LncRNA HCG11 promotes 5-FU resistance of colon cancer cells through reprogramming glucose metabolism by targeting the miR-144-3p-PDK4 axis

Abstract

BACKGROUND:

Colorectal cancer (CRC), one of the most common human malignancies, is a leading cause of the cancer-related mortality. 5-FU is a first-line chemotherapeutic agent against CRC. Although CRC patients responded to 5-FU therapy initially, a part of patients succumbed to CRC due to the acquired drug resistance. Thus, investigating molecular mechanisms underlying chemoresistance will contribute to developing novel strategies against colorectal cancer.

OBJECTIVE:

Accumulation evidence revealed pivotal roles of long non-coding RNAs (lncRNAs) in tumorigenesis and chemoresistance of CRC. However, the precise roles and molecular mechanisms of lncRNA-HCG11 in CRC remain unclear. This study aimed to investigate the biological roles and underlying mechanisms of HCG11 as well as its molecular targets in regulating the cellular metabolism processes, which facilitate the chemoresistance of CRC.

METHODS AND RESULTS:

This study uncovers that HCG11 was significantly upregulated in CRC tumors tissues and cell lines. Moreover, HCG11 was elevated in 5-FU resistant CRC tumors. Silencing HCG11 inhibited colon cancer cell proliferation, migration, invasion and glucose metabolism and sensitized CRC cells to 5-FU. In addition, we detected increased HCG11 expression level and glucose metabolism in the established 5-FU resistant CRC cell line (DLD-1 5-FU Res). Furthermore, microRNA-microArray, RNA pull-down and luciferase assays demonstrated that HCG11 inhibited miR-144-3p which displays suppressive roles in colon cancer via sponging it to form a ceRNA network. We identified pyruvate dehydrogenase kinase 4 (PDK4), which is a glucose metabolism key enzyme, was directly targeted by miR-144-3p in CRC cells. Rescue studies validated that the miR-144-3p-inhibited glucose metabolism and 5-FU sensitization were through targeting PDK4. Finally, restoration of miR-144-3p in HCG11-overexpressing DLD-1 5-FU resistant cells successfully overcame the HCG11-faciliated 5-FU resistance via targeting PDK4.

CONCLUSION:

In summary, this study reveals critical roles and molecular mechanisms of the HCG11-mediated 5-FU resistance through modulating the miR-144-3p-PDK4-glucose metabolism pathway in CRC.

Keywords

Colorectal cancer 5-FU resistance lncRNA-HCG11 PDK4 glycolysis Warburg effect

1. Introduction

Colorectal cancer, one of the most diagnosed malignancies, is associated with poor prognosis and high mortality rate [1]. Currently, surgical resection is the primary approach for CRC patients at early stage [2]. In addition, combination therapy with chemotherapy and/or radiotherapy reduced the risk of recurrence and improved overall survival of CRC patients at metastatic or advanced stages [3]. 5-fluorouracil (5-FU) is a traditionally first-line anti-cancer drug for CRC patients through interfering DNA synthesis, resulting in DNA damage and cancer cell death [4]. Over the last decades, 5-FU has been applied as an effectively chemotherapeutic agent for colorectal cancer [5]. However, development of intrinsic and acquired 5-FU resistance remains a major barrier to its clinical applications, leading to severe side-effects due to intolerable dosages [6, 7]. Therefore, understanding the precise mechanism of 5-FU resistance is critical to developing therapeutic strategies for colon cancer patients.

Long non-coding RNAs (lncRNAs) are a clan of endogenous RNAs with relatively large size ( $>$ 200 nt) and have no ability of protein coding [8]. Abnormal expression levels of lncRNAs are known to be involved in tumor processes including tumorigenesis, proliferation, migration, invasion, metabolism and chemoresistance [9]. LncRNA HLA complex group 11 (HCG11) has been reported to be dysregulated in diverse cancers such as gastric cancer [10], prostate cancer [11, 12], ovarian cancer [13], cervical cancer [14], glioma [15, 16] and liver cancer [17, 18], indicating HCG11 plays vital functions in human malignancies. Among them, studies reported that HCG11 was abnormally elevated in liver cancer and promoted the progressions of liver cancer via modulating the insulin-like growth factor 2 mRNA binding protein 1 (IGF2BP1) [18]. In addition, HCG11 was negatively correlated to the prognosis of gastric cancer patients [10], suggesting HCG11 functions as an oncogene in gastric cancer. However, the specific roles of HCG11 in colorectal cancer are largely unknown.

The “Warburg effect” describes an aberrant glucose metabolism feature of cancer cells that they demand elevated glucose supply and metabolize the majority of glucose into lactate [19]. Moreover, reprogramming of glucose metabolism renders cancer cells advantages for survival under chemotherapeutic pressures [20]. Thus, targeting the dysregulated glucose metabolism could be a potential strategy to effectively treat cancer patients [20]. It was known that the pyruvate dehydrogenase kinase 4 (PDK4), which negatively regulates the convert of pyruvate to acetyl-CoA [21], plays important roles in glucose metabolism and cancer progressions. In this study, the roles of HCG11 in glucose metabolism and 5-FU resistance of colon cancer will be evaluated. Targets of HCG11 will be explored. These results manifested important roles of the HCG11-miR-144-3p-PDK4 axis in 5-FU resistance, offering novel insights for CRC patient treatment.

2. Materials and methods

2.1 Collection of CRC patient specimens

A total of forty paired human colon tumor specimens and adjacent normal tissues were obtained from CRC patients during surgery in the Department of Chinese PLA General Hospital (Beijing, China) from July 2016 to August 2018. This study was approved by the Institutional Review Board of Chinese PLA General Hospital. Written informed consent from patients was provided. CRC patients did not receive other local or systemic therapies prior to surgery. Patient samples were examined by three independent pathologists for the diagnosis of CRC cancers. After dissection, tissues were frozen immediately in liquid nitrogen and stored at $-$ 80 ${}^{\circ}$ C freezer until use.

2.2 Cell culture and reagents

Human colorectal cancer cell lines DLD-1, LoVo, Caco2, HT-29, HT-116 and SW480, as well as normal colon epithelial cell FHC were obtained from the American Type Culture Collection (ATCC). Cells were maintained in RPMI-1640 medium (Invitrogen, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Waltham, MA, USA), 50 U/ml Penicillin and 50 $\mu$ g/ml Streptomycin liquid (Invitrogen, Waltham, MA, USA) at 37 ${}^{\circ}$ C with 5% CO ${}_{2}$ atmosphere. The 5-FU resistant colon cancer cell line was established according to the previous reports [22]. DLD-1 cells were exposed to gradient concentrations of 5-FU for 6 months. Survival cell clones were collected and pooled for downstream experiments. Resistant cells were re-selected every three month by 5-FU treatment at the IC50 concentration. 5-fluorouracil reagent was purchased from Sigma (Shanghai, China). Western blot rabbit antibody against PDK4 (ab71240) was purchased from Abcam (Cambridge, UK) and rabbit antibody against $\beta$ -actin (#4970) was purchased from Cell Signaling Technology Inc. (Danvers, MA, USA).

2.3 Bioinformatics analysis

Bindings between HCG11 and miR-144-3p, miR-144-3p and PDK4 3’UTR were predicted from the starBase 2.0 of ENCORI: http://starbase.sysu.edu.cn/ [23]. Expression levels of miR-144-3p in colon tumor or normal colon tissues were analyzed from the ualcan.path.uab.edu [24].

2.4 Transfections of miRNA, siRNA and plasmid DNA

Colon cancer cells were seeded into 6-well plates at density of 5 $\times$ 10 ${}^{5}$ cells/well for 24 hours to obtain 60–70% cell confluence. The siRNAs against HCG11 and negative control were obtained from Genechem (Shanghai, China). The sequences of HCG11 and negative control are: 5’-UUCUCCGAACGUGUCACGUTT-3’ and 5’-GCCAGAAUGUUCCUAUUUATT-3’. The miR-144-3p precursor and control miRNA were constructed by Genepharma (Shanghai, China). The sequence of precursors and mature hsa-miR-144-3p is available in miRBase (http://www.mirbase.org/) using the accession number MIMAT0000436. PDK4 plasmid was purchased from Origene.org. Transfections were performed using the Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Plasmid DNA was transfected at 2 $\mu$ g/well. MiRNA and siRNAs were transfected at 25 nM. Forty-eight hours post transfection, cells were collected for downstream analysis.

2.5 RNA isolation and qRT-PCR

Total RNA was extracted from CRC tissues and cells using TRIzol Reagent (Invitrogen). Quality and quantity of RNA samples were assessed by Nanodrop. Complementary DNA (cDNA) was synthesized from extracted total RNA (1 $\mu$ g) using the PrimeScript™RT-PCR Kit (Takara) according to the manufacture’s instruction. Quantitative PCR was carried out with the Taqman 2X Universal PCR Master Mix (Thermo Fisher Scientific, Shanghai, China) on an ABI Prism 7900HT (Applied Biosystems, Foster City, CA, USA). The thermal profile was: 95 ${}^{\circ}$ C for 1 min and 40 cycles at 95 ${}^{\circ}$ C for 15 s, 58 ${}^{\circ}$ C for 20 s, and 72 ${}^{\circ}$ C for 20 s. Primers used for qRT-PCR were: HCG11, forward 5’-AGG AGT GGTTGC ATT TGG GA-3’, reverse 5’-CCC ACC ACG CAG TGA ATA GT-3’; PDK4, forward 5’-CCC GCTGTCCATGAAGCAGC-3’, reverse 5’-CCAATGTGGCTTGGGTTTCC-3’; $\beta$ -actin, forward 5’-CTGAGAGGGAAATCGTGCGT-3’, reverse 5’-CCACAGGATTCCATACCCAAGA-3’; miR-144-3p, forward 5’-GGCCGGCGTACAGTATAGATGA-3’, reverse 5’-GTGCAGGGTCCGAGGT-3’; U6, forward 5’-GCTTCGGCAGCACATATACTAAAAT-3’, reverse 5’-CGCTTCACGAATTTGCGTGTCAT-3’. All primers were synthesized from Sigma-Aldrich (Shanghai, China). Expression levels of PDK4 and HCG11 were quantified to the housekeeping genes $\beta$ -actin. Expression levels of miR-144-3p were normalized to human U6 which is used as an internal control. Relative expression level was calculated by the 2 ${}^{-\Delta\Delta\text{Ct}}$ method. Experiments were performed in triplicate and repeated three times.

2.6 RNA pull-down assay

The RNA pull-down assay was performed according to previous reports [25]. Briefly, HCG11 and its antisense RNA as well as a scramble control were biotin-labeled from RiboBio Co. Ltd (Guangzhou, China). Probs were incubated with CRC cell lysate for 2 hours. Mixture was incubated with Streptavidin-coupled agarose beads (Thermo Fisher Scientific, Shanghai, China) for 2 hours to down the RNA-RNA complexes. The amount of miR-144-3p in the RNA-RNA complexes was measured by qRT-PCR. Experiments were repeated three times.

2.7 Luciferase assay

Cells (5 $\times$ 10 ${}^{4}$ ) were seeded in 24-well plates for 24 hours. Cells were co-transfected with control miRNA or miR-144-3p plus pGL3-reporter luciferase reporter containing wild-type HCG11 or PDK4 3’UTR or binding site mutant HCG11 or PDK4 3’UTR using lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). Forty-eight hours post-transfection, luciferase activity was determined using a Dual luciferase reporter assay system (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Relative luciferase activity was normalized to the ratio of Firefly to Renilla. Experiments were performed in triplicate and repeated three times.

2.8 MicroRNA microarray analysis

Total RNA was extracted using the RNeasy Kits (Qiagen) followed by purification using the RNeasy column (Qiagen) with a modified protocol. After quantification, RNAs were labeled by the Exiqon miRCURY LNATM arrays, Hy3/Hy5 Power Labeling Kit which enzymatically adds a fluorescent label to the 3’-end of the miRNAs, followed by hybridization onto an microRNA array in a SureHyb hybridization chamber (Agilent) according to the manufacturer’s protocols. Images were obtained using an Agilent microarray scanner. The Bioconductor package LIMMA and R program were performed to analyze the raw data and statistical analysis. The background-corrected intensity data were then normalized. MiRNAs that were up or down regulated more than 1.5-fold were selected as potential targets. Heatmap of the up or downregulated miRNAs was created using the GraphPad Prism 7 software (La Jolla, CA, USA).

2.9 Measurement of glucose metabolism

The glucose metabolism rate was assessed via glucose uptake assay, extracellular acidification rate (ECAR), and oxygen consumption (OCR). The glucose uptake was determined using a glucose test kit (Applygen Technologies, Beijing, China) according to the manufacturer’s instruction. The ECAR and OCR were measured using the Seahorse XF Glycolysis Stress Test Kit, XF Cell Mito Stress Test Kit and Seahorse XF96e analyzer (Agilent, Santa Clara, CA, USA) according to the manufacturer’s instructions. Results were normalized according to cell protein amount of each group. Experiments were performed in triplicate and repeated three time.

2.10 Wound healing assay

CRC cell motility capacity was determined by wound healing assay. Colon cancer cells (1 $\times$ 10 ${}^{6}$ cells/well) were seeded into 12-well plates at 37 ${}^{\circ}$ C for 24 hours. A straight line was scratched by a P200 pipette tip. Medium was replaced and cultured for 16 hours. Cell migration was observed under a microscope. The distance of cell migration was calculated as the ratio of distance from experiment group to that from control group. Experiments were performed in triplicate.

2.11 Transwell assay

Transwell chambers (Invitrogen, Waltham, MA, USA) were used to determine cell invasive capacity. Cell culture medium (0.5 ml) supplemented with 15% FBS was added to the center of chambers. Colon cancer cells (5 $\times$ 10 ${}^{3}$ ) with normal cell culture medium were suspended onto the inserts. After sixteen hours, cells on the upper surface of the membrane were wiped off. Cells on the back of the inserts were fixed by methional and stained with DAPI. The number of invaded cells was counted under a microscope. Experiments were repeated three times.

2.12 Cell viability assay

Cell viability in response to 5-FU treatment was determined by MTT assay (Sigma-Aldrich, Shanghai, China) according to the manufacturer’s instructions. Briefly, colon cancer cells (1 $\times$ 10 ${}^{3}$ cells/well) were seeded into 96-well plate and cultured with regular medium for 24 hours. Forty-eight hours post 5-FU treatment, 20- $\mu$ l MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide) solution was added into each well. Cells were incubation at 37 ${}^{\circ}$ C for 4 h. Dimethyl sulfoxide (DMSO) (100 $\mu$ l) was added to dissolve the formed formazan and the mixture was incubated for 1 h. Absorbance was measured at 450 nm using a SpectraMax M5 microplate reader (Molecular Devices, Sunnyvale, CA, USA). The relative cell viability was calculated based on the ration of the absorbance from experiment group to that from control group. Experiments were performed in triplicate and repeated three times.

2.13 Cell apoptosis assay

Cell apoptosis/death rate of CRC cells in response to 5-FU treatment was determined using an Annexin V/FITC Apoptosis Detection Kit I (BD Biosciences, San Jose, CA, USA) according to the manufacturer’s instruction. After 5-FU treatments, cells were trypsinized and centrifuged at 2000 rpm for 3 minutes. Cells were washed by PBS then resuspended in binding buffer (300 ml). Cells were stained with 5 ml propidium iodide (PI) and 5 ml Annexin V FITC for 15 minutes in the dark. Data was obtained and analyzed using a FACScan flow cytometer and Flowjo software (BD Biosciences). Experiments were repeated three times.

Figure 1.

HCG11 is upregulated in colon cancer. (A) Expression levels of HCG11 from CRC tumors ( $n=$ 40) and adjacent normal colon tissues ( $n=$ 40) were detected by qRT-PCR. (B) Expression levels of HCG11 were detected in one normal colon epithelial cell, FHC and six CRC cancer cells by qRT-PCR. (C) Expression levels of HCG11 from 5-FU sensitive ( $n=$ 17) and resistant ( $n=$ 23) CRC tumors were detected by qRT-PCR. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01; ${}^{***}$ , $p<$ 0.001.

Figure 2.

HCG11 is positively associated with CRC progresses and 5-FU resistance. (A, B) HCG11 was knocked down by siRNA in CRC cells. Cell proliferation rates were examined in 0, 24, 48 and 72 hours. (C) The above transfected cells were subjected to wound healing assay and (D) transwell assay. (E) The glucose uptake, lactate product and (F) oxygen consumption rate were examined in DLD-1 cells without or with HCG11 knockdown. (G) DLD-1 and (H) LoVo cells without or with HCG11 knockdown were treated with 5-FU at the indicated concentrations for 48 hours. Cell viability was determined by MTT assay. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01.

2.14 Western blot

Total proteins were extracted from CRC cells using RIPA buffer (Thermofisher, Waltham, MA, USA) with 1 $\times$ protease inhibitors cocktail (Beyotime Institute of Biotechnology, Shanghai, China). After centrifugation at 10,000 g for 10 min, proteins were collected, and concentrations of protein were quantified by Bradford’s method. Equal amount of protein from each sample was subjected to SDS-PAGE and transferred to PVDF membranes. After blocking with 5% nonfat milk for 1 hour at room temperature, membranes were incubated with primary antibodies (1:1000) at 4 ${}^{\circ}$ C for overnight. Membranes were washed 3 $\times$ 10 min with PBST then incubated with HRP-conjugated secondary antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) for 1 hour. Protein bands were detected using the enhanced chemiluminescence kit (ECL kit; Santa Cruz Biotechnology, Inc.) $\beta$ -actin was used as an internal control. Experiments were repeated three times.

2.15 Statistical analysis

Data were analyzed using GraphPad Prism 7 software (La Jolla, CA, USA) and were displayed as mean $\pm$ standard deviation (SD). The analysis of differences between two groups was performed using Student’s $t$ -test and differences among multiple groups were analyzed by One-way analysis of variance (ANOVA) followed by Tukey’s test. Survival curve of patients was analyzed by Kaplan-Meier plot from ualcan.path.uab.edu. The linear relationships between miR-144-3p and HCG11, miR-144-3p and PDK4, HCG11 and PDK4 were analyzed using Pearson’s correlation coefficient. $p<$ 0.05 was considered to be statistically significant. All results were repeated three independent times.

Figure 3.

5-FU resistant colon cancer cells display upregulated glucose metabolism and HCG11. (A) The 5-FU sensitivities of DLD-1 parental and 5-FU resistant cells were determined by MTT assay and (B) clonogenic assay. (C) HCG11 expression levels were detected in DLD-1 parental and 5-FU resistant cells by qRT-PCR. (D) The glucose uptake, (E) extracellular acidification rate (ECAR) and (F) oxygen consumption rate (OCR) of DLD-1 parental and 5-FU resistant cells were measured. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01; ${}^{***}$ , $p<$ 0.001.

3. Results

3.1 HCG11 is upregulated in colon cancer

Previous studies uncovered oncogenic roles of HCG11 in multiple cancers [10, 11, 12, 13, 14, 15, 16, 17, 18]. Yet, the precis functions of HCG11 in colon cancer have not been elucidated. To evaluate the biological roles of HCG11 in colon cancer, expression levels of HCG11 in 40 paired CRC and adjacently non-tumorous tissues were examined. Results from qRT-PCR demonstrated that HCG11 was significantly elevated in colon cancer tissues (Fig. 1A). Consistently, HCG11 was apparently upregulated in six CRC cancer cell lines compared with normal colon epithelial cell line, FHC (Fig. 1B). In addition, HCG11 was detected to be significantly upregulated in 5-FU resistant CRC tumor specimens compared with those from 5-FU sensitive CRC tumor tissues (Fig. 1C). Summarily, these results suggest that HCG11 plays oncogenic roles and is positively associated with 5-FU resistance of colon cancer.

Table 1
Characteristics of CRC patients enrolled in this study

Characteristic	Low NLR		High NLR
	NLR $<$ 1.94 ( $n=$ 21)		NLR $\geqslant$ 1.94 ( $n=$ 19)
Age (yr)	61.2 $\pm$ 8.8		59.3 $\pm$ 10.7
Gender
Male	11	(52.38%)	9	(47.36%)
Female	10	(47.62%)	10	(52.63%)
Stage (T)
T3	19	(90.48%)	15	(79.94%)
T4	2	(9.52%)	4	(21.06%)
Stage (N)
N0	5	(23.81%)	2	(10.53%)
N1	11	(52.38%)	8	(42.1%)
N2	5	(23.81%)	9	(47.37%)
Histology
Well	7	(33.33%)	5	(26.32%)
Moderate	12	(57.14%)	10	(52.63%)
Poorly	2	(9.52%)	4	(21.05%)
Tumor size
$\leqslant$ 3 cm	15	(71.43%)	5	(26.32%)
$>$ 3 cm	6	(28.57%)	14	(73.68%)
Metastasis
No	15	(71.43%)	6	(31.58%)
Yes	6	(28.57%)	13	(68.42%)

Figure 4.

HCG11 sponges miR-144-3p to downregulate its expression. (A) DLD-1 cells without or with HCG11 silencing were subjected to microRNA microArray analysis. (B) Results from microArray were validated by qRT-PCR in DLD-1 and LoVo cells. (C) Prediction of the binding of HCG11 with miR-144-3p from StarBase. (D) Pearson’s correlation coefficient analysis illustrated a negative correlation between HCG11 and miR-144-3p in CRC tumor specimens. (E) DLD-1 and LoVo cells were transfected with HCG11 siRNA for 48 hours, expression levels of miR-144-3p were detected by qRT-PCR. (F) Biotin pull-down assay was performed in DLD-1 and LoVo cells. Expression levels of miR-144-3p was assessed by qRT-PCR. (G) Dual-luciferase reporter assay was performed in DLD-1 and (F) LoVo cells with co-transfection of WT-HCG11 or Mut-HCG11 with control miRNA or miR-144-3p. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01; ${}^{***}$ , $p<$ 0.001.

Figure 5.

miR-144-3p is negatively associated with CRC progresses and 5-FU resistance. (A) Expression levels of miR-144-3p in normal or CRC specimens were analyzed from UALCAN. (B) Expression levels of miR-144-3p were detected in CRC tumor specimens ( $n=$ 40) and adjacent normal tissues ( $n=$ 40) by qRT-PCR. (C) Expression levels of miR-144-3p were detected in normal colon epithelial cell line and six colon cancer cells by qRT-PCR. (D) Expression levels of miR-144-3p were detected in DLD-1 parental and 5-FU resistant cells by qRT-PCR. (E) DLD-1 and (F) LoVo cells were transfected with control miRNA or miR-144-3p for 48 hours. Cells were treated with 5-FU at the indicated concentrations for 48 hours. Cell viabilities were determined by MTT assay. (G) Glucose uptake, (H) ECAR and (I) OCR were measured in DLD-1 cells without or with miR-144-3p overexpression. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01; ${}^{***}$ , $p<$ 0.001.

Figure 6.

miR-144-3p directly targets PDK4 in CRC cells. (A) Prediction of miR-144-3p binding sites on 3’UTR of PDK4 from StarBase. (B) mRNA expression levels of PDK4 in CRC tumor specimens and adjacent normal colon tissues were detected by qRT-PCR. (C) DLD-1 and (D) LoVo cells were transfected with control siRNA or siPDK4 for 48 hours. Cell viabilities in response to 5-FU treatments were determined by MTT assay. (E) Pearson’s correlation coefficient analysis revealed a negative correlation between miR-144-3p and PDK4 mRNA in CRC tumor specimens. (F) DLD-1 and LoVo cells were transfected with control miRNA or miR-144-3p precursor for 48 hours, protein levels of PDK4 were determined by Western blot. (G, H) Dual-luciferase reporter assay was performed in CRC cells by co-transfection of luciferase vector containing 3’UTR of WT-PDK4 or Mut-PDK4 with control miRNA or miR-144-3p. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01; ${}^{***}$ , $p<$ 0.001.

3.2 HCG11 promotes proliferation, invasion, glycolysis, and 5-FU resistance of colon cancer cells

To investigate the direct functions of HCG11, LoVo and DLD-1 cells were transfected with siHCG11 to specifically knockdown its expressions (Fig. S1). Expectedly, CRC cells with HCG11 silencing displayed remarkedly inhibited cell proliferation rate (Fig. 2A and B), migration and invasion capacities (Fig. 2C and D). Given the dysregulated glucose metabolism affected the chemosensitivity of tumors [19], we investigated the roles of HCG11 in glucose metabolism of colon cancer cells. The glucose uptake and lactate production, two readouts for glycolysis reactions from HCG11-low expressing cells, were attenuated (Fig. 2E). Moreover, the oxygen consumption rate (OCR), which reflects the overall mitochondrial functions, was significantly increased in HCG11-silenced cells (Fig. 2F), suggesting HCG11 drives a glucose metabolic switch from oxidative phosphorylation to anerobic glycolysis in colon cancer cells. Given the acquired chemoresistance is a hallmark of cancer cells, we examined the effects of HCG11-knockdown in 5-FU sensitivity of CRC cells. Consistent with the above results, silencing HCG11 significantly sensitized CRC cells to 5-FU (Fig. 2G and H). The 5-FU IC50s of DLD-1 and LoVo control cells were 6.16 $\mu$ M and 2.83 $\mu$ M, respectively. DLD-1 and LoVo cells with HCG11 knockdown showed significantly dropped IC50s to 1.93 $\mu$ M and 0.65 $\mu$ M, respectively (Fig. 2G and H). These results suggest targeting HCG11 might be an effective approach for colon cancer therapy.

3.3 5-FU resistant CRC cells display upregulated HCG11 and glucose metabolism reprogramming phenotypes

We then explored the underlying mechanisms of the HCG11-promoted 5-FU resistance. A 5-FU resistant cell line (5-FU Res) was established from DLD-1 by selecting survival cells under stepwise increased concentrations of 5-FU treatments for 6 months. The DLD-1 5-Fu resistant cells displayed higher survival rate under 5-FU treatments compared with parental cells detected by cell viability assay and clonogenic assay (Fig. 3A and B). The 5-FU IC50 of DLD-1 5-FU Res cells was 24.4 $\mu$ M, which is four-fold higher than that of DLD-1 parental cells (6.1 $\mu$ M) (Fig. 3A). Expectedly, we observed significant upregulation of HCG11 in 5-FU resistant cells compared with parental cells (Fig. 3C). Moreover, the glucose uptake (Fig. 3D) and extracellular acidification rate (Fig. 3E) were remarkedly elevated in DLD-1 5-FU Res cells. On the contrary, the oxygen consumption rate (OCR) of 5-FU resistant cells was significantly attenuated (Fig. 3F), suggesting the glucose metabolism switch of CRC cells contributes to 5-FU resistance.

3.4 HCG11 downregulates miR-144-3p in CRC cells via sponging it as a ceRNA

Considering that lncRNAs function through sponging miRNAs to block their expressions [8, 9], we investigated the molecular targets of HCG11 in 5-FU resistant colon cancer. A microRNA-microArray analysis was performed. Results in Fig. 4A demonstrated that among the miRNAs which were differentially regulated in HCG11-silencing DLD-1 cells, miR-144-3p was detected to be upregulated by HCG11 knockdown. The microArray results were further validated by qRT-PCR in DLD-1 and LoVo cells (Fig. 4B). Moreover, bioinformatics analysis showed miRNA-144-3p contains putative HCG11 binding sites (Fig. 4C), suggesting HCG11 might downregulate miR-144-3p via direct binding it to form a ceRNA network. To assess the clinical correlation between HCG11 and miR-144-3p, Pearson’s correlation coefficient analysis was performed. Expectedly, a significantly negative correlation between HCG11 and miR-144-3p was observed in CRC tumor specimens (Fig. 4D). Furthermore, DLD-1 and LoVo cells with lower HCG11 expression levels displayed significantly inhibited miR-144-3p expressions (Fig. 4E). To confirm the binding of HCG11 on miR-144-3p, an RNA-pull down assay was performed. qRT-PCR results clearly demonstrated that the biotin-labeled antisense HCG11 probe effectively pulled down enriched amount of miR-144-3p from DLD-1 and LoVo cells, compared with those from HCG11 sense RNA probe and scramble control (Fig. 4F). To validate whether HCG11 directly binds on miR-144-3p, luciferase assays were performed by co-transfection of CRC cells with luciferase vector containing wild type HCG11 (WT-HCG11) or binding site mutant HCG11 (Mut-HCG11) plus control miRNA or miR-144-3p. Expectedly, luciferase activities of CRC cells with co-transfection of WT-HCG11 plus miR-144-3p were significantly suppressed compared with that from the Mut-HCG11 plus miRNA-144-3p co-transfection (Fig. 4G and H). Taken together, the above results elucidated that HCG11 downregulates miR-144-3p in CRC cells through sponging it as a ceRNA.

3.5 Overexpression of miR-144-3p sensitizes colon cancer cells to 5-FU

To determine the effects of miR-144-3p on the 5-FU sensitivity in colorectal cancer, we examined miR-144-3p expression levels in CRC tumors specimens and normal colon tissues. As we expected, miR-144-3p was significantly downregulated in colon tumors (Fig. 5A and B). In addition, downregulated miR-144-3p was observed in six CRC cancer cell lines compared with normal colon epithelial cell line, FHC (Fig. 5C). MiR-144-3p was detected to be remarkedly low-expressed in DLD-1 5-FU resistant cells (Fig. 5D), suggesting miR-144-3p plays suppressive roles in 5-FU resistant colon cancer. Consequently, the effects of miR-144-3p overexpression on the 5-FU sensitivity were evaluated. CRC cells with miR-144-3p overexpression exhibited significantly increased 5-FU sensitivity (Fig. 5E and F). Moreover, overexpression of miR-144-3p effectively blocked the glucose uptake (Fig. 5G), ECAR (Fig. 5H) and OCR (Fig. 5I) in colon cancer cells. Summarily, the above results uncovered a tumor suppressive role of miR-144-3p in CRC cells. Moreover, overexpression of miR-144-3p contributes to sensitization of CRC cells to 5-FU.

Figure 7.

Rescue of PDK4 overcomes the miR-144-3p-attenuated glucose metabolism and 5-FU sensitization in CRC cells. (A) DLD-1 cells were transfected with control, miR-144-3p alone or plus PDK4, expression levels of PDK4 were detected by Western blot. (B) Glucose uptake, (C) lactate product and (D) oxygen consumption were measured from the above transfected cells. (E) The above transfected cells were treated with 5-FU for 48 hours, 5-FU sensitivity was determined by MTT assay and (F) Annexin V apoptosis assay. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01.

Figure 8.

The HCG11-miR-144-3p-PDK4 axis in 5-FU resistance. (A) Correlation between HCG11 and PDK4 expression levels was analyzed from CRC tumor specimens and adjacent normal colon tissues. (B) DLD-1 5-FU Res cells were transfected with control, HCG11 alone or combined with miR-144-3p for 48 hours, expression levels of PDK4 were detected by Western blot. (C) Glucose uptake, (D) lactate product and (E) oxygen consumption from the above transfected cells were examined. (F) The above transfected cells were treated with 5-FU at the indicated concentrations, 5-FU sensitivity was determined by Annexin V apoptosis assay. ${}^{*}$ , $p<$ 0.05; ${}^{**}$ , $p<$ 0.01.

3.6 miR-144-3p directly targets PDK4 to regulate glucose metabolism and 5-FU resistance of CRC cells

Accumulating studies reported that miRNA functions through binding to the 3’UTR of their target mRNAs, leading to blocking the mRNA transcriptions [26]. We then analyzed the mRNA targets of miR-144-3p from the non-coding RNA database, starBase2.0. Interestingly, we observed the 3’UTR of PDK4, which regulates the reaction of pyruvate to lactate or acetyl-CoA during the glucose metabolism [21], contains a putative miR-144-3p binding sites (Fig. 6A). We then assessed the clinical relevance of PDK4 in colon cancer. Expression level of PDK4 was significantly upregulated in CRC tumor specimens compared with normal colon tissues by qRT-PCR (Fig. 6B). To verify the biological roles of PDK4 in 5-FU sensitivity of CRC cells, PDK4 was silenced by siRNA in DLD-1 and LoVo cells. Expectedly, CRC cells with PDK4 silencing were more sensitive to 5-FU compared with control cells (Fig. 6C and D). We then evaluated the clinical correlation between miR-144-3p and PDK4. The Pearson’ correlation coefficient analysis demonstrated that higher PDK4 expression level was significantly associated with lower miR-144-3p expression level in CRC tumor specimens (Fig. 6E), supporting that miR-144-3p targets PDK4 in colon cancer. Consistently, overexpression of miR-144-3p effectively downregulated PDK4 protein expressions in CRC cells (Fig. 6F). To further validate whether miR-144-3p could target the 3’UTR of PDK4, luciferase reporter assays were performed by co-transfection of luciferase vector containing 3’UTR of WT-PDK4 or binding site mutant-PDK4 with control miRNA or miR-144-3p into CRC cells. Expected results demonstrated overexpression of miR-144-3p significantly blocked the luciferase activity of vector containing WT-PDK4 3’-UTR but not that of the binding site mutant PDK4 3’-UTR (Fig. 6G and H).

To investigate whether the miR-144-3p-modulated glucose metabolism switch and 5-FU sensitization were through targeting PDK4, rescue experiments were conducted. Transfection of PDK4 in miR-144-3p-overexpressing DLD-1 cells successfully rescued PDK4 protein expression level (Fig. 7A). Expectedly, restoration of PDK4 in miR-144-3p-overexpressing CRC cells significantly recovered the glucose uptake (Fig. 7B), lactate product (Fig. 7C) and oxygen consumption (7D). Furthermore, DLD-1 cells with PDK4 rescue showed decreased 5-FU sensitivity compared with those from cells with miR-144-3p overexpress alone by MTT assay and Annexin V apoptosis assay (Fig. 7E and F). Taken together, these data validated that miR-144-3p sensitizes CRC cells to 5-FU through targeting the PDK4-glucose metabolism pathway.

3.7 Inhibition of HCG11 sensitizes 5-FU resistant CRC cells via the miR-144-3p-PDK4 axis

We evaluated whether the HCG11-promoted 5-FU resistance was through the miR-144-3p-PDK4-glycolysis axis. Bioinformatics analysis from the lncRNA database, Starbase, indicated HCG11 expression level is positively associated with PDK4 in colon cancer (Fig. S2). Meanwhile, qRT-PCR results clearly verified that CRC patients with higher HCG11 expression level were associated with PDK4 upregulation (Fig. 8A). Thus, mechanism rescue experiments were performed by co-transfection of DLD-1 5-FU Res cells with control, HCG11 overexpression plasmid alone or plus miR-144-3p. Western blot results showed that overexpression of HCG11 significantly upregulated PDK4 expression level, which was further blocked by co-transfection of miR-144-3p (Fig. 8B). Expectedly, co-transfection of HCG11 with miR-144-3p recovered the glucose uptake (Fig. 8C), lactate product (Fig. 8D) and oxygen consumption (Fig. 8E). Consequently, under 5-FU treatments, we detected significant recovery of 5-FU sensitivity in miR-144-3p-rescued DLD-1 5-FU Res cells compared with that from cells with HCG11 overexpression alone (Fig. 8F). Summarily, the above results consistently demonstrated that the HCG11-promoted 5-FU resistance was through modifying the miR-144-3p-PDK4-glucose metabolism axis.

4. Discussion

Colorectal cancer is considered one of the most common human malignancies, leading to high death rate worldwide [1]. In the last two decades, the overall prognosis and survival rates of CRC patients have remarkedly improved, due to early screening and advanced therapeutic approaches [2, 3]. Despite 5-Fluoropyrimidine has been widely used in clinical treatment of CRC, development of intrinsic and acquired chemoresistance limits its clinical applications [4, 5]. Currently, while a number of studies described multiple molecules which are responsible for 5-FU resistance [6, 7], the precise mechanisms of 5-FU resistance in CRC are still under investigation. Therefore, understanding the precise mechanism of 5-FU resistance is critical for developing therapeutic strategies for colon cancer patients. This study uncovered non-coding RNA-modulated molecular targets and mechanisms of 5-FU resistance in colon cancer.

Accumulating evidence revealed that lncRNAs play vital roles in the occurrence and development of cancers [9]. However, the biological roles and mechanisms of HCG11 in colon cancer have not been investigated. Studies reported that HCG11 was dysregulated in variously malignant tumors [10, 11, 12, 13, 14, 15, 16, 17, 18]. Interestingly, HCG11 functioned as either oncogene or tumor suppressor in these studies due to tissue specific expressions [10, 11, 12, 13, 14, 15, 16, 17, 18]. Our present study for the first time showed that HCG11 was significantly upregulated in colon tumors and cell lines. Moreover, HCG11 expression level was positively associated with 5-FU resistance of CRC patients, suggesting HCG11 plays an oncogenic role in CRC. Silencing HCG11 effectively sensitized CRC cell to 5-FU. Consistent results demonstrated that HCG11 was significantly upregulated in 5-FU resistant colon cancer cells compared with parental CRC cells, indicating

HCG11 is a potential biomarker and therapeutic target of CRC.

Mounting studies reported that lncRNAs acted as sponges of miRNAs to downregulate their expressions [27]. Here we report miR-144-3p, which acts as a tumor suppressor in colon cancer cells, was a target of HCG11 in CRC cells. MicroRNA-microArray and RT-qPCR assay consistently demonstrated that overexpression of HCG11 significantly downregulated miR-144-3p expression levels. Subsequently, RNA-pull down and luciferase assay validated the direct bind of HCG11 with miR-144-3p in CRC cells. We further identified the glucose metabolism key enzyme, PDK4, as a direct mRNA target of miR-144-3p in CRC cells. Rescue experiments clearly verified the miR-144-3p-promoted 5-FU sensitization was through direct targeting PDK4.

Reprogramming of glucose metabolism characterized with increased anaerobic glycolysis and suppressed oxidative phosphorylation, a phenomenon called “Warburg effect”, was a new hallmark of cancer [19]. In addition, accumulation evidence manifested that dysregulated glucose metabolism is tightly correlated with chemoresistance [20], suggesting targeting glucose metabolism could be an effectively anti-cancer approach. From the established 5-FU resistant CRC cell line, we observed significantly upregulated glycolysis rate and decreased mitochondrial oxidative phosphorylation in 5-FU resistant cells. Consequently, HCG11 was detected to be upregulated and miR-144-3p was suppressed in 5-Fu resistant CRC cells compared with parental cells. Importantly, mechanisms rescue results showed silencing HCG11 suppressed glycolysis and recovered oxidative phosphorylation of CRC cells through regulating the miR-144-3p-PDK4 axis, suggesting targeting the HCG11-promoted glucose metabolism pathway could potentially overcome chemoresistance of CRC. This study still has limits that the current molecular mechanisms of the HCG11-promoted glucose metabolism and 5-FU resistance were assessed only by in vitro cell model. The in vivo xenograft mice experiments which could verify the in vitro results were lacking. In addition, since lncRNA and miRNA could target multiple molecules, we cannot exclude other signaling pathways which are potentially regulated by HCG11 or miR144-3p in the 5-FU sensitivity of CRC cells.

In summary, we demonstrate that lncRNA HCG11 plays oncogenic roles in 5-FU resistant CRC. Mechanically, HCG11 sponges miR-144-3p as a ceRNA, leading to de-repression of the miR-144-3p target, PDK4 in colon cancer. Our discovery will contribute to the development of novel approaches for treatment of chemoresistant CRC.

Authors’ contributions

Quan-Li Han: Investigation, Methodology, Writing – review& editing; Mu-Hong Deng and Zhi Cui: Investigation, Methodology; Qi Wang: Statistical Analysis.

Supplementary data

The supplementary files are available to download from https://dx-doi-org.web.bisu.edu.cn/10.3233/CBM-210212.

sj-docx-1-cbm-10.3233_CBM-210212.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-210212.docx

Footnotes

Acknowledgments

None.

Conflict of interest

The authors declare no competing financial and non-financial interests.

References

Udayasiri

D.K.

Skandarajah

and Hayes

I.P.

, Laparoscopic compared with open resection for colorectal cancer and long-term incidence of adhesional intestinal obstruction and incisional hernia: A systematic review and meta-analysis, Dis Colon Rectum 63 (2020), 101–112.

Hangaard Hansen

Gögenur

Tvilling Madsen

and Gögenur

, The effect of time from diagnosis to surgery on oncological outcomes in patients undergoing surgery for colon cancer: A systematic review, Eur J Surg Oncol 44 (2018), 1479–1485.

Tang

Price

T.J.

Shapiro

Gibbs

Haller

D.G.

Arnold

et al., Adjuvant therapy for resected colon cancer 2017, including the IDEA analysis, Expert Rev Anticancer Ther 18 (2018), 339–349.

McQuade

R.M.

Stojanovska

Bornstein

J.C.

and Nurgali

, Colorectal cancer chemotherapy: The evolution of treatment and new approaches, Curr Med Chem 24 (2017), 1537–1557.

Wei

Yang

Cao

and Zhao

, The combination of curcumin and 5-fluorouracil in cancer therapy, Arch Pharm Res 41 (2018), 1–13.

Vodenkova

Buchler

Cervena

Veskrnova

Vodicka

and Vymetalkova

, 5-fluorouracil and other fluoropyrimidines in colorectal cancer: Past, present and future, Pharmacol Ther 206 (2020), 107447.

Marjaneh

R.M.

Khazaei

Ferns

G.A.

Avan

and Aghaee-Bakhtiari

S.H.

, The role of microRNAs in 5-FU resistance of colorectal cancer: Possible mechanisms, J Cell Physiol 234(2019), 2306–2316.

Jarroux

Morillon

and Pinskaya

, History, discovery, and classification of lncRNAs, Adv Exp Med Biol 1008 (2017), 1–46.

Bhan

Soleimani

and Mandal

S.S.

, Long noncoding RNA and cancer: A new paradigm, Cancer Res 77(2017), 3965–3981.

10.

Zhang

Huang

Wang

Yao

et al., LncRNA HCG11 promotes proliferation and migration in gastric cancer via targeting miR-1276/CTNNB1 and activating Wnt signaling pathway, Cancer Cell Int 19 (2019), 350.

11.

Zhang

Wan

Kong

Zhai

et al., Downregulation of long non-coding RNA HCG11 predicts a poor prognosis in prostate cancer, Biomed Pharmacother 83 (2016), 936–941.

12.

Wang

Y.C.

W.Y.

Dong

C.H.

Pei

and Ma

Y.L.

, lncRNA HCG11 regulates cell progression by targeting miR-543 and regulating AKT/mTOR pathway in prostate cancer, Cell Biol Int (2019).

13.

X.F.

D.M.

Zhao

Y.X.

Zhang

and Jin

, Knockdown of lncRNA HCG11 suppresses cell progression in ovarian cancer by modulating miR-144-3p/PBX3, Eur Rev Med Pharmacol Sci 24 (2020), 11032–11040.

14.

Zhang

Mao

and Li

, Long noncoding RNA HCG11 inhibited growth and invasion in cervical cancer by sponging miR-942-5p and targeting GFI1, Cancer Med 9 (2020), 7062–7071.

15.

Zhang

Cao

Kou

Che

Zhou

Chen

et al., Long non-coding RNA HCG11 suppresses the growth of glioma by cooperating with the miR-4425/MTA3 axis, J Gene Med 21 (2019), e3074.

16.

Chen

Bao

Zhang

Lin

Huang

and Wang

, Long non-coding RNA HCG11 modulates glioma progression through cooperating with miR-496/CPEB3 axis, Cell Prolif 52 (2019), e12615.

17.

M.L.

Zhang

and Ma

L.T.

, LncRNA HCG11 accelerates the progression of hepatocellular carcinoma via miR-26a-5p/ATG12 axis, Eur Rev Med Pharmacol Sci 23 (2019), 10708–10720.

18.

Zheng

Liu

and Li

, Modulation of IGF2BP1 by long non-coding RNA HCG11 suppresses apoptosis of hepatocellular carcinoma cells via MAPK signaling transduction, Int J Oncol 51 (2017), 791–800.

19.

Taefehshokr

Hemmat

Hajazimian

Isazadeh

Dadebighlu

et al., The pivotal role of MicroRNAs in glucose metabolism in cancer, Pathol Res Pract 217 (2021), 153314.

20.

Bhattacharya

Mohd Omar

M.F.

and Soong

, The warburg effect and drug resistance, Br J Pharmacol 173 (2016), 970–979.

21.

Leclerc

Pham

D.N.

Lévesque

Truongcao

Foulkes

W.D.

Sapienza

et al., Oncogenic role of PDK4 in human colon cancer cells, Br J Cancer 116 (2017), 930–936.

22.

Jung

J.H.

Taniguchi

Lee

H.M.

Lee

M.Y.

Bandu

Komura

et al., Comparative lipidomics of 5-Fluorouracil-sensitive and -resistant colorectal cancer cells reveals altered sphingomyelin and ceramide controlled by acid sphingomyelinase (SMPD1), Sci Rep 10 (2020), 6124.

23.

J.H.

Liu

Zhou

L.H.

and Yang

J.H.

, starBase v2.0:Decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP-seq data, Nucleic Acids Res 42 (2014), D92–97.

24.

Chandrashekar

D.S.

Bashel

Balasubramanya

S.A.H.

Creighton

C.J.

Ponce-Rodriguez

Chakravarthi

B.V.S.K.

and Varambally

, UALCAN: A portal for facilitating tumor subgroup gene expression and survival analyses, Neoplasia 19 (2017), 649–658.

25.

Torres

Becquet

Guillen

Boyer

Moreno

Blanchard

M.P.

et al., RNA pull-down procedure to identify RNA targets of a long non-coding RNA, J Vis Exp (2018), 57379.

26.

Sarraf

J.S.

Puty

T.C.

da Silva

E.M.

Allen

T.S.R.

Sarraf

Y.S.

de Carvalho

L.E.W.

et al., Noncoding RNAs and colorectal cancer: A general overview, Microrna 9 (2020), 336–345.

27.

Zhang

Qian

Wang

et al., Analysis of lncRNA-associated ceRNA network reveals potential lncRNA biomarkers in human colon adenocarcinoma, Cell Physiol Biochem 49 (2018), 1778–1791.

LncRNA HCG11 promotes 5-FU resistance of colon cancer cells through reprogramming glucose metabolism by targeting the miR-144-3p-PDK4 axis

Abstract

BACKGROUND:

OBJECTIVE:

METHODS AND RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Collection of CRC patient specimens

2.2 Cell culture and reagents

2.3 Bioinformatics analysis

2.4 Transfections of miRNA, siRNA and plasmid DNA

2.5 RNA isolation and qRT-PCR

2.6 RNA pull-down assay

2.7 Luciferase assay

2.8 MicroRNA microarray analysis

2.9 Measurement of glucose metabolism

2.10 Wound healing assay

2.11 Transwell assay

2.12 Cell viability assay

2.13 Cell apoptosis assay

2.15 Statistical analysis

3.1 HCG11 is upregulated in colon cancer

Table 1 Characteristics of CRC patients enrolled in this study

3.3 5-FU resistant CRC cells display upregulated HCG11 and glucose metabolism reprogramming phenotypes

3.4 HCG11 downregulates miR-144-3p in CRC cells via sponging it as a ceRNA

3.5 Overexpression of miR-144-3p sensitizes colon cancer cells to 5-FU

3.7 Inhibition of HCG11 sensitizes 5-FU resistant CRC cells via the miR-144-3p-PDK4 axis

4. Discussion

Authors’ contributions

Supplementary data

sj-docx-1-cbm-10.3233_CBM-210212.docx - Supplemental material

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Characteristics of CRC patients enrolled in this study