lncRNA LINC02323 predicts adverse neoadjuvant chemotherapy outcomes of gastric cancer patients and regulates cell sensitivity to 5-fluorouracil by negatively modulating miR-139-3p

Abstract

Background/Objective

Drug resistance is a challenging problem in the clinical chemotherapy of gastric cancer. Identification of predictive biomarkers for chemotherapy outcomes could improve therapeutic efficacy and patient prognosis. This study aimed to assess the significance of long non-coding RNA (lncRNA) LINC02323 in gastric cancer progression and neoadjuvant chemotherapy and to explore its potential regulatory mechanism.

Materials and Methods

This study enrolled 117 patients with gastric cancer who received neoadjuvant chemotherapy combined with surgical treatment and 35 patients with benign gastroscopic results. The expression of LINC02323 in gastric mucosal tissues of study subjects was analyzed by PCR, and its association with chemotherapy efficacy and cancer development was evaluated. Gastric cancer cells were treated with 5-FU, and the effect of LINC02323 on cell growth and motility under 5-FU treatments was evaluated using CCK8 and transwell assays.

Results

LINC02323 was upregulated in gastric cancer patients, which was related to advanced T stage, occurrence of lymph node metastasis, and less pathological response to chemotherapy. LINC02323 serves as a prognostic biomarker for predicting poor overall survival of gastric cancer patients receiving neoadjuvant chemotherapy. Silencing LINC02323 suppressed the proliferation and motility of gastric cancer cells treated with 5-FU and induced cell apoptosis, indicating the enhanced sensitivity of gastric cancer cells to 5-FU. miR-139-3p was negatively regulated by LINC02323 and could reverse the function of LINC02323 in 5-FU-treated gastric cancer cells.

Conclusion

Upregulated LINC02323 expression in gastric cancer is associated with malignant progression, adverse prognosis, and chemotherapy resistance. Silencing LINC02323 could enhance the sensitivity of gastric cancer cells to 5-FU by negatively modulating miR-139-3p expression.

Introduction

Among various malignant tumors, gastric cancer has a high incidence and mortality rate. It is characterized by insidious onset, rapid development, and easy postoperative recurrence [1]. In the past decades, the therapy and nursing of gastric cancer have greatly improved; however, the lack of specific therapeutic targets and effective target drugs has led to unsatisfactory prognosis. The occurrence and development of gastric cancer are complex processes and are affected by a variety of factors, such as the tumor microenvironment, Helicobacter pylori, and interactions between host genes [2,3]. Gastroscopy is an important means of gastric cancer screening, but most patients are diagnosed at an advanced stage or even with distant metastasis, and therefore lose the opportunity for surgery. Although chemotherapy and radiotherapy can slow the progression of gastric cancer to some extent, drug and radiation resistance remains a challenging problem in the clinic. Neoadjuvant chemotherapy has been proposed for the treatment of patients with gastric cancer at progressive stages, which could improve their prognosis [4,5]. However, chemotherapy sensitivity remains the main factor that limits therapeutic efficiency. Therefore, the exploration of novel molecular targets for gastric cancer therapy is of great significance for revealing the resistance mechanism and providing optional early intervention strategies for the clinical management of gastric cancer.

Transcriptomic studies have shown that non-coding RNAs (ncRNAs) account for the majority of the human genome. Long noncoding RNAs (lncRNAs) are critical tumor regulators in various human cancers. Identification of gastric cancer development-related lncRNAs could provide potential therapeutic targets. With the development of molecular biology and bioinformatics technologies, numerous studies have established expression profiles to identify candidate biomarkers for gastric cancer [6,7]. LINC02323, located on chromosome 14 between 545253 and 55826, was enriched in the GSE224056, GSE163416, and GSE172032 datasets, which enriched differentially expressed ncRNAs in gastric cancer. LINC02323 was previously suggested to regulate epithelial-mesenchymal transition, boost cell biological function, and predict adverse outcomes in human cancers [8–10]. However, whether the abnormal expression of LINC02323 in gastric cancer indicates its functional role in tumor progression, especially in chemotherapy resistance, remains unknown.

This study investigated the expression of LINC02323 in gastric cancer and evaluated its significance in cancer development and chemotherapy resistance, with the aim of exploring a potential target for gastric cancer therapy. Based on the recognized competing endogenous RNA (ceRNA) theory, the downstream target of LINC02323 was studied to reveal the underlying regulatory mechanism. The study design was summarized as Figure S1.

Materials and methods

Study subjects

This study enrolled 117 patients diagnosed with advanced gastric cancer (T2 and above, with any N stage) through gastroscopic biopsy between 2016 and 2020. according to the following criteria: 1) patients receiving neoadjuvant chemotherapy combined with radical surgery, 2) patients with complete clinical records, and 3) patients who had never received any chemotherapy or radiotherapy before confirmed diagnosis. Patients with one of the following criteria were excluded: 1) patients diagnosed with distant metastasis or early-stage disease (T1 with any N stage), 2) patients who lost the opportunity for surgical treatment, and 3) patients who received radical surgical treatment without neoadjuvant chemotherapy. All patients received surgical therapy combined with neoadjuvant chemotherapy. As the oxaliplatin combined with tiggio (SOX) regimen has been gradually revognized as a more suitable chemotherapy regimen for Asian population, therefore, the neoadjuvant chemotherapy regimens for all included patients were SOX regimens. The medication strategy was: intravenous 130 mg/m² oxaliplatin daily and oral 40 mg/m² tiggio twice daily, and the medication was performed once every three weeks. The response of patients to chemotherapy was evaluated using CAP grading (0–3, a higher score indicates resistance). Patients were followed-up for 3–60 months after surgery through telephone or outpatient review.

Another 35 individuals who underwent gastroscopic biopsy and were diagnosed with inflammation were included as the control group. All participants provided signed informed consent, and the study was approved by the Ethics Committee of Xuzhou Municipal Hospital Affiliated to XuZhou Medical University.

Baseline information and sample collection

The baseline information of the two groups, including age, sex, smoking, drinking history, and body mass index (BMI), were collected to ensure that the information was matched. The clinicopathological features of gastric cancer patients, including differentiation, T stage, lymph node metastasis, and CAP grades, were completely collected.

Gastric mucosal tissues were collected during gastroscopic biopsy examination from the two groups and immediately frozen in liquid nitrogen. For subsequent analyses, the tissues were stored in a −80 °C refrigerator for a long time.

Cell culture

HGC-27 (undifferentiated gastric cancer cells), MKN-45 (poorly differentiated gastric cancer cells), MKN-74 (gastric adenocarcinoma cells), HSC-43 (gastric adenocarcinoma cells), and GES-1 (normal gastric mucosal epithelial cells) cells were obtained from the Institute of Basic Medical Cell Resource Center of the Chinese Academy of Medical Sciences. The cells were maintained in RPMI-1640 culture medium supplemented with 10% FBS. The cell culture was conducted at 37 °C, and the culture medium was replaced every two days. Cells were available for the following experiments after reaching 70–80%.

Cell transfection

Cell transfection was performed with LINC02323 siRNA (5′-GCCTCCTACATCTTCAGAT-3′), miR-139-3p inhibitor (5′-ACUCCAACAGGGCCGCGUCUCCA-3′), or their negative controls (si-LINC02323 NC, 5′-TTCTCCGAACGTGTCACGT-3′ or miR-139-3p NC, 5′-UUGUACUACACAAAAGUACUGGUACUUUUGUGUAGUACAAUU-3′) using Lipofectamine 2000 (Invitrogen). Cell transfection was conducted at room temperature, and the cells were available 48 h after transfection. The efficiency was evaluated the expression of LINC02323 and miR-139-3p in transfected cells.

Real-time quantitative PCR

Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and assessed using the OD260/280 value. Reverse transcription was performed using isolated RNA and Primer Script RT Master Mix (TaKaRa) to obtain cDNA. PCR amplification was conducted on a CFX Connect Real-Time System PCR (Bio-Rad, USA) using the TB Green Premix Ex Taq II reagent (TaKaRa, China). The primer sequences were summarized in Table S1. The reaction conditions were: initial denaturation at 95 °C for 10 min followed by 40 cycles at 95 °C for 15 s and an annealing temperature of 60 °C for 60 s. The relative expression levels were calculated using the 2^-ΔΔct method normalized to GAPDH (for LINC02323) and U6 (for miR-139-3p).

Cell viability assay

Cells (5 × 10³ cells/well) were seeded in 96-well plates and pre-incubated with culture medium for 24 h at 37 °C. The culture medium was replaced with 5-fluorouracil (5-FU, 0, 5, 10, 25, 50, 100, and 200 μg/mL, Sigma Aldrich, USA)-containing culture medium and incubated for another 24 h. Then, CCK8 reagent was added and the OD₄₅₀ was measured after 2 h. The CK group was set without 5-FU treatment and the blank group was set without cells. Cell survival rate = [(OD_teatments-OD_blank)/(OD_CK-OD_blank)] × 100

Cell motility assay

Cells were seeded into 24-well Transwell upper chambers supplied with FBS-free culture medium. The bottom chamber was filled with a complete culture medium containing 20 μg/mL 5-FU. The chambers were incubated at 37 °C for 24 h and the cells in the upper chamber were removed. Subsurface cells were fixed, stained, and counted under a microscope. Five random fields were selected for each treatment group.

Dual-luciferase reporter assay

The binding sites of LINC02323 and miR-139-3p were predicted using the lncRNASNP v3 database (http://gong_lab.hzau.edu.cn/lncRNASNP3#!/). The wild-type LINC02323 vector was constructed by cloning the binding sites into the pGL3 plasmid (Promega, USA), while the mutant sites were cloned to establish the mutant-type vector. The vectors were co-transfected with miR-139-3p mimic, inhibitor, or negative controls into gastric cancer cells using Lipofectamine 2000 (Invitrogen, USA), and the relative luciferase activity of LINC02323 was detected using the TransDetect Double-Luciferase Reporter Assay Kit (TransGen Biotech, China) with Renilla as an internal reference.

Statistical analyses

Differences were compared using Student’s t-test (for comparison between the two groups after evaluating to be normal distribution by P-P plot) and one-way analysis of variance (ANOVA, for comparison among multiple groups, p < 0.05) using GraphPad Prism software (version 9.0). The chi-square test was used to assess the association between LINC02323 and the clinicopathological features of gastric cancer patients. Kaplan-Meier and Cox regression analyses were employed for the analysis of follow-up data to evaluate the prognostic significance of LINC02323.

Results

Expression of LINC02323 in gastric cancer patients

A total of 391 differentially expressed ncRNAs were filtered from the GSE224056 dataset, including 153 downregulated and 238 upregulated ncRNAs in gastric cancer (Figure S2a). From the GSE163416 dataset, 681 dysregulated ncRNAs were enriched, including 353 downregulated and 328 upregulated ncRNAs (Figure S2b). A total of 235 downregulated and 156 upregulated ncRNAs were enriched in the GSE172032 dataset (Figure S2c).

Among the enriched dysregulated ncRNAs, 27 differentially expressed ncRNAs were finally enriched at the intersection of the GSE224056, GSE163416, and GSE172032 datasets, where LINC02323 showed significant potential as a biomarker (Figure 1a).

Figure 1.

(a). Venn plots merged enriched ncRNAs and 27 ncRNAs were enriched in the intersection of three datasets. b-c. LINC02323 was upregulated in gastric cancer patients compared with the control group (b), especially in patients with no pathological response to neoadjuvant chemotherapy (c). **p < 0.01, ^****p < 0.0001.

LINC02323 was significantly upregulated in gastric cancer patients compared to patients with benign gastroscopic biopsy results (Figure 1b). According to the CAP grades of patients with gastric cancer, patients were defined as those with pathological response (CAP = 0–2) and without pathological response (CAP = 3). Patients without a pathological response showed higher LINC02323 expression relative to patients with a pathological response (Figure 1c). Additionally, patients were further grouped into low-LINC02323 and high-LINC02323 groups according to their average expression in gastric cancer patients [11,12]. A significant association was observed between LINC02323 and T stage (p = 0.005), lymph node metastasis (p = 0.045), and CAP grade (p = 0.019) in patients with gastric cancer (Table 1).

Table 1.

Association of LINC02323 with gastric cancer patients’ clinicopathological features.

	Cases (n = 117)	Low-LINC02323	High-LINC02323	P-value
Age				0.802
≤ 60	55	28	27
> 60	62	28	34
Gender				0.727
Male	67	33	34
Female	50	23	27
Differentiation				0.234
Well + moderate	86	44	42
Poor	31	12	19
T stage				0.005
T2-T3	84	47	37
T4a	33	9	24
Lymph node metastasis				0.045
Absent	77	42	35
Present	40	14	26
CAP grades				0.019
0–2	62	36	26
3	55	20	35

The high-LINC02323 group also showed a significantly lower overall survival rate than the low-LINC02323 group (Figure 2a). The prognostic significance of LINC02323 (HR = 8.004, 95% CI = 2.351–27.255) in gastric cancer was also confirmed by Cox regression analysis, as well as T stage (HR = 4.032, 95% CI = 1.353–12.015), lymph node metastasis (HR = 6.161, 95% CI = 2.064–18.386), and CAP grade (HR = 2.866, 95% CI = 1.051–7.811) in gastric cancer patients (Figure 2b).

Figure 2.

Higher LINC02323 levels indicated an adverse 5-year overall survival rate of gastric cancer patients (a) and served as an independent prognostic factor together with T stage, lymph node metastasis, and CAP grades (b).

LINC02323 negatively regulated miR-139-3p

From the Gene Expression Omnibus (GEO) database, 2001 and 194 abnormally expressed miRNAs in gastric cancer were filtered from the GSE23739 (Figure S3a) and GSE78091 (Figure S3b) datasets, respectively. The downstream target miRNAs of LINC02323 were predicted using the lncBoook and lncRNASNP databases. miR-139-3p, miR-765, and miR-1225-3p were enriched in both dysregulated miRNAs and LINC02323 downstream targets (Figure 3a), and only miR-139-3p was downregulated in gastric cancer (Figure S4a–c).

Figure 3.

a. Venn plots merging enriched dysregulated miRNAs in gastric cancer from the GEO database and LINC02323 downstream miRNAs from the lncBoook and lncRNASNP databases. b-c. Increased LINC02323 (b) and reduced miR-139-3p (c) levels were observed in gastric cancer cells. d-e. miR-139-3p negatively regulated the luciferase activity of LINC02323 (d) but showed no significant effect on the expression levels (e). f. Silencing LINC02323 could suppress miR-139-3p expression, which was reversed by transfection with the miR-139-3p inhibitor. ***p < 0.001 relative to CK; ^NSP > 0.05, ^##p < 0.01 relative to si-LINC02323.

In gastric cancer cells, significant upregulation of LINC02323 (Figure 3b) and miR-139-3p was significantly downregulated (Figure 3c). LINC02323 was predicted to bind to miR-139-3p through several sites, and the luciferase activity of the wild-type LINC02323 vector was negatively regulated by miR-139-3p in both HGC27 and MKN45 cells (Figure 3d).

In contrast, LINC02323 and miR-139-3p were found to be regulated in gastric cancer cells through cell transfection. LINC02323 was silenced by siRNA and was not affected by the miR-139-3p inhibitor (Figure 3e). The knockdown of LINC02323 significantly accelerated the expression of miR-139-3p, which was reversed by the miR-139-3p inhibitor (Figure 3f).

LINC02323 regulated the 5-FU sensitivity of gastric cancer cells via modulating the miR-139-3p/ABCC9 axis

Under the treatment of 5-FU, the viability of gastric cancer cells decreased with increasing 5-FU concentrations. Knockdown of LINC02323 significantly enhanced the suppressive effect of 5-FU on HGC27 and MKN45 cells (Figure 4a). In contrast, the knockdown of LINC02323 promoted apoptosis (Figure 4b) and suppressed the migration (Figure 4c) and invasion (Figure 4d) of 5-FU-treated gastric cancer cells, indicating the increased sensitivity of gastric cells to 5-FU. Additionally, miR-139-3p knockdown reversed the regulatory effect of LINC02323 on 5-FU treated cell viability (Figure 4a), apoptosis (Figure 4b), and motility (Figure 4c and d). Additionally, ABCC9 was predicted as a downstream target of miR-139-3p with several binding sites, and miR-139-3p was found to negatively regulate the luciferase activity of ABCC9 (Figure 5a). The co-regulatory effect of LINC02323 and miR-139-3p on the expression of ABCC9 was also demonstrated in 5-FU treated HGC27cell (Figure 5b).

Figure 4.

Silencing LINC02323 enhanced the suppressed cell proliferation (a), promoted cell apoptosis (b), and inhibited cell migration (c) and invasion (d) by 5-FU in HGC27 and MKN45 cells. ***p < 0.001 relative to CK; ^NSP > 0.05, ^##p < 0.01, ^###p < 0.001 relative to si-LINC02323.

Figure 5.

ABCC9 Was predicted to bind with miR-139-3p, and its luciferase activity was negatively regulated by miR-139-3p (a). Co-regulatory effect of LINC02323 and miR-139-3p on the expression of ABCC9 in 5-FU treated HCG27 cell. ***p < 0.001 relative to CK; ^NSP > 0.05, ^##p < 0.01 relative to si-LINC02323.

Discussion

Owing to improvements in next-generation sequencing and molecular biology, big data analysis based on public databases provides a reliable means of filtering biomarker candidates. This study integrated dysregulated ncRNAs in gastric cancer from three GEO databases, GSE224056, GSE143416, and GSE172032. Among the filtered ncRNAs, LINC02323 has been previously reported to be involved in the progression of various malignant tumors. LINC02323 accelerates the development of lung adenocarcinoma and lung squamous cell carcinoma and is correlated with poor prognosis of patients with lung cancer [9,10]. LINC02323 has been identified as a tumor promoter in ovarian cancer, facilitating tumor cell growth and migration [8]. Here, significant upregulation of LINC02323 was observed in patients with gastric cancer, which was closely related to advanced T stage, lymph node metastasis, and CAP grades. Moreover, higher LINC02323 levels also indicated a poorer 5-year overall survival rate of the enrolled patients and were identified as a prognostic biomarker together with CAP grade, lymph node metastasis, and T stage. This study focused on the role of LINC02323 in neoadjuvant chemotherapy of gastric cancer, a pre-chemotherapy before surgical treatment, therefore, patients with T2-T4a stages were included in this study. Considering the effect of chemotherapy on the expression of ncRNAs, LINC02323 was analyzed in the tissues collected before neoadjuvant treatment, ensuring the original state of LINC02323 and better evaluating its correlation with patient outcomes of chemotherapy [13,14]. CAP grading was employed to evaluate the tumor regression of gastric cancer patients according to the following criteria: CAP0 represents complete response without tumor cell remaining; CAP1 represents almost complete response with few remaining; CAP2 represents partial response with a resident tumor but definite regression; CAP3 represents poor or no response with extensive residual tumor [15,16]. A comparison of LINC02323 expression in the enrolled gastric cancer patients revealed that patients with no pathological response to neoadjuvant chemotherapy showed higher LINC02323 levels. Therefore, LINC02323 was speculated to correlate with chemotherapy efficiency, which was further validated in gastric cancer cells. In addition, LINC02323 upregulation was observed in the tissues collected during gastroscopy, which is easily obtained and provides potential biomarkers for predicting chemotherapeutic efficacy.

5-FU is a broad-spectrum antitumor agent commonly used in chemotherapy [17]. Resistance to 5-FU has become a major challenge in chemotherapy and neoadjuvant chemotherapy of gastric cancer. Recent studies have reported the function of lncRNAs in the regulation of chemotherapy resistance. For example, lncRNA EIF3J-DT was reported to induce resistance of gastric cancer cells to 5-FU by regulating autophagy [18]. CRNDE can alleviate resistance in gastric cancer by targeting SRSF6 and indirectly regulating PICALM alternative splicing [19]. The mechanism by which lncRNAs regulate chemotherapy resistance involves various processes such as cell apoptosis, autophagy, epithelial interstitial transformation, and regulation of related proteins [20–23]. Herein, silencing LINC02323 was found to enhance the inhibitory effect of 5-FU on gastric cancer cell growth and motility. Consistently, silencing LINC02323 promoted cell apoptosis under the treatment of 5-FU, suggesting that knockdown of LINC02323 could improve the sensitivity of gastric cancer cells to 5-FU and therefore suppress their resistance.

According to ceRNA theory, lncRNAs display functional roles by negatively modulating downstream miRNAs or target genes. LINC02323 was previously shown to sponge miR-1343-3p and regulate the development of lung and ovarian cancers [8–10]. In the present study, the sponging miRNAs of LINC02323 were predicted, and we further filtered the dysregulated miRNAs in gastric cancer. miR-139-3p was finally focused, which was downregulated in gastric cancer and negatively modulated by LINC02323. Moreover, miR-139-3p acts as a tumor suppressor in gastric cancer and mediates the regulatory effect of several lncRNAs on gastric cancer progression [24–27]. Although silencing miR-139-3p had no significant effect on LINC02323 expression, it attenuated the enhanced antitumor effect of 5-FU by LINC02323 knockdown. The regulatory effects of LINC02323 on cell stemness, autophagy, and related genes and proteins, such as multidrug resistance protein 1 (MDR1), multidrug resistance-associated protein 1 (MRP1), and other members of ATP-binding cassette family [28–30]. Among the ATP-binding cassette family, ABCC9 was predicted to bind with miR-139-3p and was negatively regulated by LINC02323 and miR-139-3p in 5-FU treated gastric cancer cells. Therefore, the miR-139-3p/ABCC9 axis was hypothesized as the downstream molecular mechanism underlying the regulation of gastric cancer development and chemotherapy resistance by LINC02323. These results provide novel targets to solve the chemotherapy resistance of gastric cancer. However, this is still a primary study revealing the clinical significance of LINC02323 in gastric cancer and its effect on chemotherapy. Although these findings implying the potential of LINC02323 as a biomarker for gastric cancer progression and chemotherapy efficiency, there is still a need expanding sample size to further verify the clinical results.

Conclusions

According to the above findings, LINC02323 upregulation in gastric cancer serves as a predictor of cancer development and chemotherapy outcomes in patients. LINC02323 regulates the sensitivity of gastric cancer cells to chemotherapy by modulating miR-139-3p.

Funding Statement

This study was funded by Shanghai Pudong New Area People’s Hospital Qihang Plan Talent Cultivation Project (PRYQH202405).

Ethical approval

This study was conducted in accordance with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Xuzhou Municipal Hospital Affiliated to XuZhou Medical University (Approval number: 2015107).

Consent to participate

All participants have signed informed consent.

Author contributions

ZXC, XDZ, ZTL and ZMW participated in the design of this study, and both performed the statistical analysis. HHZ conducted the study and collected important background information. ZTL was a major contributor to writing the manuscript. All authors have read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analyzed during this study are included in this article. Further enquiries can be directed to the corresponding author.