cm	26	9	17		6	20
T stage				0.011			0.010
T1-T2	31	20	11		17	14
T3-T4	21	6	15		4	17
FIGO stage				0.569			0.264
Stage I	20	9	11		10	10
Stage II/III/IV	32	17	15		11	21
Lymph node metastasis				0.244			0.105
No	34	19	15		11	23
Yes	18	7	11		10	8
Distance metastasis				0.158			0.147
No	31	18	13		10	21
Yes	21	8	13		11	10

Knockdown of LINC00565 Inhibited in vitro Proliferative Ability and in vivo Tumorigenesis in EC

LINC00565 knockdown model was established in KLE and HEC-1B cells by the transfection of sh-LINC00565. Transfection efficacy of sh-LINC00565 was verified (Figure 2A). Knockdown of LINC00565 decreased the viability and colony number in EC cell lines, indicating that LINC00565 could promote the proliferative ability of EC (Figure 2B and C). In addition, the average tumor volume and tumor weight were lower in nude mice administrated with KLE cells transfected with sh-LINC00565 than those of controls, showing the reduced tumor growth rate (Figure 2D and E). The tumor tissues in nude mice were harvested after sacrifice at the end of the experimental period. Protein level of KLF9 was markedly higher in EC mice with in vivo knockdown of LINC00565 compared with those of controls (Figure 2F).

Figure 2

Knockdown of LINC00565 inhibited in vitro proliferative ability and in vivo tumorigenesis in EC. (A) Transfection efficacy of sh-LINC00565 in KLE and HEC-1B cells; (B) Viability in KLE and HEC-1B cells transfected with sh-NC or sh-LINC00565; (C) Colony number in KLE and HEC-1B cells transfected with sh-NC or sh-LINC00565 (magnification: 10×); (D) Average tumor volume that was weekly recorded in nude mice administrated with KLE cells transfected with sh-NC or sh-LINC00565; (E) Tumor weight in nude mice administrated with KLE cells transfected with sh-NC or sh-LINC00565; (F) Protein level of KLF9 in EC tissues collected from nude mice administrated with KLE cells transfected with sh-NC or sh-LINC00565. Data were expressed as mean±SD. *P < 0.05, **P < 0.01.

KLF9 Was the Downstream Gene of LINC00565

Wild-type and mutant-type LINC00565 vectors were constructed based on the binding sequences in the 3ʹUTR of KLF9. Decreased luciferase activity was observed in wild-type LINC00565 after KLF9 overexpression in EC cell lines (Figure 3A). Protein level of KLF9 was up-regulated by knockdown of LINC00565 in EC cells (Figure 3B). Contrary to LINC00565, KLF9 was down-regulated in EC tissues and cell lines (Figure 3C and F). KLF9 level was negatively linked to LINC00565 level in EC tissues (Figure 3D). Moreover, lowly expressed KLF9 predicted a poor prognosis in EC patients (Figure 3E).

Figure 3

KLF9 was the downstream gene of LINC00565. (A) Binding sequences in the 3ʹUTR of LINC00565 and KLF9. Luciferase activity in KLE and HEC-1B cells co-transfected with NC/pcDNA-KLF9 and LINC00565-WT/LINC00565-MUT; (B) Protein level of KLF9 in KLE and HEC-1B cells transfected with sh-NC or sh-LINC00565; (C) Differential expressions of KLF9 in EC and paracancerous tissues; (D) A negative correlation between relative expressions of LINC00565 and KLF9 in EC tissues; (E) Kaplan–Meier curves depicted based on KLF9 levels in EC patients; (F) KLF9 levels in EC cell lines. Data were expressed as mean±SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Knockdown of KLF9 Abolished the Regulatory Effects of Silenced LINC00565 on in vitro Proliferative Ability and in vivo Tumorigenesis in EC

To uncover the involvement of LINC00565 and KLF9 in the development of EC, LINC00565 and KLF9 co-silence model was established. Protein level of KLF9 was lower in EC cells co-transfected with sh-LINC00565 and si-KLF9 than those co-transfected with sh-LINC00565 and si-NC (Figure 4A). Increased viability and colony number were observed in EC cells with co-silenced LINC00565 and KLF9 compared with those with only LINC00565 knockdown (Figure 4B and C). Subsequently, in vivo effects of LINC00565 and KLF9 on the growth rate of EC were detected. Compared with nude mice administrated with KLE cells co-transfected with sh-LINC00565 and si-NC, the average tumor volume and tumor weight were larger in those with co-silence of LINC00565 and KLF9 (Figure 4D and E). Downregulated KLF9 level was examined in EC tissues collected from mice with in vivo co-knockdown of LINC00565 and KLF9 than those with solely knockdown of LINC00565 (Figure 4F).

Figure 4

Knockdown of KLF9 abolished the regulatory effects of silenced LINC00565 on in vitro proliferative ability and in vivo tumorigenesis in EC. (A) Protein level of KLF9 in KLE and HEC-1B cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9; (B) Viability in KLE and HEC-1B cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9; (C) Colony number in KLE and HEC-1B cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9 (magnification: 10×); (D) Average tumor volume that was weekly recorded in nude mice administrated with KLE cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9; (E) Tumor weight in nude mice administrated with KLE cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9; (F) Protein level of KLF9 in EC tissues collected from nude mice administrated with KLE cells co-transfected with sh-LINC00565 and si-NC, or sh-LINC00565 and si-KLF9. Data were expressed as mean±SD. *P < 0.05, **P < 0.01.

Discussion

EC derives from women’s endometrial epithelium, accounting for about 20–30% of female reproductive system tumors. Its incidence has increased year by year.1,2 Current researches believe that the occurrence and development of EC are chronic and continuous processes, involving intracellular and epigenetic changes, oncogene activation, inactivation of tumor suppressors, hormones and their receptors, etc.2–4 Nevertheless, its specific pathogenesis is not clear. Great breakthroughs in the early diagnosis and recurrence prediction of EC are lacked, nor as effective targeted drugs. Differential expression and vital functions of lncRNAs broaden our understanding of tumor researches.5–10 LncRNAs are able to influence apoptosis signaling, tumor invasiveness and metastasis.13–16 They are featured by long sequences, diverse spatial structures and complex functions. Through three levels, lncRNAs regulate target gene expressions. First of all, they irreversibly and genetically change gene functions without altering their DNA sequences at the epigenetic level. This process mainly involves regulations of DNA methylation and demethylation, RNA interference, histone modification, chromosome remodeling or other methods.11,15 Secondly, lncRNAs transcriptionally regulate gene expressions by mediating surrounding genes, thus leading to chromosome remodeling. Besides, they can bind basal transcription factors, thereafter inactivating promoters and activating accessory proteins.14–16 Thirdly, lncRNAs post-transcriptionally process, mediate and modify target genes by complementary base pairing.15,16

So far, multiple lncRNAs have been discovered to be related to tumor development.19 LINC00565 is a tumor-associated lncRNA involved in the development of ovarian cancer and gastric cancer.17,18 We collected 52 paired EC tissues and paracancerous tissues and detected the relative levels of LINC00565 and KLF9 in EC tissues. LINC00565 was up-regulated and KLF9 was down-regulated in EC tissues. It is speculated that LINC00565 was an oncogene and KLF9 was a tumor suppressor involved in the development of EC. By analyzing clinical data of included EC patients, it is found that LINC00565 level was linked to tumor size, pathological staging and prognosis in EC patients. Furthermore, we constructed LINC00565 knockdown model in KLE and HEC-1B cells by lentivirus transfection. In vitro experiments demonstrated the promotive effect of LINC00565 on proliferative ability in EC cells. LINC00565 was also capable of stimulating tumor growth in nude mice bearing EC.

To uncover the molecular mechanisms of LINC00565 in regulating the development of EC, the downstream gene of LINC00565 was searched. Our findings confirmed that KLF9 was the target binding LINC00565. Knockdown of LINC00565 markedly upregulated KLF9 level. Notably, knockdown of KLF9 abolished the regulatory effects of silenced LINC00565 on EC cell proliferation and tumorigenesis in nude mice bearing EC. To sum up, LINC00565 stimulated the malignant development of EC by negatively regulating KLF9.

Conclusions

To sum up, LINC00565 is upregulated in EC tissues and closely linked to tumor size, pathological staging and poor prognosis of EC patients. Besides, LINC00565 stimulates the proliferative ability of EC cell lines by downregulating KLF9.