This article is mentioned in:

Introduction

Esophageal cancer is a common digestive tract cancer worldwide and ranks seventh in morbidity and sixth in mortality rates among all cancer types. Approximately 604,000 new cases were recorded in 2020 worldwide (1). According to differences in histopathology, esophageal cancer can be divided into two subtypes: Esophageal adenocarcinoma (EAC) and esophageal squamous cell carcinoma (ESCC). ESCC is the most common type of esophageal cancer worldwide, which is primarily distributed in the Asian esophageal cancer belt, including Japan, northern China, Iran, and parts of Africa (2). With a 5-year survival rate ranging from 15 to 25% in China, there is an urgent need to improve patient survival by improving early diagnosis and treatment (3). Therefore, it is vital to study the pathogenesis of ESCC, to identify prognostic markers and effective drugs for patients.

To date, several genome-wide profiles of ESCC have been performed by whole genome, exome, and targeted sequencing. Compared with EAC, the copy number variations (CNVs) landscape of ESCC is more akin to that of other types of squamous carcinomas, such as head and neck squamous cell carcinoma (HNSCC) and cervical squamous cell carcinoma (CESC) (4–6). In addition to well-established cancer genes (EGFR, CDKN2A/2B, CCND1), researchers have also revealed that other genes (CBX4, ZNF750, CDCA7, FAM84B) may play important roles in ESCC (7–10). Moreover, generous focal CNV segments (amplifications and deletions) have been identified. Certain gene signatures in these segments can lead to metastasis or poor prognosis (11). Integrated gene enrichment analysis of somatic mutations and CNVs has found pivotal cancer pathways involved in ESCC progression, including the receptor tyrosine kinase (RTK)-Ras-PI3K, JAK-STAT, Akt, NOTCH, Wnt, cell cycle, and p53 signaling pathways (6,7,12). All these findings highlight potential therapeutic molecular targets for the management of ESCC (13).

Although certain reports have shed light on specific genes related to the molecular carcinogenic mechanism or targeted treatment of ESCC (6,12), a systematic review concerning CNVs has not been conducted yet to the best of our knowledge. The present review aimed to summarize the implications of CNVs on the formation and progression of ESCC; to describe the identified genes that are amplified or deleted in ESCC; and to discuss their roles as potential biomarkers. A summary of the present review is presented in Table I. The current review may help to understand the genomic characteristics and variations in ESCC, thus providing a guideline for targeted therapy.

Table I. Highlights of the article.

Table I.

Highlights of the article.

Highlights	(Refs.)
The complex formation mechanisms of CNVs including NAHR, NHEJ, FoSTeS, retrotransposon, amongst others.	(17–22)
The diversity of detection methods for CNVs including WGS, WES, TS, chromosome microarrays, amongst others.	(30–35)
The universality of CNVs affected genes including NFE2L2, ZNF750, amongst others.	(41–99)
The potential value of clinical application of CNV events including cetuximab and nimotuzumab in EGFR,	(100–110)

[i] Apatinib and Endor in VEGFR, amongst others.

Mechanisms of formation of CNVs and their effects on gene expression

CNVs are widely distributed across the human genome and refer to the copy number change by deletion, insertion, inversion, translocation, duplication, and other types of chromosomal rearrangements (14). CNVs are generally DNA segments of >1 kb in size, with a mean fragment length of ~2.9 kb (15,16). CNVs are mainly generated by non-allele homologous recombination (NAHR), non-homologous end joining (NHEJ), a DNA replication-based mechanism such as fork stalling and template switching (FoSTeS), and long interspersed nuclear element-1 (L1) retrotransposition (17). NAHR is the principal mechanism for the formation of CNVs (18). DNA is damaged and repaired via a DNA double-stranded break repair mechanism to produce NHEJ following exposure to ionizing radiation or oxidative stress (19,20). In 2007, Lee et al (21) proposed a FoSTeS model to explain the complex CNV rearrangements, which may explain the process of producing more complex and longer CNVs. Retrotransposon is primarily produced by spontaneous transposon L1s (22).

CNVs play a prominent role in tumorigenesis in various cancer types, including breast cancer, ovarian cancer, hepatocellular carcinoma, colorectal cancer, and bladder carcinoma (23–27). Amplified genomic regions are often detected to harbor oncogenes, while tumor suppressor genes are often located in the deleted regions. The molecular mechanisms of pathogenesis primarily include gene dosage, gene disruption and fusion, position effect, and recessive allele mutations (18). The gene dosage effect is characterized by changes in gene expression levels along with the CNVs. Gene breakups and breakpoint fusions are caused by interruptions directly by CNVs, resulting in loss of function or gain of function mutations. The position effect leading to rearrangement regulates distal and proximal gene expression, and affects the translation of adjacent genes (28,29).

Detection methods for CNVs

In recent years, with the rapid development of sequencing technologies, the methods of detecting CNVs have made notable progress. Common detection methods for CNVs can be primarily divided into two categories: i) Genome-wide detection of unknown CNVs and ii) specific detection of known CNVs. Microarray and sequencing are commonly used to detect genome-unknown CNVs.

Currently, the mainstream approaches for analyzing CNVs are gene sequencing-based methods. Sequencing techniques comprise Sanger DNA sequencing, next-generation sequencing (NGS), and third-generation sequencing (TGS), of which NGS and TGS are commonly used. Whole genome sequencing (WGS), whole exome sequencing (WES), and targeted sequencing (TS) are frequently utilized for CNV detection by NGS or TGS. Each kind of sequencing may be applied to different aspects. WGS is a high-resolution and reliable approach to detect CNVs with greater coverage and increased accuracy (30,31). Compared to WGS, WES needs less time and is cheaper, but its application is limited. Since exomes account for 1–1.5% of the human genome, WES is a better choice when only detecting coding region copy number changes. TS focuses on select DNA segments that can be used to identify somatic changes (32). In recent decades, NGS has revealed aberrant genes and signaling pathways, and has certain advantages regarding throughput capacity and price. TGS can be used to screen out single molecular genes even without PCR sequencing, and can generate longer and more accurate reads with faster speed (33,34). It can read >15 kb on average, which provides the possibility of obtaining longer sequences and avoids the need for fragment rejoining (35). By TGS, CNV fragment length, and specific sites can be detected more precisely. However, it has some drawbacks such as low throughput, and high cost and error rate.

Chromosome microarrays primarily consist of comparative genomic hybridization and single nucleotide polymorphisms. Candidate-gene screening is generally utilized to identify the CNVs of specific sites. High-throughput gene sequencing, chromosome microarray, and candidate-gene screening can be used for further detection (Fig. 1).

Figure 1. Diagram of the mechanisms, forms, effects, and detection methods of copy number variations.

Affected protein-coding and non-coding CNV genes in ESCC

CNVs are involved in the initiation and development of ESCC (5). CNVs at the chromosomal level were present in 97.9% of patients with ESCC (4). Previous studies, some of them conducted by our group, have identified frequent CNV fragments, including gains at 3q, 5p, 7p, 7q, 8p, 8q, 12p, 14q, 16p, 17p, 18p, 20p, 20q, Xp, and Xq, and losses at 3p, 4p, 4q, 5q, 9p, 10p, 11p, 13q, 18q, 19p, 19q, and 21q (4–8). Certain genes have been identified with copy number changes, including CCND1, CTTN, MYEOV, EGFR, FGFR1, ZNF750, SOX2, and FBXW7, in ESCC (7,10,36). There are several studies focused on the screening of non-coding RNA, including microRNA (miR)-21, miR-23a, miR-342-3p, miR-483-5p, miR-548d-5p, miR-574-5p, miR-595, miR-766, miR-940, miR-1224-3p, miR-1260, miR-1280, miR-4448, and miR-4707-5p, which are overexpressed with amplifications, while let-7c, miR-139-3p and miR-139-5p are downregulated with losses (37,38). Due to genomic and epigenomic aberrations, several signaling pathways such as disrupted G1-S cell cycle and amplified PI3K/Akt/mTOR could be therapeutic targets for further treatment (36,39,40). Our previous study constructed an evolutionary tree of ESCC and found three models of evolution: i) Metastasis-to-metastasis; ii) linear spread; and iii) explosive metastasis models. The majority of amplifications or deletions of driving genes that play critical roles in tumor pathogenesis are trunk variations. Furthermore, cell cycle-regulated genes occur in the early stages of tumor progression in particular (12). Our group found that some local deletions were closely associated with a poor prognosis, including 13q12.11, 13q14.2, 17q25.3, and 22q13.33 (10). As mentioned above, CNVs can be clustered according to different locations at a chromosomal arm level. The following sections summarize important CNVs identified in ESCC (Fig. 2 and Table II).

Figure 2. Overview of copy number variation-affected genes during the progression of esophageal squamous cell carcinoma. These genes have a close association with the characteristics of cancer, such as proliferation, anti-apoptosis, mobility, angiogenesis, and heterogeneity. Genes colored red indicate amplification or gain, while those colored green indicate deletion or loss. M, biomarker for diagnosis; T, therapeutic target.

Table II. Copy number variations associated with esophageal squamous cell carcinoma.

Table II.

Copy number variations associated with esophageal squamous cell carcinoma.

Chromosome location	Associated genes	Variation type	Frequency (%)	Function	Mechanisms	Clinical implications	(Refs.)
1p36.32	TP73-AS1	Gain	2.1	Oncogene	Promote proliferation, suppress apoptosis, and increase the chemical sensitivity to cisplatin and 5-FU	An independent prognostic biomarker and a potential therapeutic target	(41)
1q32.1	AGPG	Gain	1.1	Oncogene	Promote glycolysis and proliferation	A promising target for metabolism-related treatment	(42)
2q31.2	NFE2L2	Gain	14.3	Oncogene	Associate with metabolic pathways, Notch signaling pathway and PI3K/Akt signaling pathway	A predictive marker used in chemoradiotherapy	(44–49)
3p21.1	CACNA2D3	Loss	9	Suppressor gene	Inhibit tumorigenesis, mobility, induce apoptosis and enhance the chemical sensitivity to cisplatin	A potential target when combined with cisplatin	(50,51)
3p22.2	DLEC1	Loss	40.2	Suppressor gene	Inhibit cell cycle and metastasis, inhibit JAK/STAT3, MAPK/ERK, Wnt/β-catenin and AKT pathways	DLEC1 methylation as a biomarker in ESCC	(52)
	RBMS3	Loss	61	Suppressor gene	Reduce the tumorigenicity of ESCC tumor cells	A novel therapeutic target	(53)
3p26	CHL1	Loss	41	Suppressor gene	Anti-proliferation, anti-metastasis. Inhibit AKT signaling pathway involved in cell adhesion	Unknown	(54)
3q26	PIK3CA	Gain	29	Oncogene	Involve in PI3K signaling pathway to promote ESCC	A predictive marker	(57,59)
	SOX2	Gain	28	Oncogene	Activate HIF-1α/STAT3 signaling pathway and promote metastasis	Indicative of a poor prognosis	(60–63)
8p11	FGFR1	Gain	9	Oncogene	Activate MEK-ERK signaling pathway	A potential therapeutic target	(13,36,64–67)
8q24	MYC	Gain	43	Oncogene	Involve in TLR7/NF-κB, PI3K/AKT and other signaling pathways	An independent prognostic factor	(11,68–71)
	TSTA3	Gain	13	Oncogene	Promote metastasis by regulating fucosylation of LAMP2 and ERBB2	An independent indicator of poor prognosis	(72,73)
	PCAT1	Gain	18	Oncogene	Sponge miR-326 as a ceRNA	A non-invasive biomarker	(38,114)
9p21.3	MTAP	Loss	28	Suppressor gene	Induce EMT through the GSK3β/Slug/E-cadherin signal axis	Unknown	(74)
	CDKN2A/2B	Loss	63/56	Suppressor gene	Deregulate G1 phase during the cell cycle	A marker of sensitivity to CDK4/6 inhibitors	(8,75–77)
11q13	MALAT1	Gain	22.2	Oncogene	Inhibit the effect of radiotherapy	A potential target in radiotherapy	(115,116)
	FGF4	Gain	54	Oncogene	Unknown	An independent prognostic factor	(80)
	ANO1	Gain	57	Oncogene	Promote proliferation, invasion and migration	A predictive marker	(37,81,82)
	mir-548k	Gain	57	Oncogene	Facilitate lymphatic metastasis and regulate microenvironment through the secretion of VEGFC	A predictive marker in early diagnosis	(6,83)
12q11.21	FAM60A	Gain	9	Oncogene	Regulates cell cycle, promote cell proliferation, and inhibit apoptosis	A therapeutic target	(84)
12q15	MDM2	Gain	5	Oncogene	Promote cell proliferation and inhibit apoptosis	A predictive marker used in radiochemotherapy	(85–88)
13q32.1	ABCC4	Gain	14.8	Oncogene	Promote the proliferation and activate oncogenic pathways	A predictive marker and therapeutic target	(89)
14q21.1	FOXA1	Gain	8.3	Oncogene	Promote invasion and migration	A predictive marker	(90–92)
17q25.3	ZNF750	Loss	14	Suppressor gene	Inhibit tumor proliferation, invasion, migration and angiogenesis	A novel therapeutic target	(93,94)
19p13.3	DIRAS1	Loss	59.3	Suppressor gene	Promote cell apoptosis and inhibit invasion	A predictive marker	(98)
20q13.12	MYBL2	Gain	51.4	Oncogene	Induce ESCC cell proliferation and regulate cell cycle	An independent indicator of poor prognosis	(38,99)

Chromosome 1

1p36.32

Previous studies on 1p36.32 have demonstrated that CNVs in long non-coding RNA (lncRNA) regions play critical roles in cancer progression. Importantly, certain lncRNAs may serve as prognostic biomarkers for clinical diagnosis. For example, lncRNA TP73-AS1 is located in 1p36.32 and is upregulated in ESCC, promoting proliferation and suppressing apoptosis. TP73-AS1 enhances the sensitivity of ESCC to cisplatin and 5-fluorouracil. In clinical samples, TP73-AS1 is associated with TNM staging and tumor localization of ESCC, but not with sex, age, lymphatic metastasis, or differentiation. Increased copy numbers of TP73-AS1 were found in early-stage ESCC, whereas decreased expression of TP73-AS1 was found in advanced TNM stages (41).

1q32.1

Situated on chromosome 1q32.1, AGPG is identified as a metabolic lncRNA. According to Liu et al (42), the copy number of AGPG in ESCC cell lines was increased compared with normal cells. As a transcriptional target for p53, AGPG plays a crucial role in promoting glycolysis and proliferation via PFKFB3, eventually leading to carcinogenesis. Furthermore, overall survival (OS) analysis revealed that advanced AGPG expression was indicative of a poor prognosis for ESCC, showing that it may be a promising target for metabolism-related treatment for ESCC.

Chromosome 2

2q31.2

Large-scale chromosome amplification of the NFE2L2 gene was commonly found in ESCC samples (43). NFE2L2 was not only amplified in ESCC but also mutated in patients (10). NFE2L2 (also known as NRF2) is a transcription factor strongly activated in ESCC. Elevated expression of NRF2 is indicative of a poor outcome, and has been found to be related to metabolism, as well as Notch and PI3K/Akt signaling (44). As an independent prognostic indicator in ESCC, upregulation of the NRF2 gene is important for lymph node metastasis, postoperative recurrence, and OS (45–47). Inhibitors such as AEM1 and ML385 can block signaling pathways and attack tumors (48,49). NRF2 is associated with chemoradiation; increased expression of NRF2 in the nucleus is predictive of a poor clinical response and shorter progression-free survival after chemoradiotherapy in patients with ESCC.

Chromosome 3

3p21-3p26

CACNA2D3, which is located at 3p21.1, is a tumor suppressor gene. Reduced expression of CACNA2D3 is significantly associated with lymph node metastasis, TNM staging, and a poorer prognosis in patients with ESCC. A functional study showed that CACNA2D3 inhibited tumorigenesis and cell migration, and induced apoptosis (50). Moreover, CACNA2D3 enhanced the sensitivity of ESCC to cisplatin by inducing calcium-mediated apoptosis and inhibiting the PI3K/Akt signaling pathway. Thus, regulating CACNA2D3 expression is a potential novel strategy for improving the effectiveness of cisplatin in patients with ESCC (51). DLEC1, which is located at 3p22.2, is frequently downregulated not only in lung and nasopharyngeal carcinoma cell lines, but also in ESCC and other primary tumors. DLEC1 is continuously expressed in normal tissues and immortalized epithelial cells. However, mutations in the DLEC1 gene have rarely been detected. Promoter CpG methylation and histone acetylation leads to low expression of DLEC1. Promoter CpG methylation of DLEC1 is frequently detected in ESCC, and is closely associated with lymph node metastasis, tumor recurrence, and progression. DLEC1 inhibits cell proliferation via cell cycle arrest and metastasis. In addition, it can inhibit cell metastasis by reversing epithelial-mesenchymal transformation (EMT) and stemness via the STAT3 signaling pathway (52). Decreased expression of RBMS3 in 3p24 leads to a poorer prognosis in ESCC. Genetic loss of RBMS3 can effectively reduce the oncogenicity of ESCC tumor cells in vivo and in vitro. It has been demonstrated that RBMS3 dysregulates the expression of MYC and CDK4, thereby inhibiting the phosphorylation of Rb (53). CHL1, which is located at 3p26, is another important tumor suppressor gene. The deletion of CHL1 is associated with poor differentiation, invasion, lymph node metastasis, tumor progression, and a worse OS rate in ESCC. The anti-proliferative and metastatic factor CHL1 can inhibit cell adhesion via the AKT signaling pathway (54).

3q26

The 3q26 segment has been found to be amplified in various tumors such as cervical, ovarian, lung, and head and neck cancer (55). The 3q26 region is notably highly amplified in ESCC, and harbors oncogenes such as PIK3CA, PRKCI, MDS1-EVI1, and SOX2 (56–58). As an oncogene, PIK3CA encodes the catalytic subunit p110-kinase of PI3K, which is highly expressed in esophageal cancer and is involved in the PI3K signaling pathway to promote the development of ESCC (57,59). Gene amplification results in overexpression of PRKCI. Immunohistochemical analysis using 180 ESCC samples showed that overexpression of PRKCI was correlated with lymph node metastasis, while upregulated expression of PRKCI was correlated with poor prognosis in ESCC (56). SRY-box transcription factor 2 (SOX2) has been evaluated in several studies. For instance, Gen et al (60,61) reported the role of SOX2 as a protooncogene in ESCC. The change in copy number of SOX2 was positively correlated with SOX2 expression levels (62). SOX2 promotes ESCC and leads to a poorer prognosis in ESCC. High expression of SOX2 is related to differentiation in ESCC (60). By regulating Slug, SOX2 activates the HIF-1α/STAT3 signaling pathway and promotes metastasis in ESCC (63). These aforementioned studies suggest that 3q26 is a vital region for the tumorigenesis of ESCC.

Chromosome 8

8p11

The FGFR1 gene, which is situated in 8p11, is significantly amplified in ESCC (36). Patients with high FGFR1 expression typically have a short disease-free survival. FGFR1 is an independent predictor of prognosis (64). In other solid tumors, FGFR1 amplification can be used as a hallmark for the prediction of the effectiveness of targeted inhibitors and therefore may similarly act as a promising therapeutic target in ESCC (36). By analyzing >500 cancer cell lines, it was suggested that gains of FGFR1 may be the most effective marker for determining the sensitivity of FGFR1 inhibitors (65). Previous in vitro experiments demonstrated that PD173074 prevented cell viability and blocked the MEK-ERK signaling pathway. Anti-FGFR1 drugs such as PD173074 may be used to inhibit ESCC growth, which may be applied in targeted therapies in the future (64,66). The FGFR1 inhibitor AZD4547 can enhance the sensitivity of ESCC cells to gefitinib (67). The results of multiple clinical trials indicated that the small-molecular inhibitor of FGFR1, dovitinib, exhibited good efficacy in ESCC (13).

8q24

The 8q24 chromosomal region contains the MYC gene. Amplification of the MYC gene is indicative of a poorer prognosis and is an independent prognostic factor. In the majority of clinical cases of ESCC, MYC was closely associated with clinical stage, degree of histological differentiation, lymph node metastasis, and invasive depth of ESCC (11,68). MYC can affect the development of ESCC through multiple signaling pathways such as the TLR7/NF-κB and PI3K/AKT signaling pathways (69–71). Furthermore, GDP-L-fucose synthase (TSTA3) was focally amplified in this region and was significantly associated with poor outcomes of patients with ESCC according to our previous study (72). TSTA3 may act as a driver gene during metastasis by regulating the fucosylation of LAMP2 and ERBB2. Thus, fucosylation inhibitors may be applied to suppress ESCC metastasis (72,73).

Chromosome 9

9p21.3

9p21.3 is the most frequently deleted region in ESCC. In 9p21.3, the MTAP gene is adjacent to CDKN2A/2B, which is a tumor suppressor gene. MTAP induces EMT through the GSK3β/Slug/E-cadherin signaling axis to promote ESCC metastasis (74). The loss of CDKN2A/2B occurs frequently, and is closely associated with lymph node metastasis and poor prognosis in ESCC (8,75). Deficiency of CDKN2A/2B is a marker of sensitivity to CDK4/6 inhibitors, which deregulates the G1 phase of the cell cycle in ESCC (76,77). The 9p region often shows a loss of heterozygosity.

Chromosome 11

11q13.2–11q13.4

The 11q13.2–11q13.4 region is frequently amplified in ESCC and other various tumors, such as HNSCC and breast cancer (78,79). Amplification of 11q13.3 indicates poor outcomes in patients with ESCC (10). The 11q13.2–11q13.4 segment contains ‘FGFs-FADD-SHANK2’. Moreover, the MYEOV, CCND1, CPT1A, CTTN, PPFIA, FADD, TMEM16, and CTTS genes are also located in this region. FGFs include FGF3, FGF4, and FGF19. Upregulated expression of SHANK2 contributes to a poor prognosis of ESCC (38). FGF4 gene amplification is associated with clinical stages and can be used to predict 5-year survival and recurrence rates. Serving as an independent prognostic factor, FGF4 upregulation is indicative of a shorter OS and disease-free survival (80). The ANO1 gene, also known as TMEM16A or DOG1, encoding the Ca2+-activated Cl− channel protein, is an oncogene that can facilitate cell proliferation, invasion, and migration. Upregulated ANO1 expression is associated with a poor prognosis, and is significantly associated with lymphatic metastasis and advanced TNM stages. Immunohistochemical experiments demonstrated that ANO1 was upregulated in tumor specimens from patients with ESCC (37). Previous in vivo and in vitro studies demonstrated that the downregulation of ANO1 via the TGF-β signaling pathway could significantly inhibit the carcinogenesis of ESCC (81). ANO1 may thus be used as a biomarker for ESCC (82).

miR-548k has been identified as an oncogene that is present in the 11q13.3–13.4 amplicon, which induces malignant phenotypes of ESCC (6). In addition, miR-548k notably facilitated lymphatic metastasis and regulated the microenvironment through increased secretion of VEGFC. miR-548k and VEGFC may thus have potential as biomarkers for the early diagnosis of their clinical features in serum detection, and may thus serve as novel targets for the prognostic prediction of ESCC (83).

Chromosome 12

12q11.21

FAM60A is also known as SIN3-HDAC Complex Associated Factor. Gains of copy numbers lead to the upregulation of FAM60A in ESCC. Thus, as a driver gene of ESCC, it may be a clinical target. FAM60A is upregulated in tumor tissues of ESCC, and is closely associated with tumor size, lymph node metastasis, and TNM staging. Knockdown of the FAM60A gene can repress FAM60A expression, arrest cell cycle events at the G2/M phase, reduce the proportion of G1 phase cells, inhibit cell proliferation, and promote apoptosis. FAM60A contributes to metastasis of ESCC, and is predictive of a poor outcome (84).

12q15

MDM2 is located on 12q15, and exhibits gains in 9% of ESCC cases. In a previous study, MDM2 gene amplification notably influenced overall prognosis and led to a shorter survival time (85). MDM2 could promote cell proliferation and apoptosis resistance (86). MDM2 negatively regulates p53 protein activity and plays a vital role in p53-mediated apoptosis, aging, and cell cycle arrest. Nutlin-3, an antagonist of MDM2, stabilizes p53 and increases p53 expression in wild-type p53 cells. Radiotherapy combined with Nutlin-3 can notably inhibit wild-type p53 cells and enhance sensitivity to radiotherapy (87). MDM2 can also be used as a predictor of chemoresistance in ESCC phase III (88).

Chromosome 13

Gains of the ABCC4 gene were found in ESCC. Decreased expression of ABCC4 can result in reduced activity of COX-2, the PGE2 receptor, and MYC, therefore inhibiting the activity and proliferation of ESCC cells. High expression of ABCC4 indicates poor prognosis. ABCC4 activates oncogenic pathways, thereby promoting carcinogenesis in ESCC. ABCC4 may thus have predictive and therapeutic value for the management of ESCC (89).

Chromosome 14

14q21.1

The FOXA1 gene, also known as HNF3A, is located in a frequently amplified region and functions as a DNA-binding transcription factor (90). FOXA1 expression is correlated with differentiation, vascular invasion, lymphatic metastasis, and subtyping (91). Knockdown of the FOXA1 gene markedly reduced invasion and migration in ESCC (91). Accordingly, FOXA1 is predictive of a poor prognosis in patients with ESCC (92).

Chromosome 17

17q25.3

The ZNF750 gene is focally deleted in ESCC (93). Decreased ZNF750 gene expression promotes tumor proliferation, invasion, migration, and angiogenesis (93,94). ZNF750 directly binds to the promoter of SNAI1 to suppress EMT in ESCC (94). Deletions of ZNF750 indicate advanced malignancy and a poor clinical outcome, which highlights its potential as a therapeutic target for the management of ESCC.

Chromosome 19

19p13.3

Deletion of 19p is quite common in lung cancer, ovarian cancer, and ESCC (95–97). The Ras-like small GTPase DIRAS1 at 19p13.3 is a tumor suppressor gene. DNA copy number loss leads to downregulation of DIRAS1, while reduced expression of DIRAS1 contributes to metastasis, and poor outcomes in patients with ESCC. DIRAS1 can promote cell apoptosis via dephosphorylating Bad via the ERK1/2 and p38 MAPK signaling pathways, while inhibiting invasion by regulating MMP2 and MMP9 (98).

Chromosome 20

20q13.12

With tandem amplified somatic CNVs, the MYBL2 gene is upregulated in tumor tissue with copy number gains, which indicates a poorer prognosis (38,99). As a nuclear protein, MYBL2 overexpression can induce cell proliferation, as well as regulate the cell cycle during the S and G2/M phases in ESCC (99). Deletion of MYBL2 reduced the levels of the cell cycle-related proteins CDK1 and cyclin B1 in ESCC cells. Knockdown of MYBL2 also inhibited ESCC cell proliferation and migration (38).

CNVs with clinical application in ESCC

Research on CNVs aims to identify candidate genes and provide clues for the development of effective ESCC clinical treatment. To date, copy number variant genes that have been used in clinical targeted therapy of ESCC include EGFR and VEGFR. Mapped on chromosome 7p11.2, amplified EGFR belongs to the ErbB family of RTKs, and can transfer the downstream signal to the RAS-RAF-MEK-ERK-MAPK and PI3K-AKT-mTOR pathways (100). With an amplification rate of 24.3%, excessive activation of EGFR leads to cell proliferation, invasion, metastasis and poor prognosis in ESCC (6,101). As a well-known valid target for various tumors, EGFR is considered to have potential in ESCC treatment. Thus far, several drugs have been applied to block EGFR signaling such as the monoclonal antibodies cetuximab and nimotuzumab, and the small molecular inhibitor gefitinib. Combination of cetuximab with other adjuvant therapies such as chemotherapeutic drugs and radiotherapy has been shown to exert a remarkable effect on reducing ESCC progression (102). However, the adverse effects need to be taken into consideration. Nimotuzumab exhibits low toxicity, and may thus be a promising treatment strategy for advanced or metastatic ESCC (103). In addition to EGFR, VEGFR also performs well as an effective therapeutic target. VEGFR binds to VEGF to form the VEGF/VEGFR complex and trigger the PI3K/AKT and MAPK/ERK signaling pathways (104). Apatinib and Endor, two anticarcinogens independently developed by Chinese scientists that target the VEGF/VEGFR axis, have been evaluated in clinical trials and may become novel drugs for patients with ESCC (105–107).

Another target, HER-2, has been used for various tumors, particularly breast cancer (108). HER-2 may have potential for the treatment of esophageal cancer as well. In a clinical phase III trial, patients with advanced esophageal adenocarcinoma received combined treatment of the HER-2 specific inhibitor lapatinib and CapeOx; OS reached 12.2 months compared with that of the untreated group (10.5 months), showing a distinct curative effect in esophageal adenocarcinoma (109). At present, clinical evaluation of lapatinib is still required in ESCC. In addition to the utilization of copy number variant genes for targeted therapy, they may also be used for chemoradiotherapy. The MYC locus is highly amplified with a rate of 43% and is expressed throughout the entire process of chemoradiotherapy, which may facilitate an increase in resistance (4). Furthermore, a focal copy-number increase in MYC is markedly associated with a reduced survival rate (110). With an increased understanding of ESCC, additional targets are expected to be identified in the future.

Conclusions and future prospectives

CNVs play crucial roles in ESCC occurrence and development by altering numerous protein-coding genes. Future studies should focus on identifying the emerging functions of non-coding RNAs [including miRNAs, lncRNAs, and circular RNAs (circRNAs)] in copy number aberrations, since protein-coding sequences constitute only 2% of the total genome sequence, while the remaining 98% corresponds to non-coding sequences. However, the function of non-coding RNA has not yet been fully elucidated with regard to CNVs.

The expression of lncRNAs is more specific, particularly in cells, tissues, and during the developmental stages, and in diseased tissues (111,112). Compared with linear RNA, circRNAs are more stable, abundant, and better conserved. Thus, they may serve as a potential target. For instance, lower expression of deleted miRNA let-7c is correlated with a poorer prognosis and reaction to chemotherapy, and it may be used to predict clinical effects. In vitro, let-7c suppresses the IL6/STAT3 signaling pathway after cisplatin exposure. In short, let-7c is a therapeutic target for ESCC and is deserved of further research and exploitation for the management of cancer (37,113). Amplified lncRNA PCAT1 acts as an oncogene and sponges miR-326 as a competing endogenous RNA. PCAT1 was notably detected in exosomes secreted by ESCC cells, highlighting its potential as a non-invasive biomarker for patients with cancer (38,114). Overexpression of the lncRNA MALAT1 may result from frequent amplification of 11q13.1 in ESCC (115). MALAT1 inhibits the effect of radiotherapy partly due to enhanced Cks1 expression. These results demonstrate that MALAT1-targeted therapy may be feasible for improving the effectiveness of radiotherapy (116). Though promising, non-coding RNA-based therapies remain in an early stage of development and need further validation in confirmatory clinical trials. Numerous studies have been conducted to investigate ESCC; however, ESCC still lacks efficient marker genes for diagnosis and subtyping, unlike HER2 in breast cancer, and EGFR and ROS in lung cancer.

Tumors are comprised of not only cancer cells but also various stromal cells, which form the microenvironment the tumor is situated in, and this contributes to cell proliferation, drug resistance, invasion, migration, and the poor outcomes of clinical treatment (117,118). In addition, intra-tumor heterogeneity may lead to differences in the therapeutic efficacy when treating malignant tumors (119). However, the heterogeneity and tumor components of ESCC have not been clearly elucidated yet. Although admixed samples of diverse cell types have been analyzed by second-generation or TGS, heterogeneity was not noticeably detected (120). Single-cell sequencing offers the possibility of exploring genuine cellular and molecular components during tumorigenesis (121). Tumor cells showed higher CNV levels compared with normal epidermal cells in ESCC analyzed at the single-cell level, which exhibited remarkable heterogeneity (122). However, further research is needed to explore the microenvironment of ESCC and to identify more effective targets for tumor therapy.

In conclusion, this review comprehensively summarized the CNV events and CNV-affected genes in ESCC, which contributed to a deeper understanding of the molecular mechanisms underlying the occurrence, development, recurrence, and metastasis of ESCC. Improving our understanding of CNVs in ESCC will be valuable for identifying biomarkers for early diagnosis and molecular typing, as well as prognostic markers, and novel therapeutic targets for the management of ESCC.

Acknowledgements

Not applicable.

Funding

This study was supported by the National Natural Science Foundation of China (grant no. 81972613 and 82072746), the Fundamental Research Program of Shanxi Province (grant no. 202203021222253), the Science and Technology Innovation Team of Shanxi Province (grant no. 202204051001024) and the Science and Technology Achievements Transformation Project of Shanxi Province (grant no. 202204021301062).

Availability of data and materials

Not applicable.

Authors' contributions

JR prepared the manuscript. LZ conceived the subject of review. PK, YW, and DG reviewed and edited the manuscript. Data authentication is not applicable. All authors have read and approved the final manuscript.

Patient consent for publication

Not applicable.

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References