Secondary Logo

Journal Logo

Research Articles

COL14A1 promotes self-renewal of human liver cancer stem cells through activation of ERK signaling

Kong, Ruijiaoa; Liu, Haidongb; Shi, Yia; Man, Qiuhonga; Liu, Shanronga,b,∗

Author Information
doi: 10.1097/JBR.0000000000000087
  • Open



Hepatocellular carcinoma (HCC), representing the major subtype of primary liver cancer, is one of the malignancies with the highest morbidity and mortality globally.[1] Due to insidious onset and the low effectiveness of screening methods, 60% to 70% of HCC patients are already in the advanced stage at the time of diagnosis, and their 5-year survival rate is less than 10%.[2] High recurrence and low survival, characteristic for HCC, are closely related to tumor heterogeneity.[3] It is believed that a subset of tumor cells, called cancer stem cells (CSCs), construct a hierarchical tumor of heterogeneous cancer cells.[4] CSCs share many characteristics with normal stem cells, including self-renewal ability and multi-directional differentiation potential. Importantly, CSCs play a crucial role in tumor deterioration and treatment resistance. Although many aberrantly expressed genes have been identified in liver CSCs in recent years,[5,6] the molecular basis of liver CSCs is still far from clearly understood.

Collagen Type XIV Alpha 1 Chain (COL14A1) is a member of the fibril-associated collagen family,[7] whose dysregulated deposition and metabolism have been shown to enhance tumor development.[8] Remarkably, collagen type XIV is specifically upregulated in human colorectal cancer compared with healthy colon.[9] Additionally, increased expression of collagen type XIV has been observed during the progression from liver fibrosis and steatohepatitis to HCC using mouse models.[10] However, research examining how COL14A1 functions in CSCs is currently lacking.

Here, our aim was to identify the role and underlying mechanism of aberrant COL14A1 overexpression in liver CSCs, with the goal of better understanding the molecular basis of HCC metastasis and recurrence.

Materials and methods

Oncomine platform

Oncomine ( is a large database of tumor gene chips with 715 datasets and 86,733 samples, integrating RNA and DNA-seq data from sources such as GEO, TCGA and published literature. The transcript levels of COL14A1 in various cancers were analyzed and assessed relative to that in normal tissue with the Oncomine 4.5 database.

Cell culture

Primary HCC cells (1#, 2#, 3#, 4#) were prepared from tumor tissues of four male HCC patients (50–70 years old) using a Human Tumor Dissociation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Informed consent was obtained from volunteers undergoing surgical resection in the Eastern Hepatobiliary Surgery Hospital (Shanghai, China). Four HCC cell lines (Hep3B, Huh7, HepG2, SMMC7721) and HEK293T cells were obtained from the Chinese Academy of Sciences Cell Bank (Beijing, China). For adherent cell culture, cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Corning, Corning, NY, USA) supplemented with 10% fetal bovine serum (Gibco, Gaithersburg, MD, USA) under 37°C and 5% CO2 conditions. For suspension CSC culture or tumorsphere formation, 1 × 105 tumor cells were trypsinized with Trypsin (0.25%; Gibco) as single-cell suspensions and seeded into 6-well plates treated with poly(2-hydroxyethyl methacrylate) (Poly-HEMA; Sigma, St. Louis, MO, USA) in DMEM/F12 (1:1) (Hyclone, Logan, UT, USA) with 20ng/mL insulin growth factor (IGF; PeproTech, Rocky Hill, NJ, USA), 1.0 ng/mL basic fibroblast growth factor (bFGF; PeproTech), and 20ng/mL epidermal growth factor (EGF; PeproTech) under 37°C and 5% CO2 conditions for 1 week. This study was approved by the Ethics Committee of Shanghai Fourth People's Hospital, Tongji University School of Medicine, China (approval No. 2019tjdx17) on February 21, 2019.

Short hairpin RNA (shRNA) plasmid construction

Sense and antisense sequences of 21-mer targets of the COL14A1 gene (Table 1; shCOL14A1#1 and shCOL14A1#2; Tsingke, Beijing, China) were inserted into oligos compatible with pLKO.1-puro empty vector (Sigma). A corresponding scrambled sequence (See Table 1, scramble; Tsingke) was used to construct the control plasmid. Next, annealed oligos were ligated into a linearized pLKO.1 vector (digested with AgeI and EcoRI). A screen for successfully ligated plasmids was performed using Sanger sequencing.[11]

Table 1 - Sequences of primers, shRNAs, and miRNAs
Name Sequence (5′–3′)
miR-7108-3p mimic Sense:ACCCGCCCGUCUCCCCACAG
miR-7108-3p inhibitor CUGUGGGGAGACGGGCGGGU
F = forward, R = reverse, snRNA = small nuclear RNA.

Cell transfection

LipofectamineTM 3000 (Invitrogen, Carlsbad, CA, USA) was used to transiently transfect plasmids and miRNA mimics or inhibitors of miR-7108-3p and negative control (GenePharma, Shanghai, China) into Huh7 and HEK293T cells at 60% to 80% confluency. In addition, lentivirus was used to establish cell lines with stable expression as previously described.[5] Cells were incubated with infectious lentiviruses for about 12 hours, and stable COL14A1-silenced cells were obtained after puromycin screening for 1 week.

RNA extraction, reverse transcription, and quantitative real-time polymerase chain reaction (qRT-PCR)

RNA was extracted from primary HCC cells and HCC cell lines using RNAiso Plus (Takara, Kyoto, Japan) and subsequently reverse transcribed with PrimeScriptTM RT Reagent Kit (Takara) and a miRcute miRNA First-Strand cDNA Synthesis Kit (Tiangen, Beijing, China). qRT-PCR was performed with TB Green Premix Ex Taq™ II (Takara) using specific primers (Table 1) by LightCycler 480 II (Roche, Basel, Switzerland). All operations were carried out in accordance with procedures recommended in the manufacturer's manual.

Western blotting

Soluble denatured proteins extracted from Hep3B and Huh7 cells were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 4% to 20% gel (EZBiolab, Carmel, IN, USA) and transfer to 0.45-μm polyvinylidene fluoride membrane (Millipore, Burlington, MA, USA). Membranes were blocked in Tris-buffered saline containing Tween-20 (TBST) and 5% bovine serum albumin (BSA) for 2 hours at room temperature, then incubated with the antibody of interest [rabbit anti-COL14A1 (PA5-54886, 1/500 dilution; Invitrogen), rabbit anti-NANOG (ab109250, 1/4000 dilution; Abcam, Cambridge, UK), rabbit anti-GAPDH (AB0037, 1/4000 dilution; Abways, Pleasanton, CA, USA), rabbit anti-ERK1/2 (ab184699, 1/5000 dilution; Abcam), or rabbit anti-Erk1 (pT202/pY204)/2 (pT185/pY187) (ab76299, 1/5000 dilution; Abcam) in 3% BSA-TBST buffer overnight at 4°C. After washing, membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (AS014, 1/5000 dilution; ABclonal, Woburn, MA, USA) at 4°C for 3 hours. Protein bands were visualized by enhanced chemiluminescence (ImageQuant LAS 4000; GE Healthcare, Piscataway, NJ, USA). Semi-quantitative analysis was performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Cell apoptosis detection

Hep3B and Huh7 cells at 80% to 90% confluency were trypsinized with Trypsin (0.25%; Gibco, Gaithersburg, MD, USA) into single cells to detect apoptotic cells with an Annexin V-FITC/PI Apoptosis Detection Kit (Elabscience, Wuhan, China). Fluorescence of annexin V bound to apoptotic cells was quantitatively determined by flow cytometry (BD FACSCantoTM II; BD Biosciences, San Jose, CA, USA).

Cell proliferation assay

A total of 4 × 104 Hep3B or Huh7 cells were seeded into 24-well plates in advance. When Hep3B or Huh7 cells reached about 60% confluency, a Click-iT™ EdU Cell Proliferation Assay Kit (Invitrogen) was used to detect proliferation according to the manufacturer's recommended protocol. Fluorescent images were obtained using OLYMPUS IX73 and acquisition software (Olympus, Tokyo, Japan).

Luciferase reporter assay

The 3′ untranslated region (3′UTR) of COL14A1 with a wild-type (WT) or mutant target site of miR-7108-3p was cloned into the pmirGLO vector (Tsingke) to generate a WT or mutated pmirGLO-COL14A1-3′UTR reporter. The sequence (position 1450-1749) of wild-type COL14A1-3′UTR was synthesized as follows: TTTCTGTCCTTCTCATAATCTCCTGATCTTTGGAGATGAAAGATCTCCTTTCATCTTCATCTGAAGGATTCTTCAGATGATTCATCTTCAGACTTCATCTGAAGGATTCAAACAATCTACTTGTGTGCCTTTGGTGGGGTGGGGCAGGGGCGGGGGGCGGTTCTAAACAAATCAGTTTTTTTGGTGTCTTTGCAGGGAATGAAAGAAGATAATGAGCAGATAATCAGCAGCCATTTCCTTTCAGTTATGGTATATGATTCTGCCTGAGAAAGCAGGATTTATGGGAAGGAAAACTTGTCT. Mutated COL14A1-3′UTR was cloned with substitution of miR-7108-3p target site (position 1597-1603). The reporter and pRL-TK vector were subsequently co-transfected into HEK293T cells with miR-7108-3p mimic or inhibitor using Lipofectamine 3000 (Invitrogen). At 48 hours post-transfection, cells were collected for luciferase activity detection using a Dual Luciferase Reporter Gene Assay Kit (KeyGen Biotech, Nanjing, China). Mimic and inhibitor of miR-7108-3p or negative control (Table 1) were synthesized by GenePharma (Shanghai, China).

Statistical analysis

Statistical analysis was performed by GraphPad Prism 7.0 (GraphPad Software Inc., La Jolla, CA, USA). Data are presented as mean±standard deviation. Differences between groups were evaluated using two-tailed Student's t-tests. P < 0.05 was considered statistically significant.


COL14A1 was overexpressed in liver CSCs

For cancer patients, the biggest fear is that malignant tumor cells continue to metastasize and spread. The reason metastasis occurs is rooted in the tumor seed, that is, CSCs left in the body following removal of primary lesions.[12] In view of the vital role of CSCs in various cancers, we have performed a lot of work in this area[5,13,14] and identified many aberrantly expressed molecules in CSCs. COL14A1, a molecule overexpressed in a variety of cancers (Fig. 1A), especially HCC tumors (Fig. 1B), was notable among them. Through suspension culture,[14] we confirmed that COL14A1 was significantly upregulated in CSCs of both HCC primary cells (Fig. 1C, top panel) and HCC cell lines (Fig. 1C, bottom panel).

Figure 1
Figure 1:
COL14A1 was overexpressed in liver CSCs. (A) COL14A1 expression in cancers was analyzed by the Oncomine Platform. Threshold (P-value) = 0.05; Threshold (Fold change) = 2; Threshold (Gene rank) = Top 10%. Cell color is determined by the best gene rank percentile for analyses within cells. (B) COL14A1 expression in HCC was analyzed by the Oncomine Platform. Fold change = 2.019; Mas Liver, a data set of HCC from Oncomine Platform. (C) COL14A1 expression in CSCs was compared with cells in HCC primary cells (top panel) and HCC cell lines (bottom panel) by qRT-PCR. Cells were HCC cells in adherent culture, while CSCs were matched CSCs obtained by suspension culture. β-actin served as the control gene. The blue columns represent data from HCC cells cultured adherently. Data are presented as mean±SD, and derived from three independent experiments. P < 0.05, ∗∗ P < 0.01, vs HCC cells in adherent culture by two-tailed Student's t-tests. COL14A1 = Collagen Type XIV Alpha 1 Chain, CSC = cancer stem cell, HCC = hepatocellular carcinoma, qRT-PCR = quantitative real-time polymerase chain reaction.

COL14A1 was required for liver CSCs to maintain their self-renewal capability

To explore the function of COL14A1 in liver CSCs, we constructed two shRNAs to silence COL14A1. Both shRNAs achieved good knockdown efficiency in HCC cell lines Hep3B and Huh7 (Fig. 2A). Notably, both mRNA and protein expression of the pluripotent transcription factor NANOG was significantly attenuated with COL14A1 depletion (Fig. 2B). Additionally, COL14A1 knockdown remarkably increased cell apoptosis (Fig. 2C) and impaired cell proliferation (Fig. 2D) of Hep3B and Huh7 cells. More importantly, tumor sphere formation, the most direct and powerful assay to evaluate the self-renewal capability of CSCs, was greatly inhibited by knockdown of COL14A1 expression (Fig. 2E). Collectively, these results clearly indicate that COL14A1 was required for the maintenance self-renewal in liver CSCs.

Figure 2
Figure 2:
COL14A1 was required for self-renewal of liver CSCs. (A) Two shRNAs, shCOL14A1#1 and shCOL14A1#2, significantly knocked down expression of both COL14A1 mRNA (top panel) and protein (bottom panel) in Hep3B and Huh7 cells. (B) NANOG expression was detected using qRT-PCR (top panel) and western blotting (bottom panel) in COL14A1-knockdown Hep3B and Huh7 cells. (C) Ratios of apoptotic COL14A1-silenced Hep3B and Huh7 cells were detected using Annexin V-FITC/PI double staining. (D) Cell proliferation potency of COL14A1-silenced Hep3B and Huh7 cells was detected using EdU staining by a fluorescence microscope. Green staining indicated EdU-labeled (EdU+) cells; blue staining indicated DAPI-labeled cells. (E) A tumorsphere formation assay was performed to detect the pluripotency of Hep3B and Huh7 cells when COL14A1 was knocked down. Images were captured using a light microscope. β-actin served as the control gene for mRNA. Data are presented as mean±SD, and derived from three independent experiments. COL14A1-silenced (shCOL14A1#1 and shCOL14A1#2) Hep3B and Huh7 cells were compared with the matched control (scramble) ( P < 0.05, ∗∗ P < 0.01, vs control (scramble); by two-tailed Student's t-test). COL14A1 = Collagen Type XIV Alpha 1 Chain, CSC = cancer stem cell; qRT-PCR = quantitative real-time polymerase chain reaction; scramble = scrambled control shRNA; snRNA = small nuclear RNA, GAPDH = glyceraldehyde 3-phosphate dehydrogenase, EdU = 5-Ethynyl-2’-deoxyuridine, DAPI = 4’,6-diamidino-2-phenylindole.

miR-7108-3p regulated COL14A1 expression in liver CSCs

To elucidate the mechanism of COL14A1 overexpression in liver CSCs, we analyzed miRNAs targeting COL14A1 mRNA. Twelve miRNAs experimentally validated to have interactions with COL14A1 mRNA have been reported in the miRTarBase database. We tested expression of these twelve miRNAs in liver CSCs, and found that miR-7108-3p was significantly downregulated in CSCs of both HCC primary cells (Fig. 3A, top panel) and the Huh7 cell line (Fig. 3A, bottom panel). Next, we performed miRNA overexpression and inhibition experiments to verify that regulation occurs. COL14A1 expression was effectively inhibited by miR-7108-3p mimic (Fig. 3B) and remarkably enhanced by its inhibitor (Fig. 3C). To further confirm the interaction, we constructed two luciferase reporter plasmids containing the 3′UTR of COL14A1 mRNA- and either the wild-type (pmirGLO-COL14A1-3′UTR-WT) or mutated (pmirGLO-COL14A1-3′UTR-Mut) putative miR-7108-3p binding site, respectively. As a result, a significant change was only observed in the luciferase activity of pmirGLO-COL14A1-3′UTR-WT, not the mutated pmirGLO-COL14A1-3′UTR-Mut (Fig. 3D). Collectively, these results suggest that lower expression of miR-7108-3p allowed more COL14A1 to be expressed in liver CSCs.

Figure 3
Figure 3:
miR-7108-3p regulated COL14A1 expression in liver CSCs. (A) miR-7108-3p, but not the other eleven miRNAs examined, was remarkably downregulated in CSCs of both HCC primary cells and Huh7 cell line compared with the matched cells. Cells were HCC cells in adherent culture, while CSCs were matched CSCs obtained by suspension culture. (B) miR-7108-3p mimic remarkably decreased COL14A1 RNA and protein expression when compared with control mimic ( P < 0.05, ∗∗ P < 0.01). (C) miR-7108-3p inhibitor remarkably increased COL14A1 RNA and protein expression when compared with control inhibitor ( P < 0.05, ∗∗ P < 0.01). (D) Luciferase activities of pmirGLO-COL14A1-3′UTR-WT and pmirGLO-COL14A1-3′UTR-Mut were analyzed when co-transfected with miR-7108-3p mimic (middle panel) or miR-7108-3p inhibitor (right panel), compared with the matched control mimic and control inhibitor, respectively. U6 snRNA and β-actin served as the control gene for miRNA and mRNA, respectively. Data are presented as mean±SD, and derived from three independent experiments. P < 0.05, ∗∗ P < 0.01 (two-tailed Student's t-test). COL14A1 = Collagen Type XIV Alpha 1 Chain, CSC = cancer stem cell, GAPDH = glyceraldehyde 3-phosphate dehydrogenase, HCC = hepatocellular carcinoma, snRNA = small nuclear RNA.

COL14A1 activated ERK signaling in liver CSCs

Analysis using GeneCards ( revealed five potential superpathways for the COL14A1 gene. Notably, ERK signaling, a critical pathway in normal cells and for stemness of certain cancer cells,[15] was among them. We found that ERK signaling was extremely activated in liver CSCs (Fig. 4A). To verify the interaction between COL14A1 and ERK signaling, we examined protein levels of total ERK1/2 (ERK1/2) and phosphorylated extracellular signal-regulated kinase (p-ERK1/2). We observed a significant change in the percentage of p-ERK1/2 to total ERK1/2 in COL14A1-silenced Hep3B and Huh7 cells compared with control cells (Fig. 4B). This finding strongly indicates that COL14A1 maintained the self-renewal of CSCs in HCC through ERK signaling.

Figure 4
Figure 4:
COL14A1 activated ERK signaling in liver CSCs (western blot assay). (A) Top panel, protein bands of p-ERK1/2 and ERK1/2 were visualized in CSCs and cells of Hep3B and Huh7 cell lines by enhanced chemiluminescence. Bottom panel, percentage of p-ERK1/2 to ERK1/2 was calculated used ImageJ, and gray value represented protein amount. Cells were HCC cells in adherent culture, while CSCs were matched CSCs obtained by suspension culture. (B) Top panel, protein bands of p-ERK1/2 and ERK1/2 in COL14A1-silenced (shCOL14A1#1 and shCOL14A1#2) and matched control (scramble) Hep3B and Huh7 cells were visualized by enhanced chemiluminescence. Bottom panel, percentage of p-ERK1/2 to ERK1/2 was calculated used ImageJ, and gray value represented protein amount. COL14A1 = Collagen Type XIV Alpha 1 Chain, CSC = cancer stem cell, ERK1/2 = extracellular signal-regulated kinase 1 and 2, GAPDH = glyceraldehyde 3-phosphate dehydrogenase, p-ERK = phosphorylated extracellular signal-regulated kinase 1 and 2.


In many solid tumors, only the population of CSCs (a minority among all cancer cells) could maintain a tumor, while the majorities were terminally differentiated and lost clonogenic potential.[16,17] Nevertheless, research directly confirming the existence of CSCs in human cancer was recently reported by Melenhorst et al,[18] who identified a shocking clinical case of leukemic relapse after chimeric antigen receptor (CAR) T-cell immunotherapy. Strikingly, the cancer cells that caused the patient's death all came from a single edited leukemic B cell, which was unintentionally introduced by the CAR gene. Therefore, CSCs are critical for tumor metastasis and recurrence.

CSCs, like normal pluripotent stem cells, have the ability to self-renew and differentiate into various types of cells.[19] CSCs maintain their stemness through multiple mechanisms, including activation of stem-cell promoting signaling[20] and inactivation of pathways inhibiting stem-cell characteristics.[21] Here, we showed that COL14A1 (negatively regulated by miR-7108-3p) could promote the self-renewal of liver CSCs through the ERK signaling pathway, which has been found to regulate NANOG expression and enhance CSC-like phenotypes (Fig. 5).[22] Based on this, therapy targeting COL14A1-ERK signaling could contribute to the elimination of liver CSCs, thus inhibiting tumor recurrence to improve the survival rate of HCC.

Figure 5
Figure 5:
Schematic diagram of the mechanism of COL14A1 promoting self-renewal of liver CSCs. MiR-7108-3p is downregulated in liver CSCs, and inhibition of miR-7108-3p releases more COL14A1 mRNA to be translated into protein, which subsequently increases levels of p-ERK1/2 to activate ERK signaling, eventually elevating NANOG expression and promoting the self-renewal of liver CSCs. COL14A1 = Collagen Type XIV Alpha 1 Chain, CSC = cancer stem cell, ERK = extracellular signal-regulated kinase, p-ERK1/2 = phosphorylated extracellular signal-regulated kinase 1 and 2.

CSC characteristics are regulated by transcription factors,[23] epigenetic regulators,[24] and extrinsic and intrinsic cell signaling factors.[25] It has been reported that collagens are aberrantly expressed in the tumoral extracellular matrix,[26] whereby they actively drive tumor growth and invasiveness.[27] In gastric cancer, as a component of extracellular matrix, COL14A1 exhibited high prevalence of nonsynonymous mutations.[28] Furthermore, an association between altered expression of COL14A1 and hepatic fibrosis has been confirmed.[29] However, the mode by which this extracellular collagen participated in HCC development and self-renewal of liver CSCs remained unknown. Here, we showed that COL14A1, which is upregulated in liver CSCs, is required to maintain their self-renewal property through activation of ERK signaling by increasing p-ERK1/2. Our results suggest that extracellular collagens represent an important layer for regulation of oncogenesis. However, how COL14A1 increases levels of p-ERK1/2 in liver CSCs remains to be further investigated to uncover this regulation.

However, this study had some limitations. Our observations of the biological function of COL14A1 were based on ex vivo cell culture, which may not be able to express the complexity of in vivo situation. Therefore, the potential of COL14A1 as a therapeutic target for HCC treatment needs to be well evaluated in animal models and clinical trials.

In conclusion, our results show that COL14A1 plays a crucial regulatory role in liver CSCs. Interfering with COL14A1 expression had a striking effect on self-renewal of liver CSCs. Further research demonstrated that COL14A1 enhances phosphorylation of ERK protein to activate ERK signaling, which elevates NANOG expression and promotes CSC self-renewal. Our research provides a new target and theoretical basis for liver CSC elimination, as well as curbing HCC metastasis and recurrence.



Author contributions

SL supervised the study. SL and RK designed the study and wrote the manuscript. RK, HL, YS and QM performed the experiments and analyzed the data. All authors edited the manuscript and approved the final version of the manuscript.

Financial support

This work was supported by the Shanghai Science and Technology Committee of China (No. 18XD1405300), the State Key Program of National Natural Science Foundation of China (No. 81730076), and the Program of Shanghai Fourth People's Hospital of China (No. SY-XKZT-2020-1009).

Institutional review board statement

This study was approved by the Ethics Committee of Shanghai Fourth People's Hospital, Tongji University School of Medicine, China (approval No. 2019tjdx17) on February 21, 2019.

Conflicts of interest

The authors declare that they have no conflicts of interest.


[1]. Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394–424.
[2]. Dutta R, Mahato RI. Recent advances in hepatocellular carcinoma therapy. Pharmacol Ther 2017;173:106–117.
[3]. Zheng H, Pomyen Y, Hernandez MO, et al. Single-cell analysis reveals cancer stem cell heterogeneity in hepatocellular carcinoma. Hepatology 2018;68:127–140.
[4]. Clarke MF. Clinical and therapeutic implications of cancer stem cells. N Engl J Med 2019;380:2237–2245.
[5]. Chen L, Kong R, Wu C, et al. Circ-MALAT1 functions as both an mRNA translation brake and a microRNA sponge to promote self-renewal of hepatocellular cancer stem cells. Adv Sci (Weinh) 2020;7:1900949.
[6]. Wang Y, He L, Du Y, et al. The long noncoding RNA lncTCF7 promotes self-renewal of human liver cancer stem cells through activation of Wnt signaling. Cell Stem Cell 2015;16:413–425.
[7]. Castagnola P, Tavella S, Gerecke DR, et al. Tissue-specific expression of type XIV collagen--a member of the FACIT class of collagens. Eur J Cell Biol 1992;59:340–347.
[8]. Pickup MW, Mouw JK, Weaver VM. The extracellular matrix modulates the hallmarks of cancer. EMBO Rep 2014;15:1243–1253.
[9]. Afik R, Zigmond E, Vugman M, et al. Tumor macrophages are pivotal constructors of tumor collagenous matrix. J Exp Med 2016;213:2315–2331.
[10]. Lai KK, Shang S, Lohia N, et al. Extracellular matrix dynamics in hepatocarcinogenesis: a comparative proteomics study of PDGFC transgenic and Pten null mouse models. PLoS Genet 2011;7:e1002147.
[11]. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 1977;74:5463–5467.
[12]. Parada LF, Dirks PB, Wechsler-Reya RJ. Brain tumor stem cells remain in play. J Clin Oncol 2017;35:2428–2431.
[13]. Liu S, Cheng K, Zhang H, et al. Methylation status of the nanog promoter determines the switch between cancer cells and cancer stem cells. Adv Sci (Weinh) 2020;7:1903035.
[14]. Liu S, Li N, Yu X, et al. Expression of intercellular adhesion molecule 1 by hepatocellular carcinoma stem cells and circulating tumor cells. Gastroenterology 2013;144:1031–1041. e10.
[15]. Du R, Shen W, Liu Y, et al. TGIF2 promotes the progression of lung adenocarcinoma by bridging EGFR/RAS/ERK signaling to cancer cell stemness. Signal Transduct Target Ther 2019;4:60.
[16]. Saygin C, Matei D, Majeti R, et al. Targeting cancer stemness in the clinic: from hype to hope. Cell Stem Cell 2019;24:25–40.
[17]. Al-Hajj M, Wicha MS, Benito-Hernandez A, et al. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 2003;100:3983–3988.
[18]. Ruella M, Xu J, Barrett DM, et al. Induction of resistance to chimeric antigen receptor T cell therapy by transduction of a single leukemic B cell. Nat Med 2018;24:1499–1503.
[19]. Batlle E, Clevers H. Cancer stem cells revisited. Nat Med 2017;23:1124–1134.
[20]. Bahr C, von Paleske L, Uslu VV, et al. A Myc enhancer cluster regulates normal and leukaemic haematopoietic stem cell hierarchies. Nature 2018;553:515–520.
[21]. Shlush LI, Mitchell A, Heisler L, et al. Tracing the origins of relapse in acute myeloid leukaemia to stem cells. Nature 2017;547:104–108.
[22]. Huang C, Yoon C, Zhou XH, et al. ERK1/2-Nanog signaling pathway enhances CD44(+) cancer stem-like cell phenotypes and epithelial-to-mesenchymal transition in head and neck squamous cell carcinomas. Cell Death Dis 2020;11:266.
[23]. Park CS, Lewis AH, Chen TJ, et al. A KLF4-DYRK2-mediated pathway regulating self-renewal in CML stem cells. Blood 2019;134:1960–1972.
[24]. Paris J, Morgan M, Campos J, et al. Targeting the RNA m(6)A reader YTHDF2 selectively compromises cancer stem cells in acute myeloid leukemia. Cell Stem Cell 2019;25:137–148. e6.
[25]. Clara JA, Monge C, Yang Y, et al. Targeting signalling pathways and the immune microenvironment of cancer stem cells - a clinical update. Nat Rev Clin Oncol 2020;17:204–232.
[26]. Hessmann E, Buchholz SM, Demir IE, et al. Microenvironmental determinants of pancreatic cancer. Physiol Rev 2020;100:1707–1751.
[27]. Liguori M, Solinas G, Germano G, et al. Tumor-associated macrophages as incessant builders and destroyers of the cancer stroma. Cancers (Basel) 2011;3:3740–3761.
[28]. Li X, Wu WK, Xing R, et al. Distinct subtypes of gastric cancer defined by molecular characterization include novel mutational signatures with prognostic capability. Cancer Res 2016;76:1724–1732.
[29]. Bracht T, Schweinsberg V, Trippler M, et al. Analysis of disease-associated protein expression using quantitative proteomics--fibulin-5 is expressed in association with hepatic fibrosis. J Proteome Res 2015;14:2278–2286.

COL14A1; ERK signaling; hepatocellular carcinoma; liver cancer stem cells; miR-7108-3p

Copyright © 2021 The Chinese Medical Association, Published by Wolters Kluwer Health, Inc. under the CCBY-NC-ND license.