MCF2L-AS1 promotes the biological behaviors of hepatocellular carcinoma cells by regulating the miR-33a-5p/FGF2 axis

Long noncoding RNA MCF2L-AS1 functions in the development of cancers like lung cancer, ovarian cancer, and colorectal cancer. Notwithstanding, its function in hepatocellular carcinoma (HCC) stays obscure. Our research probes its role in MHCC97H and HCCLM3 cell proliferation, migration, and invasion. qRT-PCR gauged MCF2L-AS1 and miR-33a-5p expressions in HCC tissues. CCK8, colony formation, Transwell, and EdU assays detected HCC cell proliferation, invasion, and migration, respectively. The xenograft tumor model was built to confirm the MCF2L-AS1-mediated role in HCC cell growth. Western blot and immunohistochemistry detected FGF2 expression in HCC tissues. Bioinformatics analysis predicted the targeted relationships between MCF2L-AS1 or FGF2 and miR-33a-5p, which were further examined through dual-luciferase reporter gene and pull-down assays. MCF2L-AS1 was expressed highly in HCC tissues and cells. MCF2L-AS1 upregulation enhanced HCC cells’ proliferation, growth, migration, and invasion and reduced apoptosis. miR-33a-5p was demonstrated as an underlying target of MCF2L-AS1. miR-33a-5p impeded HCC cells’ malignant behaviors. MCF2L-AS1 overexpression reversed miR-33a-5p-mediated effects. MCF2L-AS1 knockdown enhanced miR-33a-5p and negatively regulated FGF2 protein. miR-33a-5p targeted and inhibited FGF2. miR-33a-5p overexpression or FGF2 knockdown inhibited MCF2L-AS1-mediated oncologic effects in MHCC97H. By modulating miR-33a-5p/FGF2, MCF2L-AS1 exerts a tumor-promotive function in HCC. The MCF2L-AS1-miR-33a-5p-FGF2 axis may provide new therapeutic targets for HCC treatment.


INTRODUCTION
Hepatocellular carcinoma (HCC) is regarded as one of the most prevalent cancers across the world with high morbidity, aggressive invasion, rapid development, and poor prognosis [1,2]. Due to the high incidence of hepatitis B virus (HBV) infection, China has a high prevalence of hepatocellular carcinoma (HCC) [3]. Patients with HCC are prone to show malignant clinicopathological features in lymphatic vessels, blood vessels and distant viscera metastasis at the early stage [4]. Elucidating the molecular biological changes in HCC development can help HCC diagnosis and treatment [5,6].
Fibroblast growth factors (FGFs) are found to be extensively expressed in the tissues of human body; FGF and FGF receptor (FGFR) usually constitute FGFR signaling pathways to participate in growth and development, wound healing and fibrosis, inflammation, and neogenesis of many tumors [7,8]. FGF2, belonging to the FGF family, is situated on human chromosome 4q26 with a total length of 38kb [9]. FGF2 is originally isolated and purified from the pituitary gland of cattle and undertakes a significant role in promoting cellular proliferation, differentiation, and vascular growth [10]. Growing evidence has shown that FGF2 is altered in tumor tissues [11]. For instance, FGF2 expression was increased in ovarian cancer (OC), which was related to AGING the clinical stage [12,13]. FGF2 could interact with FGFR to trigger downstream signaling and regulate HCC invasiveness [14,15].
miRNAs are thought to impact over 90% of proteincoding gene loci in the human genome through diverse regulatory approaches, such as combining with the 3'UTR of target messenger RNAs (mRNAs) and affecting gene expression post-transcriptionally [16,17]. miR-33a-5p, one member of miRNAs, participates in tumor progression. For example, miR-33a-5p targets Snail family transcriptional repressor 2 (SNAI2) to suppress the migration of melanoma cells [18]. Xing et al. have discovered that miR-33a-5p can mediate the proliferation, migration and apoptosis of cardiomyocyte progenitor cells via targeting NODAL modulator 1 (NOMO1) [19]. On the other hand, long non-coding RNAs (lncRNAs), another class of ncRNAs with over 200 nts in length with tissue specificity [20], undertake an essential role in many biological behaviors including tumorigenesis, neuroscience and ontogeny, especially in tumor cell proliferation, apoptosis, invasion, migration and chemotherapeutic drug resistance [21][22][23][24]. lncRNA MCF2L antisense RNA 1 (MCF2L-AS1) is a newly reported molecule in recent years. Kong et al. indicated that MCF2L-AS1 aggravates colorectal cancer development via miR-105-5p/RAB22A signaling [25]. lncRNA MCF2L-AS1 exhibits a high level under the conditions of non-small cell lung cancer, and knocking down MCF2L-AS1 impedes the cancer stem cell (CSC)like traits of NSCLC cells [26]. However, lncRNA MCF2L-AS1 profile in both liver cancer tissues and cells, as well as its regulatory mechanism, is still poorly understood.
The intention of the research is to delve into the influence of MCF2L-AS1, miR-33a-5p and FGF2 on liver cancer cell proliferation, invasion, migration and apoptosis, as well as to examine the regulatory mechanism of the MCF2L-AS1-miR-33a-5p-FGF2 axis in the context of HCC. We hope this study provides a new strategy for HCC treatment.

lncRNA MCF2L-AS1 was altered in hepatocellular carcinoma
The GEPIA database confirmed that lncRNA MCF2L-AS1 has a certain expression ability in most organs in the human body, including the liver ( Figure 1A). Enhanced MCF2L-AS1 expression in liver cancer tissue samples ( Figure 1B, 1C). We also detected MCF2L-AS1 expression in tissue samples from liver cancer patients by qRT-PCR, unveiling that MCF2L-AS1 profile was vigorously up-regulated vis-à-vis the normal control group ( Figure 1D). Through the database, we also found MCF2L-AS1 could also be expressed in liver cancer cells ( Figure  1E). In addition, an elevated MCF2L-AS1 profile bore a relation to the poorer survival rate of liver cancer patients ( Figure 1F). Therefore, lncRNA MCF2L-AS1 might act as a tumor-promoting factor in liver cancer.

lncRNA MCF2L-AS1 promoted tumor cell growth in vivo
Next, we conducted in vivo assays for investigating the function of MCF2L-AS1. As suggested by our statistics, MCF2L-AS1 overexpression enhanced tumor volume and weight in the xenograft tumor model ( Figure 3A-3C). RT-PCR confirmed that MCF2L-AS1 was upregulated in the tumor tissues (versus the Vector group, Figure 3D). Next, the TUNEL assay was performed for evaluating apoptosis. It was found that MCF2L-AS1 overexpression reduced TUNEL-positive cell rate ( Figure 3E). Furthermore, we performed in vivo assays using HCCLM3 cells. Similarly, MCF2L-AS1 overexpression promoted cell growth and reduced apoptosis ( Figure 3F-3J). Therefore, lncRNA MCF2L-AS1 showed tumor-promotive effects in liver cancer.

DISCUSSION
Immunotherapy based on targeted molecules has been rapidly developed and shows potential in HCC treatment [27,28]. Presently, we placed emphasis on the function of MCF2L-AS1 in HCC and found that it derives HCC development via affecting cell proliferation, migration, invasion and apoptosis. The underlying mechanism of MCF2L-AS1 in HCC might depend on regulating the miR-33a-5p-FGF2 axis, which offers a new understanding of HCC progression ( Figure 8).
With deeper investigation, the role of MCF2L-AS1 in malignant tumors begins to attract people's attention [29], and its potential as a tumor diagnostic marker and therapy target is also enhanced [30]. In our study, MCF2L-AS1 profile was abnormally boosted in liver cancer tissues and cells through the database. The survival curve indicated that the high profile of MCF2L-AS1 bore a relation to the poor survival rate of patients. Controlling tumor cell proliferation and metastasis has always been strategies in treating HCC [31,32]. Therefore, we conducted both ex vivo and in vivo assays for confirming the function of MCF2L-AS1 in HCC cells. We found that MCF2L-AS1 exerted promotive effects on proliferation, invasion and migration of MHCC97H cells, and reduce apoptosis. These findings denoted that MCF2L-AS1 might be implicated in the pathological process of liver carcinoma.
Competing endogenous RNA (ceRNA) refers to one miRNA that is able to mediate multiple target genes, which constitutes a kind of competitive relationship [33].
Fibroblast growth factors (FGFs) are found to be extensively expressed in the tissues of human body; FGF and FGF receptor (FGFR) usually constitute FGFR signaling pathways to participate in growth and development, wound healing and fibrosis, inflammation, and neogenesis of malignant tumors [42]. Altered FGF/ FGFR signaling is also observed in HCC [43]. A clinical analysis showed that patients with the FGF2 rs308447 TT genotype had shorter overall survival than patients with the CC or CT genotype (p=0.016) and that FGF2 rs308379 A allele carriers had shorter overall survival than patients with the TT genotype (p=0.020), suggesting that the FGF2 rs308379 A allele and advanced tumor stage were independent prognostic factors for overall survival in patients with HCC [14]. In a recent clinical trial, the potent and selective fibroblast growth factor receptor (FGFR) 1-3 inhibitor pemigatinib has a manageable safety profile and pharmacodynamic and clinical activity in patients with FGFR fusions/ rearrangements and mutations [44]. FGF2 has been found to be upregulated in HCC and induces oncogenic functions, such as promoting HCC metastasis, epithelialmesenchymal transition, and angiogenesis [15,45,46]. For example, E2F transcription factor 3 (E2F3) has a combination with the promoter of FGF2 and induces FGF2 overexpression and the malignant progression of HCC cells. The underlying mechanism involves initiating FGFR1-PI3K/AKT and MEK/ERK signaling pathways [47]. Those studies all suggest that FGF2/EGFR pathway is a therapeutic target in HCC. miRNAs have been reported to function in the miRNA-mRNA axis by modulating their own target gene expressions, which modulates biological processes like cell differentiation, proliferation, growth, apoptosis, invasion and migration [41,48,49]. Interestingly, our bioinformatic analysis showed that miR-33a-5p can target FGF2 and inhibit its expression at both mRNA and protein levels. Previous studies have revealed that miR-33a-5p can target multiple genes, such as carnitine O-octanoyltransferase (CROT) [50], PTGS2 [40], PNMA family member 1 (PNMA1) [51], and Wnt inhibitors Dickkopf-1 (DKK1) [37] et al. Presently, we confirmed that FGF2 is a direct target of miR-33a-5p, and MCF2L-AS1 enhances FGF2 level by directly targeting miR-33a-5p, and the MCF2L-AS1-miR-33a-5p-FGF2 axis is a potential diagnostic and therapy target in HCC.
However, several limitations also should be noted in our study. First, more clinical samples of HCC are necessary for confirming the predictive role of MCF2L-AS1-miR-33a-5p-FGF2 axis in HCC patients. Second, the underlying mechanism of MCF2L-AS1-miR-33a-5p-FGF2 axis needs investigation. Third, in vivo experiments are necessary for confirming this axis in tumor cell growth and metastasis.

CONCLUSIONS
In summary, we found that MCF2L-AS1 negatively modulated miR-33a-5p and up-regulated FGF2 protein so as to influence MHCC97H proliferation, apoptosis, migration, and invasion. The MCF2L-AS1-miR-33a-5p-FGF2 axis was likely to be implicated in the pathological processes of liver cancer, and MCF2L-AS1 possibly functioned as a potential biomarker for the early screening and diagnosis of cancer.

Clinical sample collection
The tumor specimens and tumor-adjacent normal hepatic tissues were harvested from 53 patients (age: 55.93±12.27) suffering from HCC in HwaMei Hospital, University of Chinese Academy of Sciences. All patients had never received chemotherapy before surgery, and all fresh samples were put into liquid nitrogen after isolation. The patients signed the informed consent. This research was conducted in compliance with the ethical guidelines of the 1975 Declaration of Helsinki and received the green light from the Ethics Committee of HwaMei Hospital, University of Chinese Academy of Sciences (No. 2018124).

Quantitative real-time polymerase chain reaction (qRT-PCR)
TRIzol (Invitrogen, Waltham, MA, USA) was used for extracting total RNA from tissues, MHCC97H and HCCLM3 cells. The total RNA concentration was determined using UV-Vis Spectrophotometry on Thermo Fisher Scientific™ NanoDrop Lite. The integrity of total RNA was confirmed by agarose gel electrophoresis. All extracted RNAs were synthesized into complementary DNA (cDNA) with the utilization of the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). The steps of this procedure were conducted according to the manufacturer's instructions. Next, RT-qPCR was performed using the SYBR Green PCR kit (Vazyme, Nanjing, China) on a CFX96TM Real-time System (Bio-Rad, Hercules, CA, USA). 45 cycles were included in the reaction, each cycle included amplification at 95° C for 5 minutes, 10 s at 60° C, and 30 s at 72° C. The relative expression of detected genes was analyzed using the 2 -ΔΔCT method. GAPDH serves as the internal control of FGF2 and MCF2L-AS1, and U6 was used for the internal control of miR-33a-5p. All experiments were performed in triplicate. The ultimate results were obtained after three repetitions of the tests (Table 2).

Dual-luciferase reporter assay
After constructing the wild-type (Wt) and mutant-type (Mut) plasmids of MCF2L-AS1 or FGF2, we extracted MCF2L-AS1-WT/MCF2L-AS1-Mut or FGF2-WT/ FGF2-Mut plasmids and transfected them into 293T cells together with miR-33a-5p inhibitors. Subsequent to the removal of the cell culture medium after 48-hour transfection, we washed the cells in PBS and added 20μl lysis buffer for cell lysates collection. The dualluciferase reporter assay system (Promega, Madison, WI, USA) examined the luciferase activity.

Cell counting kit-8 (CCK8) assay
MHCC97H and HCCLM3 cells were evenly spread onto 96-well (1×104 cells/well) plates with 200 µl of cell suspension per well and were routinely cultured for 24h. As described by the CCK-8 kit's instructions, 10 µl CCK-8 solution was applied to each well for 1-2h incubation at a temperature of 37° C. A microplate reader (Bio-Rad, Beijing, China) gauged the absorbance (450 nm).

Colony formation experiment
MHCC97H and HCCLM3 were seeded onto 60 mm dishes with 600 cells per dish. The culture medium was renewed every 2-3 days. Ten days later, the cells were washed by PBS twice and fixed by 4% paraformaldehyde. Crystal violet (0.5%) was used for cell staining. Then the colonies (each colony encompassed over 50 cells) were observed under a light microscope and counted.

Transwell assay
After the Matrigel gel was kept at 4° C for 12-24 hours, the liquefied Matrigel and the medium were diluted at a ratio of 1:6. The bottom of the upper chamber was coated with Matrigel Diluent (50 μl), which was dried in the incubator for 4h. MHCC97H cells were adjusted to the cell density: 1×106 cells/ml. Then, the upper chamber was added with cell suspension (100 μl). 10% FBS (600 μl) was applied to the medium in the bottom chamber, followed by 24-hour culture in an incubator (5% CO2, 37° C). After removing the chamber, we rinsed it three times with PBS, fixed it with ethanol (95%) for 5 min, and stained it in crystal violet staining solution (0.5%) for 10min. PBS was used to rinse off the staining solution of unbound cells. Cotton swabs were adopted to gently wipe off cells on the upper layer of the filter membrane. The cells on the lower layer were monitored under a microscope (Olympus, Tokyo, Japan).

EdU assay
The EdU staining kit (Beyotime, Shanghai, China) evaluated cell proliferation. MHCC97H and HCCLM3 cells were grown in 24-well plates (1×105 per well) for 24 hours. The EdU solution was added to each well for 2 hours. PBS flushed the cells twice. After MHCC97H and HCCLM3 were immobilized in 4% paraformaldehyde for 30-60 min and flushed twice in PBS, the cells were incubated along with 0.3% H2O2 (20min, 37° C) and rinsed three times in PBS. Afterwards, MHCC97H cells were incubated for 30 minutes at a temperature of 37° C after the addition of 50μL EdU staining solution. The cells were rinsed 3 times by PBS. The nuclei were stained by DAPI solution (Beyotime, Shanghai, China) and rinsed in PBS (3 times). The EdU-positive cells were mounted for observation under a fluorescence microscope (Olympus, Tokyo, Japan).

Western blot
Both clinical samples and HCC cells were collected, and the total protein was extracted from the tissues and cells using RIPA lysis (Beyotime, Shanghai, China). After determination of cell concentration, equal content  total protein (20 μg) from each group were separated based on molecular weight using gel electrophoresis. Next, the proteins are transferred from the gel onto a PVDF membranes. The membranes were blocked with 5% skim milk to prevent non-specific binding of the primary antibody for one hour at room temperature. Junction, NJ, USA) was used for bands exposing. The intensity of the band is measured using densitometry to quantify the amount of target protein present.

Mouse xenograft model
Eighteen BALB/c nude mice (5-9 weeks; 14-21g) were kept at 22° C with 40-75% humidity under a 12h light/dark cycle. Water and food were available for all mice. The nude mice were randomized to two groups: MCF2L-AS1 and negative control (Vector). MHCC97H and HCCLM3 cells were transfected respectively and adjusted to a concentration of 1×107cells/mL. Cell suspension (0.2 ml) was subcutaneously inoculated in the left anterior upper arm armpit of each nude mouse. After inoculation, we measured the max diameter (a) and min diameter (b) of each tumor in the nude mice utilizing a vernier caliper every 3 days in accordance with the formula: V=1/2×a×b2. After 3 weeks, all animals were sacrificed; the tumor was stripped, weighed, photographed and fixed with formaldehyde. The TUNEL staining kit (Beyotime, Shanghai, China) was utilized for evaluating apoptosis in the tumor tissues.

Analysis of data
Statistical data were exhibited as mean ± standard deviation with the use of the SPSS Statistics 22.0 software. T-test contrasted the statistics of two groups; one-way ANOVA compared data among ≥3 groups. P<0.05 means statistical significance. All tests were duplicated at least 3 times independently.

Availability of data and materials
The data and materials in the current study are available from the corresponding author on reasonable request.

AUTHOR CONTRIBUTIONS
HO and LM conceived and designed the project. HO conducted RT-PCR, Western blot, EdU staining, Colony formation assay, Transwell assay; YQ performed Dualluciferase reporter assay, CCK8 assay, Mouse xenograft model, and TUNEL staining. YQ and LM conducted clinical sample collection and cell culture. YQ analyzed and interpreted the data. The first draft of the manuscript was written by LM and HO. All authors read and approved the manuscript.

CONFLICTS OF INTEREST
All authors declare no conflicts of interests.