Fig 2: Schematic diagram of this study. Enhancement of m5C methylation induced by NSUN2 upregulation in retinoblastoma cells leads to increased PFAS mRNA and protein expression and higher adenosine monophosphate (AMP) and guanosine monophosphate (GMP) content, which contribute to the aggressiveness related to retinoblastoma progression.

Fig 3: PFAS modulates m5C modification of NSUN2. (A) Multiomics analysis identified PFAS as a downstream target of NSUN2. (B) IGV tracks from m5C‐meRIP‐seq analysis showing m5C enrichment of PFAS. Biological duplicates were analysed. (C) m5C‐MeRIP‐qPCR assay of m5C status in PFAS in retinoblastoma cells and retinal pigment epithelium cells. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. *p < .05, **p < .01. (D) m5C‐MeRIP‐qPCR assay of m5C status in PFAS in NSUN2‐deficient retinoblastoma cells. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. (E) qPCR data showing PFAS expression in retinoblastoma cells (Y79 and WERI‐RB1) following NSUN2 knockdown. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. *p < .05, **p < .01. (F) Western blot and statistical analysis showing PFAS expression in retinoblastoma cells (Y79 and WERI‐RB1) following NSUN2 knockdown. The data are representative of experimental triplicates. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. **p < .01, ***p < .001. (G) Immunohistochemical staining images showing PFAS expression in the control group and NSUN2 knockdown group. Scale bars: left panel, 50 μm; right panel, 25 μm. (H) Correlation analysis of NSUN2 expression and PFAS expression in a cohort of retinoblastoma samples (n = 76). Significance was determined by Pearson correlation analysis (R = .505, p = 3.32e‐06).

Fig 4: ALYREF recognizes m5C‐methylated PFAS and promotes its RNA stability (A) Half‐life of PFAS in NSUN2‐deficient retinoblastoma cells treated with actinomycin (10 μg/μL) for 0–10 h. ***p < .001, ****p < .0001. (B) RIP‐qPCR assay of PFAS expression in retinoblastoma cells (Y79) induced by ALYREF and YBX1. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. ***p < .001, N.S. indicates no significance. (C) Single‐cell transcriptome profiling revealing the correlation between relative ALYREF protein expression and different biological processes in retinoblastoma. ***p < .001. (D) Correlation analysis of ALYREF expression and PFAS expression in a cohort of retinoblastoma samples (n = 76). Significance was determined by Pearson correlation analysis (R = .494, p = 5.80e‐06). (E) qPCR data showing PFAS expression in retinoblastoma cells (Y79 and WERI‐RB1) following ALYREF knockdown. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. ** p < .01, ***p < .001, ****p < .0001. (F) Western blot and statistical analysis showing PFAS expression in retinoblastoma cells (Y79 and WERI‐RB1) following ALYREF knockdown. The data are representative of experimental triplicates. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. ***p < .001, **** p < .0001. (G) Half‐life of PFAS in ALYREF‐deficient retinoblastoma cells treated with actinomycin (10 mg/mL) for 0–10 h. *p < .05, **p < .01, ***p < .001. (H) A CCK‐8 assay was employed to evaluate the proliferation of retinoblastoma cells (Y79 and WERI‐RB1) after ALYREF knockdown. The data are presented as the mean ± SD of experimental triplicates. Significance was determined by an unpaired two‐tailed Student's t test. **p < .01, ***p < .001. (I and J) A soft agar assay was employed to evaluate the proliferation of retinoblastoma cells (Y79 and WERI‐RB1) following ALYREF knockdown. Representative images from three experimental replicates are shown. The data are presented as the mean ± SD. Significance was determined by an unpaired two‐tailed Student's t test. **p < .01, ***p < .001. (K and L) Images of BALB/c nude mice and eyeballs containing xenografts derived from ALYREF‐deficient Y79 cells. Statistical analysis of the eyeball weight data. The data are presented as the mean ± SD. Significance was determined by an unpaired two‐tailed Student's t test. ***p < .001, ****p < .0001.