Expression of m6A Reader Protein YTHDC2 in Non-small Cell Lung Cancer in a Chinese Clinical Cohort

Summary

This study evaluates YTHDC2 expression in non-small cell lung cancer using bioinformatics and qRT-PCR. While database analyses suggest downregulation and prognostic relevance, clinical validation shows no significant difference. Findings highlight inconsistencies, indicating limited diagnostic accuracy and the need for further validation before clinical application.

Abstract

Non-small cell lung cancer (NSCLC) remains a leading cause of cancer-related mortality worldwide, with limited biomarkers available for early diagnosis and prognosis. N6-methyladenosine (m6A) RNA modification and its reader proteins, such as YTHDC2, play critical roles in gene regulation and tumorigenesis. This study aims to evaluate the expression profile of YTHDC2 in NSCLC tissues using bioinformatics and quantitative real-time PCR (qRT-PCR), assess the relationship between YTHDC2 and clinicopathological characteristics, and explore its potential clinical and biological relevance for future research. Gene expression and survival data from databases were analyzed for prognostic accuracy. YTHDC2 expression was quantified in NSCLC tissues in a Chinese clinical cohort, and clinicopathological characteristics were analyzed. Public database analyses showed that YTHDC2 was downregulated in NSCLC tissues (p < 0.05) and was associated with patient survival, although its prognostic performance was poor (AUC ≈ 0.5). qRT-PCR analysis of 19 paired tumor and adjacent normal tissues showed no statistically significant difference (p = 0.537) in YTHDC2 expression between cancer and adjacent tissues. Collectively, public database analyses suggest that YTHDC2 may be downregulated and have potential prognostic relevance in NSCLC, but independent clinical validation in our cohort did not confirm significant differential expression. These findings highlight substantial heterogeneity between large-scale datasets and real-world cohorts, indicating that YTHDC2 is unlikely to serve as a reliable standalone diagnostic or prognostic biomarker and may require integration with additional molecular markers for clinical applicability.

Introduction

Non-small cell lung cancer (NSCLC), encompassing squamous cell carcinoma and adenocarcinoma, is the most frequently diagnosed lung cancer and a leading cause of global cancer mortality¹. Global cancer statistics show that lung cancer is one of the main causes of cancer deaths². Recent developments in diagnosis, surgery, radiotherapy, and molecular targeted therapy have not improved the prognosis of NSCLC significantly because the primary diagnosis occurs in the late stage, and the disease is often asymptomatic in the early stages³. This underscores the need for new biomarkers to identify, prognosticate, and assess treatment effectiveness in NSCLC patients.

m6A is the first identified and most abundant RNA modification in eukaryotes and has been implicated in RNA stability, translation, and transcription⁴. m6A modifications are dynamic and reversible, facilitated by methyltransferases (writers), demethylases (erasers), and specific binding proteins (readers)⁵. Of these, YTHDC2, a major m6A reader protein, modulates RNA stability and translation rate⁶. Recent studies indicate that YTHDC2 is implicated in the regulation of multiple cancer types, either by suppressing cell growth and invasion or by inducing cell death⁷. Studies have also highlighted the clinical relevance of other m6A reader proteins, including YTHDF1, YTHDF2, and YTHDC1, in lung cancer progression, immune regulation, and prognosis, underscoring the importance of systematically evaluating individual m6A readers in independent patient cohorts^8,9,10. In NSCLC, there are significantly low levels of YTHDC2, which has been associated with high tumor stages, lymph node metastasis, and poor prognosis for patients, indicating that it could be an anti-tumor factor¹¹. In vitro and in vivo experiments also reveal that high expression of YTHDC2 inhibits cell proliferation and metastasis in lung cancer⁷.

In this study, YTHDC2 regulation in NSCLC was investigated through integrated computational analyses and experimental validation via qRT-PCR. The relationship between YTHDC2 expression and clinicopathological characteristics and patient prognosis was also investigated. Despite increasing evidence from public databases suggesting dysregulation of YTHDC2 in NSCLC, published studies have reported inconsistent findings regarding the magnitude and clinical significance of YTHDC2 dysregulation, and independent validation in well-characterized real-world clinical cohorts, particularly in Asian populations, remains limited. Moreover, most previous studies have relied predominantly on public transcriptomic datasets or experimental models, with relatively few integrating bioinformatics findings with independent clinical validation. By combining large-scale public transcriptomic analyses with independent qRT-PCR validation in a Chinese clinical cohort, this study aimed to assess the reproducibility of prior findings and to address the translational gap between public database analyses and real-world clinical samples. Therefore, this study aimed to systematically evaluate YTHDC2 expression using integrated bioinformatics and clinical validation approaches and to assess its potential biological and clinical relevance in NSCLC. It was hypothesized that YTHDC2 expression is dysregulated in NSCLC, and that integrating bioinformatics analyses with independent clinical validation would provide a more reliable assessment of its diagnostic and prognostic significance than either approach alone.

Protocol

The study was approved by the Ethical Committee of Hainan Medical University (approval number HMC1984.24) and was in accordance with the Declaration of Helsinki (as revised in 2013).

Study subjects

A study of 50 lung cancer patients diagnosed between 2017 and 2024 at Sanya Central Hospital in Hainan Province was conducted using the Chinese Medical Association Clinical Guidelines for the Diagnosis and Treatment of Lung Cancer (2024 Edition). CSCO staging (2024) aligned with the AJCC 8th Edition for cross-study comparability, as validated in Chinese cohorts¹², with two independent oncologist reviews to ensure consistency. The patient selection process, tissue availability, RNA quality assessment, and final sample inclusion for qRT-PCR analysis are summarized in Figure 1.

Study data

This study used bioinformatics to analyze YTHDC2 expression levels in NSCLC and their relationship with clinicopathological features, offering insights into potential mechanisms. Bioinformatics analyses were performed exclusively using publicly available RNA sequencing data from The Cancer Genome Atlas (TCGA), including lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), and corresponding normal lung tissue samples downloaded through the GDC Data Portal (https://portal.gdc.cancer.gov/). The downloaded dataset included RNA sequencing expression data together with available clinical variables (patient identifier, sample identifier, age, sex, pathological stage, survival status, and overall survival time) for eligible TCGA-LUAD and TCGA-LUSC cases. Samples lacking gene expression or survival information were excluded from downstream survival and ROC analyses. The dataset used in this study is provided as Supplementary File 1. No clinical specimens collected at Sanya Central Hospital were used for the bioinformatics analyses. Detailed procedures for the GEPIA, Kaplan–Meier Plotter, and survivalROC analyses are provided in the Bioinformatics analysis section below. The complete bioinformatics workflow, including data acquisition, preprocessing, gene expression analysis, survival analysis, and ROC analysis, is summarized in Figure 2.

Inclusion and exclusion criteria

The inclusion and exclusion criteria of the patients are presented in Table 1. A power analysis using statistical power analysis software determined that at least 26 cases per group would provide 80% power to detect a moderate effect size (d = 0.8, α = 0.05, two-tailed). While initial enrollment aimed for 50 pairs in the lung cancer group, the final qRT-PCR analysis included 30 tumor samples and 19 normal adjacent tissue samples after exclusions for tissue or RNA quality. This reduction in sample size reflects real-world clinical constraints and highlights the importance of RNA integrity and tissue availability in translational studies. At n = 19 for comparisons, the minimum detectable effect size is d = 1.0 (80% power, α = 0.05). Thus, the experiment was sufficiently powered to detect large, but not small-to-moderate, differences in YTHDC2 expression.

Tissue specimens and quantitative real-time PCR (qRT-PCR)

The pathological diagnoses were determined in a double-blind manner by two different pathologists. Tumor purity was estimated by pathologists (>70% malignant cells) and confirmed via ESTIMATE (TCGA). Adjacent normal tissues were macrodissected to minimize stromal contamination. The histological types of lung cancer included lung adenocarcinoma and squamous cell carcinoma, with 26 cases of lung adenocarcinoma and 4 cases of squamous cell carcinoma. The clinical staging of the 50 lung cancer patients was conducted according to the staging criteria outlined in “Chinese Medical Association Clinical Guidelines for Lung Cancer (2024 Edition)”. Pathologically confirmed NSCLC specimens (n = 50) represented typical clinical distributions (see Table 2). Of the 50 patients initially enrolled, 30 tumor tissue samples and 19 matched adjacent normal tissue samples met the RNA quality criteria. Because paired statistical analysis requires matched specimens from the same patient, the comparison of tumor versus adjacent normal tissue expression was performed using the 19 available matched pairs. These clinical specimens were used exclusively for experimental validation by qRT-PCR and were analyzed independently of the public TCGA datasets used for bioinformatics analysis. These selected specimens were processed immediately after pathological confirmation and handled under RNase-free conditions before RNA extraction. This subset reflects the cases in which both adequate tissue quantity and high-quality total RNA (RNA integrity number, RIN >7.0) could be obtained. Total RNA concentration and purity were measured before reverse transcription, and only samples with adequate RNA quality (RIN >7.0) were included for downstream analysis. Equal amounts of total RNA were reverse-transcribed into complementary DNA (cDNA) according to the manufacturer's protocol prior to quantitative PCR. This process ensured the validity of the gene expression data analyzed. The qRT-PCR experiments were carried out on a real-time PCR thermocycler using a probe-based quantitative PCR assay. All reactions were performed in triplicate, together with no-template controls to ensure analytical reproducibility. PCR amplification was performed under the following cycling conditions: an initial enzyme activation/denaturation step at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 s and annealing/extension at 60°C for 60 s. Fluorescence signals were acquired at the end of each amplification cycle. All reagents and consumables were obtained from commercial suppliers (see Table of Materials). The primer and probe sequences used in qRT-PCR are provided in Table 3.

Bioinformatics analysis

GEPIA database analysis of YTHDC2 gene expression

The GEPIA database was used to analyze YTHDC2 expression in NSCLC. The GEPIA web server (http://gepia.cancer-pku.cn/) was accessed through a web browser. The Expression DIY module was selected, the gene symbol "YTHDC2" was entered, the LUAD and LUSC datasets were selected, the default normalization parameters were retained, and differential expression box plots were generated directly through the GEPIA interface. Statistical significance was defined as p < 0.05.

Kaplan-Meier plotter database for survival analysis of lung cancer patients

The study analyzed the relationship between YTHDC2 expression and prognosis in lung cancer patients using the Kaplan-Meier Plotter database. The lung cancer dataset was selected, the gene symbol "YTHDC2" was entered, the auto-selected best cutoff option was applied, and Kaplan–Meier curves for overall survival and post-progression survival were generated using the default analysis settings. Patients were automatically stratified into high- and low-expression groups using the optimal cutoff determined by the Kaplan–Meier Plotter platform, and hazard ratios with corresponding 95% confidence intervals were generated using the platform's default settings.

Running R packages in R software for ROC curve plotting

RNA sequencing expression data and the corresponding clinical metadata for eligible TCGA-LUAD and TCGA-LUSC cases were downloaded from the GDC Data Portal. The downloaded datasets were merged by patient identifier and imported into R for downstream analyses. The survivalROC package was used to generate time-dependent ROC curves at 1-, 3-, and 5-year prediction time points, and the corresponding area under the curve (AUC) values were calculated to evaluate the prognostic performance of YTHDC2 expression. Only TCGA-LUAD and TCGA-LUSC patients with available RNA-seq expression and survival information were included in the survival ROC analysis. The local clinical cohort was not used for survival prediction because long-term follow-up data were unavailable.

Tissue YTHDC2 expression by qRT-PCR

To analyze gene expression levels, qRT-PCR was performed. Equal volumes of cDNA were added to each reaction according to the manufacturer's recommended reaction conditions. Amplification was performed using a probe-based quantitative PCR assay, and fluorescence data were collected automatically at the end of each amplification cycle. Briefly, the RNA extraction process involved several steps, including sample preparation, deparaffinization, removal of residual liquid, proteinase K digestion, incubation, centrifugation, DNase treatment, DNase I addition, and ethanol precipitation. The sample was then bound to a silica-based RNA purification column and centrifuged at 8,000 × g for 30 s. The column was then washed with wash buffer 1, wash buffer 2, and wash buffer 2 diluted with ethanol, and dried at 13,000 × g for 2 min. The RNA was then eluted by adding 70 µL of RNase-Free Water to the center of the column membrane, followed by centrifugation at 13,000 × g for 1 min. Each primer pair demonstrated a single amplification product, which was confirmed by melting-curve analysis before relative gene expression was calculated. The primer efficiency (90–110%) was validated using standard curves prior to sample analysis. Melting curve analyses confirmed the presence of single amplicons and the absence of primer dimers. Relative YTHDC2 expression was calculated using the 2^-ΔCt method, in which Ct values were normalized to the endogenous reference gene GAPDH. Because expression values were presented as normalized expression levels rather than fold changes relative to a calibrator sample, the results are reported as 2^−ΔCt values. GAPDH was selected as the housekeeping gene because its expression showed minimal variability (CV < 5%) compared with the tested alternatives (ACTB, CV = 12%; 18S rRNA, CV = 18%), consistent with reference gene selection criteria in m6A studies.

Statistical analysis

Statistical analyses were performed using statistical software, including the creation of graphics and charts. To analyze quantitative and categorical data on the YTHDC2 gene expression in NSCLC patients, R software was used for bioinformatics analyses and ROC curve generation. ROC curves and AUC were used to evaluate the diagnostic performance of YTHDC2 expression in predicting survival. For regression-based analyses, effect estimates (odds ratios) with corresponding 95% confidence intervals were reported where applicable. The Wilcoxon signed-rank test was used to compare YTHDC2 expression between paired tumor and adjacent normal tissue samples, while Pearson correlation analysis was used to assess the association between YTHDC2 expression and clinicopathological features. Correlation coefficients (r) and corresponding p-values were reported. Because the qRT-PCR data consisted of paired tumor and adjacent normal tissue samples from the same patients and the gene expression data were not normally distributed, the Wilcoxon signed-rank test was used to compare YTHDC2 expression levels between paired tissues. This test does not assume normality of the data and is commonly used for skewed biological data. All statistical tests were two-sided, and p <0.05 was considered statistically significant. Continuous variables were assessed for normality before analysis. Continuous variables are presented as mean ± standard deviation or median (interquartile range), as appropriate.

Results

Expression of the YTHDC2 gene in tumors based on the TCGA database

Figure 3 illustrates the association between NSCLC and YTHDC2 by examining YTHDC2 expression levels across different cancers using the GDC tool in the TCGA database.

Expression of the YTHDC2 gene in NSCLC in the GEPIA database

483 lung adenocarcinoma tissue samples, 347 normal lung tissue samples, 486 lung squamous cell carcinoma tissue samples, and 338 normal lung tissue samples were obtained from additional screening of YTHDC2 expression levels in lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) within the GEPIA database. Through statistical analyses, both lung adenocarcinoma and lung squamous cell carcinoma tissues were found to exhibit considerably lower levels of YTHDC2 expression than normal lung tissues (p < 0.05), as illustrated in Figure 4.

GEPIA analysis of YTHDC2 expression across pathological stages

The GEPIA (Gene Expression Profiling Interactive Analysis) database was used to generate a stage plot to evaluate YTHDC2 differential expression across NSCLC stages. As shown in Figure 5, no statistically significant differences in YTHDC2 expression were observed across pathological stages of NSCLC (p = 0.644).

Kaplan-Meier analysis of YTHDC2 and lung cancer survival

The Kaplan-Meier Plotter database was used to perform a Kaplan-Meier survival analysis of the YTHDC2 gene. According to the findings, patients with high YTHDC2 expression had significantly higher overall survival (OS) than those with low expression (p < 0.05). As shown in Figure 6A (OS), patients with lung cancer who expressed higher levels of YTHDC2 had a better prognosis, and this difference was statistically significant (p < 0.05). According to Figure 6B (PPS), the group with high YTHDC2 expression had a better prognosis than the group with low expression in the Post-Progression Survival (PPS) dataset; the difference was statistically significant (p < 0.05).

YTHDC2 survival rate prediction model

Time-dependent ROC analysis based on TCGA RNA-seq expression and survival data revealed limited predictive capability of YTHDC2 expression alone for NSCLC survival, with AUC values of 0.50 (95% CI: 0.38-0.62), 0.51 (95% CI: 0.39-0.63), 0.52 (95% CI: 0.40-0.64), and 0.52 (95% CI: 0.39–0.65) at 1, 3, 5, and 8 years, respectively (Figure 7). All AUC confidence intervals included 0.5, indicating performance equivalent to random chance. These findings indicate that YTHDC2 expression alone lacks discriminatory power for survival prediction and should not be considered a standalone prognostic biomarker. Integration with clinicopathological variables or multi-gene signatures may improve predictive performance.

YTHDC2 expression in cancer vs. normal tissues by qRT-PCR

YTHDC2 gene expression was investigated in both malignant and adjacent normal tissues from NSCLC patients using qRT-PCR. The normalized YTHDC2 expression (2^−ΔCt) was 3.24 ± 2.34 in malignant tissues and 3.60 ± 1.70 in nearby normal tissues. There was no significant difference in YTHDC2 expression between 19 paired NSCLC tumor tissues and their matched adjacent normal tissues (p = 0.537), as shown in Figure 8. While bioinformatics analysis of large datasets suggested significant downregulation of YTHDC2 in NSCLC, qRT-PCR in our cohort showed no significant difference, highlighting potential discrepancies due to cohort size, sample heterogeneity, and technical variability. Additionally, the low AUC (0.5) indicates that YTHDC2 alone lacks diagnostic or prognostic accuracy.

YTHDC2 expression and clinical pathological features in NSCLC

This study comprised 50 individuals with NSCLC. There were 32 male patients (64.00%) and 18 female patients (36.00%), with a mean age of 63.10 ± 9.85 years. They ranged in age from 37 to 86. Of the patients, 25 had smoked in the past, and 25 had never smoked. Nine cases of squamous cell carcinoma (18.00%) and forty-one cases of lung adenocarcinoma (82.0%) were among the pathological types. 22 patients were in stages III–IV (44.00%) and 28 in stages I–II (56.00%) according to the CSCO clinical staging. Twenty-four patients did not have lymph node metastases, while twenty-six individuals (54.00%) did. Nine patients had poorly differentiated cancers (18.00%), while 41 patients had well-to-moderately differentiated tumors (82.00%).

The association between clinical pathological characteristics and YTHDC2 expression levels in NSCLC tissues was examined. Differences in YTHDC2 expression were observed between pathological subtypes and according to lymph node metastasis status (p < 0.05). However, because only four squamous cell carcinoma samples were available, the pathological subtype comparison should be interpreted cautiously and considered exploratory. As shown in Figures 9A,B, YTHDC2 expression was significantly higher in squamous cell carcinoma than in lung adenocarcinoma and was significantly higher in NSCLC tissues with lymph node metastasis than in those without lymph node metastasis (5.70 ± 2.53 vs. 3.83 ± 0.91, p = 0.027). However, because only four squamous cell carcinoma samples were included, the comparison between pathological subtypes should be interpreted with caution and considered exploratory pending validation in larger cohorts. Higher YTHDC2 expression in patients with lymph node metastasis may indicate a potential association between YTHDC2 and lymph node metastatic status; however, this finding should be interpreted cautiously due to the limited sample size and requires validation in larger independent cohorts. However, age, smoking history, CSCO clinical stage, and histological differentiation did not significantly affect YTHDC2 expression (p > 0.05). Additional information is included in Table 4, Table 5, and Figure 9.

Binary logistic regression of lymph node metastasis in NSCLC

Consistent with the correlation analysis, YTHDC2 expression was significantly associated with lymph node metastasis (p < 0.05). Using lymph node metastasis as the dependent variable and age, gender, smoking history, tumor stage, pathological type, YTHDC2 expression level, and differentiation degree as independent variables, a binary logistic regression analysis was conducted in NSCLC patients. The results showed that YTHDC2 expression in NSCLC patients was significantly associated with lymph node metastasis (p = 0.027, OR = 2.286, 95% CI: 1.101–4.748).

In the general clinical data, the tumor stage of NSCLC patients was statistically significantly correlated with lymph node metastasis (p = 0.007, OR = 27, 95% CI: 2.504-291.186). However, age, gender, smoking history, pathological type, and differentiation degree were not significantly correlated with lymph node metastasis in NSCLC (p > 0.05), as shown in Table 6.

Data Availability: The datasets supporting the findings of this study are available in the Zenodo repository (DOI: 10.5281/zenodo.21409961). The repository includes the TCGA clinical metadata, sample annotations, data retrieval specifications, and analysis manifest used for the bioinformatics analyses. Additional data are available from the corresponding author upon reasonable request.

Figure 1: Flow diagram of patient selection and tissue inclusion for qRT-PCR analysis. Please click here to view a larger version of this figure.

Figure 2. Workflow of the bioinformatics analyses performed to evaluate YTHDC2 expression and prognostic significance in NSCLC Please click here to view a larger version of this figure.

Figure 3: Expression of the YTHDC2 Gene in the TCGA Database. YTHDC2 expression levels in lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) were compared with those in normal lung tissues using RNA sequencing data obtained from The Cancer Genome Atlas (TCGA) through the Genomic Data Commons (GDC) Data Portal and analyzed using the GEPIA web platform. Data were analyzed through the GEPIA platform. Statistical significance was determined with a p-value < 0.05. Please click here to view a larger version of this figure.

Figure 4: Differential expression of YTHDC2 in NSCLC based on GEPIA database analysis. YTHDC2 expression levels in lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) were compared with those in normal lung tissues using data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) projects via the GEPIA platform. Box plots represent normalized gene expression levels. Statistical significance was determined with a p-value < 0.05. Please click here to view a larger version of this figure.

Figure 5: Stage-wise expression analysis of YTHDC2 in NSCLC using the GEPIA database. YTHDC2 expression levels were analyzed across different pathological stages (I–IV) of NSCLC using the GEPIA platform. The stage plot illustrates variation in gene expression across tumor stages. No statistically significant differences in YTHDC2 expression were observed among stages (p = 0.644). Please click here to view a larger version of this figure.

Figure 6: Kaplan–Meier survival analysis of YTHDC2 expression in NSCLC patients (Kaplan–Meier Plotter). (A) Overall survival (OS): Kaplan–Meier survival curves comparing overall survival in NSCLC patients stratified into high (n = 140) and low (n = 364) YTHDC2 expression groups, using the optimal cut-off setting (“auto-select best cutoff”) provided by the Kaplan–Meier Plotter tool. The log-rank test was used to assess statistical significance (p = 0.0093); hazard ratio (HR) = 0.61; 95% confidence interval (CI): 0.42–0.89. Elevated YTHDC2 expression is associated with significantly better overall survival. (B) Post-progression survival (PPS): Kaplan–Meier survival curves comparing post-progression survival in NSCLC patients stratified into high (n = 181) and low (n = 296) YTHDC2 expression groups, using the optimal cut-off setting (“auto-select best cutoff”) provided by the Kaplan–Meier Plotter tool. The log-rank test was used for statistical comparison (p = 4.2 × 10⁻⁵); hazard ratio (HR) = 0.63 (95% confidence interval: 0.51–0.79). High YTHDC2 expression is associated with significantly longer post-progression survival. Please click here to view a larger version of this figure.

Figure 7: YTHDC2 survival rate prediction model. Receiver operating characteristic (ROC) curves evaluating YTHDC2's predictive accuracy for overall survival in TCGA-LUAD and TCGA-LUSC patients with available RNA-seq expression and survival data at 1-, 3-, 5-, and 8-year time points. AUC values: 1-year = 0.50, 3-year = 0.51, 5-year = 0.52, and 8-year = 0.52. Dashed diagonal line indicates random chance (AUC = 0.5). Analysis performed using the survival ROC package in R. Please click here to view a larger version of this figure.

Figure 8: YTHDC2 expression in paired NSCLC tumor and adjacent normal tissues measured by qRT-PCR. Relative YTHDC2 expression was calculated using the 2^–ΔCt method and normalized to GAPDH. Expression values represent normalized expression levels rather than fold changes relative to a calibrator sample. Statistical analysis was performed using the 19 matched tumor–adjacent normal tissue pairs available for paired comparison. Data are presented as individual paired observations with mean ± standard deviation. Differences between paired samples were analyzed using the Wilcoxon signed-rank test (p = 0.537). Please click here to view a larger version of this figure.

Figure 9: Association between YTHDC2 expression and clinicopathological characteristics in NSCLC patients. (A) YTHDC2 expression in different pathological types, showing higher expression in squamous cell carcinoma than in adenocarcinoma (p < 0.05). (B) YTHDC2 expression according to lymph node metastasis status, demonstrating higher expression in patients with lymph node metastasis than in those without metastasis (p < 0.05). (C) YTHDC2 expression across different histological differentiation grades, with no statistically significant difference (p = 0.181). (D) YTHDC2 expression across CSCO clinical stages, with no statistically significant difference (p = 0.08). YTHDC2 expression levels were measured by quantitative real-time PCR (qRT-PCR) and calculated using the 2^–ΔCt method, normalized to GAPDH. Data are presented as mean ± standard deviation (SD). Statistical comparisons were performed using the Wilcoxon rank-sum test. *p < 0.05 was considered statistically significant. Please click here to view a larger version of this figure.

S. No.	Inclusion criteria	Exclusion criteria
1	Age ≥18 years, regardless of gender.	Diagnosis of other malignancies.
2	Confirmed NSCLC diagnosis according to the Chinese Medical Association Clinical Guidelines for Lung Cancer (2024 Edition).	Presence of chronic respiratory or cardiovascular diseases (e.g., COPD, pulmonary heart disease, heart failure).
3	Availability of complete clinical data, including demographic details, serum tumor markers, and enhanced chest CT imaging, with signed informed consent.	Severe organ dysfunction, including renal insufficiency (eGFR <30 ml/min/1.73m²) or liver cirrhosis.
4	Treatment-naive patients with no prior tumor-directed therapies.	Special populations, including pregnant or lactating women (confirmed by β-hCG testing where applicable).
5	Availability of adequate tumor and/or adjacent tissue samples suitable for molecular analysis.	Poor specimen quality, including insufficient tissue or degraded RNA (confirmed by pathologist review).

Table 1: Inclusion and exclusion criteria for NSCLC patients in the study

Tissue type	Paired samples analyzed (n)	Relative expression (Mean ± SD)
Tumor tissue	19	3.24 ± 2.34
Matched adjacent normal tissue	19	3.60 ± 1.70

Table 2: Expression of YTHDC2 in Paired NSCLC Tumor and Adjacent Normal Tissues. Relative expression of YTHDC2 was measured using the 2^–ΔCt method and normalized to GAPDH. Data are presented as mean ± standard deviation (SD). Statistical comparison of the 19 matched tumor–adjacent normal tissue pairs was performed using the Wilcoxon signed-rank test (p = 0.537).

Target gene	Primer/Probe	Sequence (5′→3′)
YTHDC2	Forward (F)	CCTGTCACCAATAAAGAGCG
	Reverse (R)	CACTGGAATCTGAGGTATGCC
	Probe (P)	AGCAAGACAAGTGGGCGACTCAA
GAPDH	Forward (F)	AATCCCATCACCATCTTCCAG
	Reverse (R)	ATGACCCTTTTGGCTCCC
	Probe (P)	CCAGCATCGCCCCACTTGATTTT

Table 3: Primer and probe sequences used for quantitative real-time PCR (qRT-PCR). YTHDC2 expression was quantified using TaqMan chemistry, with GAPDH as the internal control.

Characteristic	Category	n	%
Pathological type	Adenocarcinoma	41	82
	Squamous cell carcinoma	9	18
CSCO staging	I–II	28	56
	III–IV	22	44
Lymph node metastasis	No	24	48
	Yes	26	52
Degree of differentiation	Low differentiation	9	18
	Moderate–high differentiation	41	82
Age	≥60 years	36	72
	<60 years	14	28
Gender	Male	32	64
	Female	18	36
Smoking history	Yes	25	50
	No	25	50

Table 4: Clinical and pathological characteristics of NSCLC patients. Data are presented as numbers (n) and percentages (%) based on the total study population (n = 50).

Clinicopathological feature	Group	n	YTHDC2 expression (Mean ± SD)	p-value
Tissue type	Tumor	30	3.24 ± 2.34	0.537
	Adjacent normal	19	3.60 ± 1.70
Pathological type	Adenocarcinoma	26	4.13 ± 1.29	0.022*
	Squamous cell carcinoma	4	7.50 ± 3.41
CSCO staging	I–II	22	4.29 ± 1.94	0.08
	III–IV	8	5.05 ± 1.51
Lymph node metastasis	No	19	3.83 ± 0.91	0.027*
	Yes	11	5.70 ± 2.53
Histological differentiation	Moderate–well differentiated	25	4.33 ± 1.90	0.181
	Poorly differentiated	5	5.18 ± 1.60
Age	≥60 years	20	4.62 ± 2.10	0.835
	<60 years	10	4.16 ± 1.12
Smoking history	Yes	12	4.62 ± 2.40	0.845
	No	18	4.38 ± 1.40

Table 5: Association between YTHDC2 expression and clinicopathological features in NSCLC patients. YTHDC2 expression was measured by qRT-PCR and calculated using the 2^–ΔCt method, normalized to GAPDH. Data are presented as mean ± standard deviation (SD). Group sizes (n) represent the number of valid samples analyzed. Statistical comparisons between groups were performed using the Wilcoxon rank-sum test. *p < 0.05 was considered statistically significant.

Variable	β	Wald	p-value	OR	95% CI
Age (≥60 vs <60)	0.07	0.01	0.93	1.08	0.20–5.68
Gender (Female vs Male)	0.53	0.4	0.525	1.7	0.33–8.67
Smoking History (Yes vs No)	0.54	0.46	0.496	1.71	0.36–8.09
Tumor Stage (III–IV vs I–II)	3.3	7.38	0.007*	27	2.50–291.18
Differentiation (Poor vs Moderate–well)	–1.29	1.62	0.204	0.28	0.04–2.02
YTHDC2 Expression	0.83	4.91	0.027*	2.29	1.10–4.75

Table 6: Multivariate logistic regression analysis of factors associated with lymph node metastasis in NSCLC patients. Binary logistic regression was performed using lymph node metastasis as the dependent variable. Odds ratios (ORs) with 95% confidence intervals (CI) are reported. Reference categories: age (<60 years), gender (male), smoking history (no), tumor stage (I–II), and differentiation (moderate–well). *p < 0.05 indicates statistical significance.

Discussion

YTHDC2, the largest N6-methyladenosine (m6A) binding protein in the YTH family, is unique for its ATP-dependent RNA helicase activity, distinguishing it as a key regulator of RNA metabolism¹³. Unlike other YTH family members, YTHDC2 is found in both the nucleus and the cytoplasm, where it regulates mRNA translation and RNA stability¹⁴. Previous experimental studies have shown that YTHDC2 suppresses lung adenocarcinoma (LUAD) progression by inhibiting ADIRF and MRPL12 mRNA expression, enhancing apoptosis, and reducing cell proliferation and metastasis^15,16. Previous experimental studies have also suggested that YTHDC2 stabilizes lncRNAs such as ZNRD1-AS1 to suppress LUAD proliferation and modulates m6A-modified mRNAs such as CYLD and SLC7A11, inhibiting NF-κB signaling pathway, cysteine uptake, and antioxidant processes, while promoting ferroptosis and anti-cancer activity^17,18.

In this study, the expression and clinical significance of the m6A reader protein YTHDC2 in NSCLC were investigated using an integrative approach combining bioinformatics analyses and experimental validation. This bioinformatics analysis of TCGA and GEPIA databases revealed a significant downregulation of YTHDC2 in NSCLC tumor tissues compared to normal lung tissues, and higher YTHDC2 expression was associated with improved overall and post-progression survival, suggesting a potential prognostic role. Interestingly, our qRT-PCR experiments using the 19 available paired tumor and adjacent normal tissue samples in the present cohort failed to show a statistically significant difference in YTHDC2 expression between tumor and adjacent normal tissues. This discordance highlights the challenges of translating findings from large-scale public datasets into independent clinical cohorts and underscores the need for cautious interpretation of in silico analyses. Understanding the reasons behind these differences is critical for accurately evaluating the utility of YTHDC2 as a biomarker and therapeutic target in NSCLC. The present study’s focus on YTHDC2 aligns with its unique helicase activity among YTH readers¹⁹, but recent pan-cancer analyses reveal context-dependent roles for other family members. For instance, Li et al. demonstrated that YTHDF1/2 are upregulated in most cancers (e.g., LIHC, LUAD) and correlate with poor prognosis, while YTHDC1/2 show tumor-suppressive effects in KIRC and BRCA²⁰. Notably, YTHDF1’s oncogenic role via Wnt/β-catenin activation contrasts with YTHDC2’s anti-metastatic function in NSCLC, highlighting the need for subtype-specific biomarker strategies. Compared to established prognostic markers (e.g., PD-L1 for immunotherapy response), YTHDC2’s predictive power may be enhanced by combining it with other m6A regulators (e.g., METTL3, YTHDF1) or immune checkpoint genes, as suggested by pan-cancer immune infiltration analyses. Although previous studies suggest that YTHDC2 regulates the NF-κB/CYLD pathway¹², the present study was limited by the absence of protein-level validation using immunohistochemistry or Western blotting, as well as functional validation. Future work should combine knockdown assays with m6A-sequencing to map direct targets.

Previous analyses revealed that YTHDC2 expression is downregulated in NSCLC tissues and correlates with poor prognosis²¹. Kaplan-Meier analysis indicated longer survival in patients with higher YTHDC2 expression, but ROC curve analysis suggested limited predictive power for survival²⁰. The survival ROC analysis in the present study was performed using TCGA RNA-seq expression and survival data, as long-term follow-up information was not available for our local clinical cohort. Importantly, the low AUC values (~0.5) observed in this study suggest that YTHDC2 expression lacks sufficient sensitivity and specificity for clinical use as a standalone biomarker²². This limitation aligns with emerging evidence that single-gene biomarkers are often insufficient for complex diseases such as NSCLC, where multi-omics approaches and integrated biomarker panels are increasingly required to achieve clinically meaningful predictive performance. Higher YTHDC2 expression was observed in NSCLC tissues with lymph node metastasis, suggesting a potential association between YTHDC2 expression and lymph node metastatic status in the present cohort. Although differences in YTHDC2 expression were also observed between adenocarcinoma and squamous cell carcinoma, this comparison should be interpreted cautiously, as only 4 squamous cell carcinoma samples were available for analysis, substantially limiting the statistical power of this subgroup comparison. No significant differences were observed in clinical stage, differentiation, age, or smoking history, likely due to study limitations, including sample size and a single-center design. Although the observed association between reduced YTHDC2 expression and lymph node metastasis suggests a potential biological role in NSCLC, its poor discriminatory power (AUC ~0.5) and inconsistent detection by qRT-PCR preclude its use as a standalone clinical marker. Future studies should explore whether combining YTHDC2 with other m6A-related proteins improves predictive value.

The bioinformatics analyses suggested that YTHDC2 expression is reduced in NSCLC and may be associated with better patient outcomes. Although reduced expression was also observed in patients with lymph node metastasis and differences were noted between pathological subtypes, these subgroup findings should be interpreted cautiously because of the limited sample size, particularly for squamous cell carcinoma. However, these findings were not confirmed by qRT-PCR analysis in our clinical cohort, possibly reflecting differences in cohort characteristics, sample size, technical variability, post-transcriptional regulation, or the absence of protein-level validation. Future research should explore YTHDC2 mechanisms in other lung cancer subtypes through in vitro and in vivo experiments. Such studies may further clarify the biological role of YTHDC2 and determine whether it has any clinical utility as part of a multimarker approach rather than as a standalone biomarker.

Limitations

This study has several limitations. It was a single-center retrospective study with a relatively small sample size, which may limit generalizability. The use of FFPE tissues may have affected RNA quality and qRT-PCR sensitivity. Additionally, only YTHDC2 mRNA expression was evaluated, and protein-level validation using immunohistochemistry (IHC) or Western blotting was not performed. Because mRNA expression does not necessarily correlate with protein abundance owing to post-transcriptional regulation, this may partially explain the discrepancy between our clinical findings and public database analyses. The study also lacked long-term follow-up data, and subgroup analyses (e.g., squamous cell carcinoma) were underpowered.

Future directions

Future research should explore YTHDC2 mechanisms in NSCLC through in vitro and in vivo experiments to better understand its functional role. Multi-center studies with larger sample sizes are required to validate the findings and improve generalizability. In addition, integrating YTHDC2 with other m6A-related regulators and clinicopathological variables may enhance its predictive and prognostic utility. Further studies incorporating multi-omics approaches, including transcriptomics, proteomics, and epigenomics, are necessary to clarify the biological mechanisms underlying YTHDC2 dysregulation. Protein-level validation by immunohistochemistry (IHC) or Western blotting should also be performed to assess concordance between mRNA and protein expression. Standardization of sample processing and validation techniques will also be critical to ensure reproducibility and clinical applicability.

Conclusion

This study provides an exploratory evaluation of YTHDC2 expression in NSCLC by integrating bioinformatics analyses with clinical qRT-PCR validation. However, its inconsistent detection across analytical platforms and limited diagnostic accuracy (AUC ≈0.5) currently preclude its use as a standalone clinical biomarker. While bioinformatics analyses suggest potential prognostic significance, independent clinical validation does not support its reliability for diagnostic or prognostic application. These findings should be interpreted cautiously and require validation in larger multicenter cohorts with protein-level and functional studies before any clinical application can be considered.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Custom Primers and Probes	Sangon Biotech Co., Ltd., Shanghai, China	Custom synthesis	Primer and probe sequences listed in Table 6
DNase I	Magen Biotechnology Co., Ltd., Guangzhou, China	DNase I (HiPure FFPE RNA Kit)	Removal of genomic DNA contamination
ESTIMATE algorithm	Yoshihara et al.	R package (Version 1.0.13)	Tumor purity estimation
FFPE RNA Extraction Kit	Magen Biotechnology Co., Ltd., Guangzhou, China	HiPure FFPE RNA Kit (R4130-02)	RNA extraction from FFPE tissue
GDC Data Portal	National Cancer Institute (NCI), NIH, USA	https://portal.gdc.cancer.gov	Download of TCGA datasets
GEPIA web server	Peking University, China	http://gepia.cancer-pku.cn	Gene expression analysis
Kaplan–Meier Plotter	Semmelweis University, Hungary	https://kmplot.com	Survival analysis
Microcentrifuge	Eppendorf AG, Hamburg, Germany	5424 R	RNA purification procedures
NanoDrop Spectrophotometer	Thermo Fisher Scientific, Waltham, MA, USA	NanoDrop 2000/2000c	RNA quantification
Probe-based qPCR Master Mix	Thermo Fisher Scientific, Waltham, MA, USA	TaqMan Universal PCR Master Mix II (4440040)	Quantitative PCR amplification
R Statistical Software	R Foundation for Statistical Computing, Vienna, Austria	Version 4.4.1	Statistical analyses and ROC analysis
Real-time PCR System	Bio-Rad Laboratories, Hercules, CA, USA	CFX Opus 96	Quantitative real-time PCR
Reverse Transcription Kit	Thermo Fisher Scientific, Waltham, MA, USA	RevertAid First Strand cDNA Synthesis Kit (K1622)	cDNA synthesis
RNA Purification Column	Magen Biotechnology Co., Ltd., Guangzhou, China	HiPure Mini Column I	Silica membrane RNA purification column
RNase-Free Water	Magen Biotechnology Co., Ltd., Guangzhou, China	RNase-Free Water	RNA elution
SPSS Statistics	IBM Corp., Armonk, NY, USA	IBM SPSS Statistics Version 27.0	Statistical analyses
survivalROC package	CRAN (R Foundation for Statistical Computing)	Version 1.0.3.1	Time-dependent ROC analysis
TCGA database	National Cancer Institute (NCI), NIH, USA	https://portal.gdc.cancer.gov	Public cancer genomics database
Wash Buffer RW1	Magen Biotechnology Co., Ltd., Guangzhou, China	Buffer RW1	RNA purification wash buffer
Wash Buffer RW2	Magen Biotechnology Co., Ltd., Guangzhou, China	Buffer RW2	RNA purification wash buffer