---
title: "m5C-linked NSUN5–YBX1–CENPF programme associates with melanoma cell states and immune features"
id: "frontiers-in-immunology-6-an-m5c-associated-nsun5-ybx1-cenpf-programme-marks-melanoma-cell-states-and"
canonical_url: "https://medichelpline.com/clinical-feed/frontiers-in-immunology-6-an-m5c-associated-nsun5-ybx1-cenpf-programme-marks-melanoma-cell-states-and"
content_type: "clinical_feed_article"
specialty: "Oncology"
source_name: "Frontiers in Immunology"
source_url: "https://www.frontiersin.org/articles/10.3389/fimmu.2026.1952519"
published_at: "2026-10-16T00:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# m5C-linked NSUN5–YBX1–CENPF programme associates with melanoma cell states and immune features
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/frontiers-in-immunology-6-an-m5c-associated-nsun5-ybx1-cenpf-programme-marks-melanoma-cell-states-and
- **Specialty:** [Oncology](https://medichelpline.com/clinical-feed/oncology.md)
- **Primary Source:** Frontiers in Immunology
- **Source URL:** [Original Journal Publication](https://www.frontiersin.org/articles/10.3389/fimmu.2026.1952519)
- **Published At:** 2026-10-16T00:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- The paper title indicates investigators reported a molecular programme involving **m5C** RNA modification and the genes **NSUN5**, **YBX1**, and **CENPF** that marks cellular states in **melanoma**. - The source did not include the manuscript text in the provided material; specific experimental design, cohort composition, or statistical endpoints were not reported. - No numeric results, p-values, sample sizes, or effect sizes were available in the supplied source. - The title asserts a link between this programme and tumour immune features; however, the source did not detail which immune cell populations, biomarkers, or spatial/functional assays were used. - The title implies correlative or profiling work (molecular programme “marks” states), but the source did not state whether analyses were observational, mechanistic perturbations, single-cell, bulk transcriptomics, or immunohistochemistry. - The authors apparently focused on cell-state annotation in melanoma and on immune characteristics, but the source omitted information on clinical correlates such as prognosis, therapy response, or stage. - Key unanswered items in the provided material include methods for detecting **m5C**, how **NSUN5**, **YBX1**, and **CENPF** were linked (co-expression, functional interaction), and whether findings were validated in independent cohorts or functional assays. - Readers should interpret the title-level claim cautiously until the full article is available; the supplied source did not permit verification of conclusions or replication details.
## Clinical Analysis & Structured Key Points
Frontiers | An m5C-associated NSUN5–YBX1–CENPF programme marks melanoma cell states and immune features ORIGINAL RESEARCH article Front. Immunol. , 16 October 2026 Sec. Cancer Immunity and Immunotherapy Volume 17 - 2026 | https://doi.org/10.3389/fimmu.2026.1952519 Published in Frontiers in Immunology Cancer Immunity and Immunotherapy 7 impact factor 11.3 citescore Editor & Reviewers Edited by W M Wenxue Ma Reviewed by A F Atsushi Fujimura C G Carlos Gomez Outline Figures and Tables Figure 1 View in article Figure 2 View in article Figure 3 View in article Figure 4 View in article Figure 5 View in article Figure 6 View in article Figure 7 View in article ORIGINAL RESEARCH article Front. Immunol. , 16 October 2026 Sec. Cancer Immunity and Immunotherapy Volume 17 - 2026 | https://doi.org/10.3389/fimmu.2026.1952519 An m 5 C-associated NSUN5–YBX1–CENPF programme marks melanoma cell states and immune features A L Aimei Liu 1,2,3 † K Y Kun Yuan 3 † S Y Sha Yi 4 † H M Hanxiao Mao 1,3 X X Xinyu Xie 3 S Y Shiyang Yin 3 H F Haonan Fan 3 A Z Ailin Zhang 3 X L Xixu Liu 3 Y H Yifu Hu 3 A X Aiguo Xu 1 * W C Wei Chen 2 * +4 more Y H Yuanmin He 1 * 1. Skin Structure and Function Key Laboratory of Luzhou, Department of Dermatology, The Affiliated Hospital of Southwest Medical University, Luzhou, Sichuan, China 2. Department of Dermatology, The Second Affiliated Hospital of Chengdu Medical College, China National Nuclear Corporation 416 Hospital, Chengdu, Sichuan, China 3. Clinical Medical College, Southwest Medical University, Luzhou, Sichuan, China 4. Department of Dermatology, Chengdu Integrated TCM & Western Medicine Hospital, Affiliated Hospital of Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, China See more Article metrics View details Abstract Background: RNA 5-methylcytosine (m 5 C)-associated programmes may help characterise melanoma cell states and their immune context. We investigated an NSUN5–YBX1–CENPF (NYC) programme using patient-matched sequencing, experimental perturbation and multiscale transcriptomic analyses. Methods: We analysed m 5 C-MeRIP-seq and matched Input RNA data from three melanoma–adjacent-normal patient pairs, using a paired edgeR interaction model for regional IP enrichment. Tissue and cell experiments included NSUN5 knockdown, CENPF-targeted MeRIP–qPCR and western blotting. Clinical analyses considered proliferation and sampling site. Trajectory and regulatory inference were restricted to malignant cells from GSE215120, with patient-level summaries. Treatment-associated changes were examined in 43 paired patients from GSE91061. Computational perturbation and transcriptomic signature reversal prioritised candidate regulators and compounds. Results: Two CENPF regions showed increased relative IP enrichment after Input adjustment (4.97- and 5.85-fold; BH q = 0.0457 and 0.0074). NSUN5 knockdown reduced CENPF m 5 C enrichment and protein abundance. NYC was associated with proliferation; its survival association did not persist after proliferation adjustment. Within malignant cells, NYC-High retained an association with an immune-resistance expression programme after G2M/E2F and depth adjustment (nine patients; BH q = 0.0156). Regulatory analyses prioritised MLX, whose simulated knockout and overexpression produced opposing reference-flow alignments (BH q = 0.0134 and 0.0201). MLX regulon activity was associated with glycolysis and oxidative phosphorylation after adjustment. In GSE91061, responders showed a greater treatment-associated NYC decrease than non-responders (mean change difference −0.209; permutation P = 0.0138). Drug screening generated candidates for experimental testing. Conclusions: NYC provides an m 5 C-associated framework connecting patient-derived molecular observations to malignant-cell states and immune features. YBX1 remains a candidate reader, and regulatory, treatment-associated and drug-prioritisation findings do not establish a complete causal pathway or a clinically validated predictive biomarker. 1 Introduction Cutaneous melanoma has a strikingly stage-dependent clinical course: early-stage disease can often be durably controlled by complete surgical excision, whereas advanced melanoma continues to impose a substantial mortality burden ( 1 , 2 ). Immune checkpoint blockade targeting CTLA-4 and PD-1 has markedly prolonged survival in a subset of patients with advanced disease, with treatment strategies evolving from monotherapy to combination regimens ( 3 – 5 ). Long-term follow-up has confirmed the durability of benefit from combined immunotherapy, whilst dual BRAF/MEK inhibition has reshaped the therapeutic landscape for patients with BRAF-mutant melanoma ( 6 , 7 ). Nevertheless, primary resistance, acquired resistance and treatment-driven co-evolution of tumour cells and their microenvironment continue to preclude curative outcomes for many patients ( 8 , 9 ). Molecular classification by The Cancer Genome Atlas further revealed extensive heterogeneity in melanoma driver alterations and transcriptional states, indicating that canonical mutations alone cannot fully account for the diversity of clinical outcomes ( 10 ). RNA 5-methylcytosine (m 5 C) provides a dynamic regulatory layer, distinct from DNA sequence alterations, through which such cellular plasticity may be understood. Transcriptome-wide m 5 C deposition, together with its installation and recognition, can influence RNA abundance, nuclear export, stability and translation ( 11 – 13 ). Y-box-binding protein 1 ( YBX1 ) has been established as an m 5 C reader that recognises modified transcripts and protects them from degradation in diverse biological contexts ( 14 – 16 ). Its broad RNA-binding capacity, together with evidence that it cooperates with NSUN-family methyltransferases to stabilise cancer-related mRNAs, positions YBX1 as a plausible molecular link between RNA modification and transcript fate ( 17 – 19 ). NSUN5 is best known for catalysing m 5 C modification of ribosomal RNA and thereby regulating protein synthesis and adaptation to cellular stress ( 20 , 21 ). More recently, NSUN5 has also been implicated in tumour immune evasion and in the progression of renal and liver cancers ( 22 – 24 ). Direct evidence further supports its ability to regulate cancer-associated mRNAs through m 5 C modification ( 25 ). Centromere protein F ( CENPF ), in turn, functions as a downstream effector of cell-cycle progression and chromosome segregation and can cooperate with FOXM1 -associated regulatory networks to sustain malignant proliferation ( 26 , 27 ). Its prognostic association and transcriptional regulation in cutaneous melanoma further support CENPF as a candidate downstream node ( 28 , 29 ). Paired sequencing of tumour and normal tissues from the same patient can capture coordinated changes in RNA modification and gene expression. However, tissue-averaged signals alone cannot determine which cell populations harbour these alterations, which malignant states they represent, or whether they are linked to the surrounding immune and stromal ecosystem. Single-cell atlases of melanoma have revealed substantial complexity in malignant cell states, T-cell exclusion and treatment-persistent transcriptional programmes ( 30 – 32 ). Studies of transcriptional regulation and cellular trajectories have further identified continuous relationships amongst invasive state transitions, recurrent melanoma cell states and dedifferentiation from benign to malignant phenotypes ( 33 – 35 ). Immune-cell composition in treatment cohorts is likewise closely associated with responses to immune checkpoint blockade ( 36 ). Integrating bulk cohorts, single-cell and spatial transcriptomic data, and gene-regulatory networks is therefore necessary to localise epitranscriptomic alterations to malignant cell populations, define their tissue context and establish priorities for state-directed therapeutic intervention. In this study, we first jointly analysed m 5 C methylated RNA immunoprecipitation sequencing (m 5 C-MeRIP-seq) and RNA sequencing data from patient-matched melanoma and normal tissues. On the basis of coordinated changes in CENPF m 5 C enrichment and expression, we proposed a NSUN5–YBX1–CENPF (NYC) working programme. We then experimentally anchored the key molecular associations in patient tissues and cellular models and integrated TCGA, single-cell and spatial transcriptomic datasets to characterise the clinical phenotype, cellular origin, regulatory architecture and tumour ecosystem of the NYC programme. Finally, we combined bulk transcriptomic signature reversal with single-cell drug-connectivity analysis in a two-stage screening framework to prioritise therapeutic candidates predicted to selectively reverse the NYC-High state. 2 Results 2.1 Integrated analysis of paired transcriptomic and epitranscriptomic profiles identifies a candidate NSUN5–YBX1–CENPF programme To derive experimentally testable molecular hypotheses from our in-house sequencing data, we performed parallel m 5 C-MeRIP-seq and RNA sequencing on three patient-matched pairs of melanoma and adjacent normal tissues. Candidate molecular relationships were jointly prioritised on the basis of three criteria: increased m 5 C enrichment, altered transcript abundance and covariation with potential m 5 C recognition or regulatory factors ( Figure 1 ). Within this discovery framework, CENPF showed higher peak-call counts and transcript abundance in melanoma, motivating a paired analysis of regional enrichment relative to local Input abundance. The m 5 C methyltransferase NSUN5 and YBX1 , considered here as a candidate reader, provided a plausible upstream regulatory framework for these changes. On the basis of these observations, we defined NSUN5 , YBX1 and CENPF as the NYC three-gene programme and subsequently evaluated its biological and translational relevance in melanoma using clinical cohorts, tissue- and cell-based experiments, and multiscale bioinformatic analyses. Figure 1 Study workflow and evidence levels of the NYC programme. The five modules distinguish measured observations, statistical associations and computational predictions. Module 1 summarises paired MeRIP-seq and Input RNA measurements from three melanoma-adjacent-normal tissue pairs, Input-adjusted regional enrichment and the candidate NYC working model. Dashed links denote candidate molecular relationships; the YBX1-CENPF interaction has not been established experimentally. Module 2 presents tissue expression and the pretreatment GSE91061 survival association; the displayed 50-patient analysis uses the revision-stage one-month survival restriction described in Methods and does not establish validated clinical prediction. Module 3 shows measured changes in CENPF MeRIP-qPCR enrichment and protein abundance after NSUN5 knockdown. Module 4 presents malignant-only NYC-Low/NYC-High state analyses, exploratory Low-rooted ordering, within-section spatial associations and MLX perturbation simulations; these do not demonstrate benign-to-malignant development or causal regulation. Module 5 separates treatment-associated NYC change in 43 paired patients from computational DrugReflector/scFOCAL prioritisation and predicted state selectivity. Large arrows indicate workflow sequence only, not causal relationships or increasing levels of evidence. Quantitative details are reported in the corresponding main figures and Methods. The subsequent analyses were structured around three interrelated questions. First, is m 5 C enrichment of CENPF mRNA reproducibly detected in melanoma, and is it associated with expression of the three NYC genes and patient outcomes? Second, does high NYC activity define a distinct malignant cell state with a characteristic tumour microenvironment and regulatory architecture? Third, can the transcriptional features of the NYC-High state be used to prioritise candidate compounds with the potential to reverse this state? This framework established a continuous line of investigation extending from discovery in patient-matched sequencing data and experimental validation to single-cell and spatial dissection, followed by in silico perturbation and drug prioritisation. 2.2 Paired Input-adjusted analysis identifies increased CENPF regional IP enrichment Across three melanoma specimens (A1–A3) and their matched adjacent normal tissues (B1–B3), the number of m 5 C peaks assigned to CENPF was markedly higher in the tumour samples. Specifically, 14, 6 and 2 CENPF -associated peaks were detected in A1, A2 and A3, respectively, whereas 0, 0 and 1 peak were detected in B1, B2 and B3, respectively. In aggregate, 22 peaks were identified in the melanoma tissues, compared with only one in the matched normal tissues ( Figure 2A ). Sample-resolved interval tracks showed the positions of called peaks across the CENPF locus ( Figure 2B ). These tracks display peak calls rather than read-coverage intensity. Figure 2 Paired CENPF enrichment reanalysis and clinical associations of the NYC programme (A) , Number of CENPF-associated m 5 C peaks called in three patient-matched pairs of adjacent normal tissues (B1–B3) and melanoma tissues (A1–A3) . Lines connect matched samples. (B) , Positions of sample-resolved m 5 C peak calls across the CENPF locus; these are called intervals rather than read-coverage tracks. (C) , Background-normalised IP/Input ratios at P1–P3 in three matched patient pairs. Shapes identify patients and lines join matched tissues. Enrichment fold changes and BH q values derive from the paired edgeR IP-by-condition interaction model, with correction across 7,937 candidate regions; they are not tests performed directly on the displayed ratios. (D) , Transcriptome-wide metagene density of m 5 C peaks across the 5′ UTR, CDS and 3′ UTR in normal and melanoma samples. (E) , Top-ranked DREME motifs identified in the two tissue groups. (F–H) , Differential expression of NSUN5, YBX1 and CENPF between normal skin and SKCM tissues. (I–K) , Overall-survival curves stratified by high versus low expression of NSUN5, YBX1 and CENPF. (L) , Overall survival in the TCGA-SKCM cohort stratified at the median NYC ssGSEA score; n = 456, HR = 1.348, 95% CI 1.032–1.760, P = 0.0283. The three originally nominated CENPF regions were P1 (chr1:214603194–214603321), P2 (chr1:214613713–214613916) and P3 (chr1:214648674–214648827), using zero-based, half-open BED coordinates. Paired edgeR quasi-likelihood interaction analysis estimated melanoma-to-normal relative IP/Input enrichment changes of 4.97, 5.85 and 2.61, respectively ( Figure 2C ). P1 and P2 met the BH threshold across 7,937 candidate regions (q = 0.0457 and 0.00742), whereas P3 did not (q = 0.1241). Patient-resolved normalised IP/Input ratios are shown alongside the model estimates. P1 was called in all three melanoma samples, P2 in two, and P3 predominantly in A1 ( Supplementary Figures S1A, B ). These results support regional relative enrichment changes after accounting for local Input abundance; they do not measure absolute methylation fractions. Full edgeR results and counts are supplied in Supplementary Data R2. Transcriptome-wide metagene analysis revealed distinct distributions of m 5 C peak density across the 5′ UTR, CDS and 3′ UTR between melanoma and normal tissues ( Figure 2D ). DREME analysis identified WGGARGA and HGGARGA as the top-ranked enriched sequence motifs across the two groups ( Figure 2E ), delineating the sequence context of m 5 C-enriched regions in the respective tissue types. Paired analysis of the Input libraries estimated a 2.20-fold increase in CENPF transcript abundance ( P = 0.00313; genome-wide BH FDR = 0.132). Across all six sequenced specimens, YBX1 expression correlated positively with CENPF expression (Spearman’s ρ = 0.829, P = 0.0416). These expression and regional-enrichment observations provide discovery-level support for the NYC working model; the six-specimen expression correlation remains exploratory. 2.3 Elevated expression and composite activity of the NYC programme are associated with adverse outcomes in melanoma Publicly available expression datasets further supported aberrant activation of the three NYC genes in melanoma. Compared with normal skin, NSUN5 , YBX1 and CENPF were all expressed at significantly higher levels in skin cutaneous melanoma tissues ( NSUN5 , P < 2.22 × 10 −16 ; YBX1 , P = 2.8 × 10 −13 ; CENPF , P < 2.22 × 10 −16 ; Figures 2F–H ). Kaplan–Meier analyses using gene-specific optimised cut-offs showed that high expression of NSUN5 , YBX1 and CENPF was individually associated with poorer overall survival ( P = 0.001, 0.009 and 0.004, respectively; Figures 2I–K ). Cross-cohort Cox analyses showed a broadly concordant direction of association with adverse prognosis, although the magnitude of the effects varied amongst patient populations ( Supplementary Figures S1C–E ). To assess the clinical relevance of the three genes as a unified biological programme, we calculated an NYC ssGSEA score using the TCGA-SKCM log-transformed TPM expression matrix, which comprised 19,344 genes and 473 patients. Patients were subsequently dichotomised at the median score into NYC-High (n = 236) and NYC-Low (n = 237) groups. Amongst the 456 patients with complete survival information, the NYC-High group had significantly shorter overall survival than the NYC-Low group (Cox HR = 1.348, 95% CI 1.032–1.760, P = 0.0283; Figure 2L ). Schoenfeld residual testing showed no evidence of violation of the proportional-hazards assumption, with P = 0.91 for both the group-specific and global tests. These results indicate that coordinated high activity of the NYC programme delineates a subgroup of patients with melanoma who have an unfavourable clinical outcome. These unadjusted associations were reassessed in a common complete-case set of 378 patients with 178 deaths. NYC correlated with Hallmark G2M checkpoint and E2F target scores (Spearman ρ = 0.734 and 0.747). The clinical-covariate-adjusted HR per SD of NYC was 1.22 (95% CI 1.04–1.43; nominal P = 0.0136), but the association did not persist after G2M or E2F adjustment (P = 0.478 and 0.580). The data therefore do not establish proliferation-independent prognostic value (Supplementary Data R1). 2.4 Tissue- and cell-based experiments support NSUN5 -dependent m 5 C enrichment of CENPF and maintenance of CENPF protein abundance We first assessed the expression of the three NYC genes at the tissue and cellular levels. Representative high-magnification fields, each traceable to its source whole-slide image, showed an overall increase in immunohistochemical staining for NSUN5 , YBX1 and CENPF from normal skin through benign naevi to melanoma, with more extensive positive staining in melanoma regions ( Figure 3A ). Corresponding whole-slide overviews are provided in Supplementary Figure S2 . These histological patterns were consistent with the direction of expression changes observed in public transcriptomic datasets. Figure 3 Tissue- and cell-based experiments support NSUN5-dependent m 5 C enrichment of CENPF and CENPF protein abundance. (A) , Representative high-magnification immunohistochemical images of NSUN5, YBX1 and CENPF in normal skin, benign naevi and melanoma. All displayed fields were traced to their corresponding source whole-slide images. (B) , Baseline western blots of HEMa, A375, A2058 and 451Lu cells. (C–E) , Densitometric quantification of baseline NSUN5, YBX1 and CENPF protein abundance. Data are shown as the mean ± s.d. with individual values from three independent biological replicates. Statistical significance was assessed by one-way ANOVA followed by Dunnett’s multiple-comparison test, with HEMa serving as the common control. (F) , Western blot screening of three independent siRNAs targeting NSUN5 in A375 and A2058 cells. (G) , CENPF-targeted MeRIP–qPCR following NSUN5 knockdown. Displayed values are % input; n = 3; statistical testing used l
## Related Clinical Research

- [PD‑1–positive, Chymase‑negative mast cell phenotype links to metastatic melanoma and worse survival](https://medichelpline.com/clinical-feed/pubmed-42912494.md) (DOI: 10.1080/2162402X.2026.2752644)
- [Moderna–Merck melanoma vaccine data leaked before ESMO presentation](https://medichelpline.com/clinical-feed/stat-news-0-stat-highly-anticipated-moderna-merck-melanoma-cancer-vaccine-data-arrived-early.md)
- [Moderna–Merck personalized mRNA cancer vaccine cuts melanoma recurrence; judges allow drugmakers i](https://medichelpline.com/clinical-feed/stat-news-7-stat-pharmalittle-we-re-reading-about-a-cancer-vaccine-an-experimental.md)
- [Spatial transcriptomics links lipid and mitochondrial metabolism to T cell exclusion in esophageal](https://medichelpline.com/clinical-feed/british-journal-of-cancer-1-spatial-transcriptomic-sequencing-identifies-lipid-metabolism-and-mitochondrial.md)
- [Pan-cancer catalog maps mutation enrichment at transcription factor binding sites](https://medichelpline.com/clinical-feed/biorxiv-11-a-pan-cancer-catalog-of-mutagenesis-at-transcription-factor-binding-sites.md)

## Navigation
- [← Back to Oncology Feed](https://medichelpline.com/clinical-feed/oncology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.