USP44 regulates irradiation-induced DNA double-strand break repair and suppresses tumorigenesis in nasopharyngeal carcinoma

Abstract

Radiotherapy is the primary treatment for patients with nasopharyngeal carcinoma (NPC), and approximately 20% of patients experience treatment failure due to tumour radioresistance. However, the exact regulatory mechanism remains poorly understood. Here, we show that the deubiquitinase USP44 is hypermethylated in NPC, which results in its downregulation. USP44 enhances the sensitivity of NPC cells to radiotherapy in vitro and in vivo. USP44 recruits and stabilizes the E3 ubiquitin ligase TRIM25 by removing its K48-linked polyubiquitin chains at Lys439, which further facilitates the degradation of Ku80 and inhibits its recruitment to DNA double-strand breaks (DSBs), thus enhancing DNA damage and inhibiting DNA repair via non-homologous end joining (NHEJ). Knockout of TRIM25 reverses the radiotherapy sensitization effect of USP44. Clinically, low expression of USP44 indicates a poor prognosis and facilitates tumour relapse in NPC patients. This study suggests the USP44-TRIM25-Ku80 axis provides potential therapeutic targets for NPC patients.

Fig. 1. Promoter hypermethylation of USP44 downregulates its expression in NPC.

a Heatmap clustering of seven hypermethylated CpG sites in the CpG islands of USP44 in normal nasopharyngeal epithelial tissues (n = 24) and NPC tissues (n = 24). Columns: individual samples; rows: CpG sites; blue: low methylation; red: high methylation. b Schematic illustration of the bisulfite pyrosequencing region in the USP44 promoter. Red region: input sequence; blue region: CpG islands; TSS: transcription start site; red text: CG sites used for bisulfite pyrosequencing; blue text: the most significantly altered CG site in the USP44 promoter. c, d Bisulfite pyrosequencing analysis of the USP44 promoter region (c) and statistical analysis of methylation levels (d) in normal (n = 8) and NPC (n = 8) tissues. e The methylation levels of the USP44 promoter region between NP69 and NPC cell lines (SUNE1, CNE1, CNE2, HNE1, and HONE1) were determined through bisulfite pyrosequencing analysis. f, g RT-PCR analysis of relative USP44 mRNA expression in the NP69 cell line and NPC cell lines (f) and in normal (n = 13) and NPC (n = 15) tissues (g). h, i Representative western blot analysis of USP44 protein expression in NP69 cells and NPC cell lines (h), together with normal and NPC tissues (i). j, k USP44 methylation levels measured by bisulfite pyrosequencing analysis (j) and relative USP44 mRNA levels measured by RT-PCR analysis (k) in NP69 cells and NPC cell lines with (DAC+) or without (DAC−) DAC treatment. Data in d and g are presented as the mean ± SEM, and those in e, f, j, and k are presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test; n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 2. USP44 enhances the radiosensitivity of NPC cells in vitro.

a In the GSEA of GSE12452 gene expression data, ionising radiation response-related pathways were enriched in the low USP44 expression group. b Clonogenic assays and survival fraction curves of SUNE1 and HONE1 cells stably transfected with USP44 or empty vector plasmids after exposure to the indicated IR dose. c, d Cell cycle distribution (c) and apoptosis rate (d) of SUNE1 and HONE1 cells transiently transfected with USP44 or the empty vector plasmids with or without exposure to 6Gy IR. The cell cycle distribution was detected at 8 h after IR and apoptosis rate was detected at 24 h after IR. Data were presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test; n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 3. USP44 ubiquitinates and degrades Ku80 by recruiting TRIM25.

a SDS-PAGE of HA-immunoprecipitated proteins separated from SUNE1 cells stably overexpressing HA-USP44. Red lines indicate the proteins of interest. b Co-IP with anti-HA or anti-FLAG antibody in SUNE1 cells revealed the exogenous association of USP44 and Ku80. c Immunofluorescence staining revealed the cellular location of exogenous HA-USP44 (green) and endogenous Ku80 (red) at 0.5 h after exposure to 6Gy IR. Scale bars, 10 μm. d USP44 inhibited Ku80 protein expression but not its mRNA expression in a dose-dependent manner. e The effect of CHX treatment and greyscale analysis of the results in 293T cells transfected with FLAG-Ku80 and HA-USP44 or the empty vector plasmids, as well as in sgNC or sgUSP44 SUNE1 cells. f The effect of MG132 and CQ treatment in 293T cells transfected with the indicated plasmids. g HEK293T cells transfected with FLAG-Ku80, HA-Ub and HA-USP44 or the empty plasmids were subjected to denature-IP and immunoblotted with the indicated antibodies. h Co-IP assay detecting the exogenous association of USP44 and TRIM25 and the endogenous association of USP44, TRIM25 and Ku80 in NPC cells. i TRIM25 inhibited Ku80 protein expression but not its mRNA expression in a dose-dependent manner. j The effect of MG132 and CQ treatment in 293T cells transfected with FLAG-Ku80 and FLAG-TRIM25 or the empty vector plasmids. k The effect of CHX treatment and greyscale analysis of the results in 293T cells transfected with FLAG-Ku80 and MYC-TRIM25 or the empty vector plasmids, as well as in sgNC or sgTRIM25 SUNE1 cells. l, m HEK293T cells transfected with the indicated plasmids or siRNAs were subjected to denature-IP and then immunoblotted with the indicated antibody. Data in d, e and i, k are presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test; n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 4. USP44 deubiquitinates and stabilises TRIM25 to promote Ku80 ubiquitination.

a USP44 promoted Ku80 protein expression but not its mRNA expression in a dose-dependent manner. b, c The effect of CHX (b), MG132 and CQ (c) treatment in 293T cells transfected with FLAG-TRIM25 and HA-USP44 or the empty vector plasmids, as well as in sgNC or sgUSP44 SUNE1 cells. d, e HEK293T cells transfected with HA-USP44 or the empty vector (d) and sgNC or sgUSP44 SUNE1 cells (e) co-transfected with FLAG-TRIM25 or MYC-TRIM25 and a vector encoding HA-WT-Ub or its mutants (HA-K48O-Ub or HA-K63O-Ub) were subjected to denature-IP and immunoblotted with the indicated antibodies. f HEK293T and NPC cells transfected with vector plasmid, HA-USP44 or HA-USP44 (C282A) were immunoblotted with the indicated antibodies. g HONE1 cells transfected with the vector plasmid, HA-USP44 or HA-USP44 (C282A) together with MYC-TRIM25 and HA-K48O-Ub were subjected to denature-IP and immunoblotted with the indicated antibodies. h Mass spectrometry analysis of TRIM25 ubiquitination sites. i HEK293T cells were transfected with the vector plasmid or HA-USP44, HA-Ub and Flag-TRIM25 WT or KR mutants, subjected to denature-IP with anti-Flag beads and then analysed by immunoblot with an anti-HA or anti-Flag antibody. j SUNE1 and HONE1 cells exposed to IR (6Gy) transfected with the indicated plasmids and siRNAs were fixed 0.5 h later and co-immunostained with the anti-Ku80 antibody. Scale bars, 10 μm. Data in a and b are presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test); n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 5. USP44-TRIM25 increases DSBs by impeding Ku80 recruitment.

The Vector + sgNC, USP44 + sgNC and USP44 + sgTRIM25 SUNE1 or HONE1 cells were stably constructed. a Representative comet images and quantitative analysis of tail moments for 6Gy-IR-induced DNA damage in the indicated SUNE1 or HONE1 cells, measured by the comet assay. Scale bars, 10 μm. b Representative images and quantitative analysis of the number of γH2AX foci in the indicated SUNE1 and HONE1 cells with or without 6Gy-IR exposure. Scale bars, 10 μm. c The indicated SUNE1 and HONE1 cells were transfected with GFP-Ku80 and then subjected to laser micro-IR and live-cell imaging. Scale bars: 10 μm. d The indicated SUNE1 cells were transfected with EJ5-GFP, infected with or without I-SceI adenovirus and analysed for GFP positivity by flow cytometry. Data in a, b and d are presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test; n = 20 (a), n = 10 (b), n = 3 (d) repeats from three independent experiments. Source data are provided as a Source Data file.

Fig. 6. Knockdown of TRIM25 reverses the radiosensitizing effect of USP44 in vitro.

SUNE1 and HONE1 cells were transiently co-transfected with HA-USP44 or the empty vector plasmids plus siTRIM25 or control siRNA. a Clonogenic assays and survival fraction curve analysis (a) and CCK-8 assay (b) of transfected SUNE1 and HONE1 cells after exposure to the indicated IR dose. The absorbance at 450 nm in c is presented as the mean ± SD of n = 4 independent experiments. c, d Cell cycle distribution (c) and apoptosis rate (d) of SUNE1 and HONE1 cells with or without exposure to 6Gy-IR. The cell cycle distribution was detected at 8 h after IR and apoptosis rate was detected at 24 h after IR. Data were presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test; n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 7. USP44 promotes the radiosensitivity of NPC cells in vivo.

The SUNE1 cells stably transfected with indicated plasmids were implanted subcutaneously into female BALB/c nude mice to construct xenograft growth models and exposed to 8Gy IR or not. a–c Macroscopic images (a), average volume (b) and average weight (c) of the excised tumours for each group (n = 10). d, e Representative images of immunohistochemical staining and IHC scores for USP44, TRIM25, Ku80 and caspase 3 expression in the excised tumours from each group (n = 5 for IR+USP44 group and n = 10 for the other three groups). Scale bars, 50 μm. f–h Macroscopic images (f), average volume (g) and average weight (h) of the excised tumours for each group (n = 6). Data in b, c, g, h are presented as the mean ± SEM, and those in e are presented as the mean ± SD; the P values were determined using the two-tailed Student’s t-test. Source data are provided as a Source Data file.

Fig. 8. Low expression of USP44 indicates a poor prognosis and is associated with tumour relapse in NPC patients.

a Representative images of immunohistochemical staining for USP44 protein expression is graded according to the intensity of staining in 376 NPC tissues. Scale bars, 50 μm. b Correlations of locoregional recurrence status with the level of USP44 expression detected by IHC. The P value was determined using the two-tailed $\chi$² test. c–e Kaplan–Meier analysis of locoregional recurrence-free survival (c), disease-free survival (d) and overall survival (e) according to the USP44 expression levels. The P values in c–e were determined using the log-rank test. f–h Forest plots of multivariate Cox regression analyses showing the significance of different prognostic variables in NPC locoregional recurrence-free survival (f), disease-free survival (g) and overall survival (h). i Proposed working model of USP44. USP44 recruits and stabilises TRIM25 by removing the K48-linked polyubiquitin chains of TRIM25, and TRIM25 degrades Ku80 by promoting its polyubiquitination and inhibits its recruitment to DSBs, which further inhibits the NHEJ pathway and enhances NPC radiosensitivity. In NPC, hypermethylation of the USP44 promoter leads to its downregulation at the mRNA and protein levels, which blocks the anticancer effect of the USP44-TRIM25-Ku80 axis. Source data are provided as a Source Data file.