Nuclear miR-451a activates KDM7A and leads to cetuximab resistance in head and neck squamous cell carcinoma

Cetuximab resistance has been a major challenge for head and neck squamous cell carcinoma (HNSCC) patients receiving targeted therapy. However, the mechanism that causes cetuximab resistance, especially microRNA (miRNA) regulation, remains unclear. Growing evidence suggests that miRNAs may act as “nuclear activating miRNAs” for targeting promoter regions or enhancers related to target genes. This study elucidates a novel mechanism underlying cetuximab resistance in HNSCC involving the nuclear activation of KDM7A transcription via miR-451a. Herein, small RNA sequencing, quantitative real-time polymerase chain reaction (qRT‒PCR) and fluorescence in situ hybridization (FISH) results provided compelling evidence of miR-451a nuclear enrichment in response to cetuximab treatment. Chromatin isolation via RNA purification, microarray analysis, and bioinformatic analysis revealed that miR-451a interacts with an enhancer region in KDM7A, activating its expression and further facilitating cetuximab resistance. It has also been demonstrated that the activation of KDM7A by nuclear miR-451a is induced by cetuximab treatment and is AGO2 dependent. Logistic regression analyses of 87 HNSCC samples indicated the significance of miR-451a and KDM7A in the development of cetuximab resistance. These discoveries support the potential of miR-451a and KDM7A as valuable biomarkers for cetuximab resistance and emphasize the function of nuclear-activating miRNAs. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00018-024-05324-x.

Wound healing assay and statistical analyses using CAL27 and CAL27_CTX cells (q and r) and HN30 and HN30_CTX cells (s and t) with and without cetuximab treatment.T tests (b, f, j, l, n, p, r and t; n=3) and two-way ANOVA (c, d, g and h; n=3) were performed.****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05; "ns" indicates no significance.These genes were chosen as the upregulated genes for screening the genes activated by miR-451a.(d) PCA of the four samples showed a closer relationship between the genetic features of CAL27_CTX and CAL27 miR-451a than between the genetic features of the other two samples.with antiagomiR-451a (CAL27_CTX antmiR-451a) group compared to that of the CAL27_CTX group (cetuximab treatment began at D7). (j) Tumour weights of the CAL27_CTX antmiR-451a group compared with those of the CAL27_CTX group.T tests were performed (a-e, g and j).Two-way ANOVA was performed to compare tumour growth at time points (i).****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05; "ns" indicates no significance.Table S2.The siRNAs and shRNAs used in this study.

Figure S2 .
Figure S2.Cetuximab-resistant HNSCC cell lines establishment (a) The survival rates of CAL27 cells and CAL27_CTX cells.(b) The survival rates of CAL27 and CAL27_CTX cells treated with 300 μg/mL cetuximab.(c-d) Cell growth curves of CAL27 cells (c) and CAL27_CTX cells (d) with and without 5-day treatment of cetuximab.(e) Comparison of survival rates between HN30 and HN30_CTX.(f) The survival rates of HN30 and HN30_CTX cells treated with 300 μg/mL cetuximab.(g) Cell growth curves of HN30 cells (g) and HN30_CTX cells (h) with and without 5-day treatment of cetuximab.(i-j) Colony formation assays (i) and statistical analysis (j) of CAL27 and CAL27_CTX cells with and without cetuximab treatment.(k-l) Transwell assays (k) and statistical analysis (l) of CAL27 and CAL27_CTX cells with and without cetuximab treatment.(m-n) Colony formation assays (m) and statistical analysis (n) using HN30 and HN30_CTX cells with and without cetuximab treatment.(o-p) Transwell assays and statistical analysis using HN30 and HN30_CTX cells with and without cetuximab treatment.(q-t)

Figure S4 .
Figure S4.Microarray data analysis of CAL27 miR-451a, CAL27 NC, CAL27 and CAL27_CTX.(a) Volcano plot showing the DEGs in the CAL27 miR-451a group compared to the CAL27 NC group.There were 5125 upregulated genes and 4713 downregulated genes.(b) The volcano plot shows the DEGs in CAL27_CTX compared to those in CAL27.There were 3558 upregulated genes and 3980 downregulated genes.(c) The intersection of the upregulated genes in the two comparisons.There were 1145 upregulated genes that were both upregulated in the CAL27 miR-451a and CAL27_CTX groups.

Figure S5 .
Figure S5.GO and KEGG analyses of the genes upregulated in the CAL27 miR-451a vs. CAL27 NC comparison and in the CAL27_CTX vs. CAL27 comparison.(a) GO analysis showing the genes upregulated in the CAL27 miR-451a group.(b) GO analysis showing the enrichment of genes upregulated in the CAL27_CTX strain.(c) KEGG analysis showing the enrichment of genes upregulated in the CAL27 miR-451a group.(d) KEGG analysis showing the genes whose expression was upregulated in the CAL27_CTX strain.

Figure S7 .
Figure S7.KDM7A enhanced HNSCC cells proliferation, migration and tumour growth upon treatment with cetuximab.(a) KDM7A knockdown via transient siRNA transfection.The western blot and qRT-PCR were performed to detect KDM7A expression at the protein and mRNA level.si-2 was eventually chosen for use.(b) KDM7A overexpression by the CMV vector validated by western blot.(cd) Statistical analysis of the colony formation assay and transwell assay results comparing KDM7Aoverexpressing CAL27 cells with CAL27 EV and KDM7A-silenced CAL27_CTX cells with CAL27_CTX siNC treated with cetuximab.(d) Statistical analysis of the Transwell assay results comparing KDM7A-overexpressing CAL27 cells and KDM7A-silenced CAL27_CTX cells treated with cetuximab.(e) Statistical analysis of the colony formation assay results comparing KDM7Aoverexpressing HN30 cells and KDM7A-silenced HN30_CTX cells treated with cetuximab.(f) Statistical analysis of the Transwell assay results comparing KDM7A-overexpressing HN30 cells and