Irinotecan alleviates chemoresistance to anthracyclines through the inhibition of AARS1-mediated BLM lactylation and homologous recombination repair

Abstract

Chemoresistance remains the major barrier to cancer treatment. Metabolic and epigenetic reprogramming are involved in this process; however, the precise roles and mechanisms are largely unknown. Here, we report that lactate-induced lactylation promotes chemoresistance to anthracyclines by regulating homologous recombination (HR) repair. Using the global lactylome, we revealed the landscape of differentially lactylated sites and proteins in cancer cells isolated from resistant and nonresistant tumors. Specifically, BLM, a crucial helicase in the HR repair process, is highly lactylated at Lys24 by AARS1 in response to chemotherapy. Mechanistically, hyperlactylation of BLM improves its stability by inhibiting MIB1-mediated ubiquitination and increasing its interaction with DNA repair factors, promoting DNA end resection and HR repair. Delactylation of BLM via the Lys24 mutation impairs HR repair and increases anthracycline chemosensitivity. Irinotecan shows synergistic effects and safety for alleviating anthracycline resistance by targeting BLM lactylation and suppressing HR repair in pancancer PDX models. A single-arm, phase I study (identifier NCT06766266) initiated by us suggested that the combination of irinotecan liposomes plus EPI is a feasible and safe treatment strategy for patients with anthracycline-resistant bladder cancer who experience recurrence. These findings exemplify how glycolytic reprogramming regulates HR repair through promoting protein lactylation and highlight the promising therapeutic potential of irinotecan for reversing anthracycline chemoresistance by suppressing BLM lactylation.

Fig. 1

Lactate and lactylation levels were increased, and DNA repair was hyperactivated in EPI-resistant tumor tissues and cells. a Metabolic analysis of energy metabolites in primary tumors from patients with nonresistant NMIBC (NR-PT, n = 4), primary tumors from patients with resistant NMIBC (R-PT, n = 4), and recurrent tumors from patients with resistant NMIBC (R-RT, n = 4), highlighting the glycolytic metabolite lactate; NMIBC represents nonmuscle-invasive bladder cancer. b Lactate concentrations in NR-PT, R-PT, and R-RT bladder tumor samples from patients with nonresistant NMIBC (n = 145) and patients with resistant NMIBC (n = 64) compared via one-way ANOVA followed by Tukey’s test. c Lactate levels were compared via paired two-tailed Student’s t tests. d Representative pan-Kla immunohistochemical staining of NR-PT, R-PT, and R-RT bladder tumor tissues; scale bar: 100 μm. e, f The integral optical densities of pan-Kla staining in NR-PT (n = 145), R-PT (n = 64), and R-RT (n = 64) bladder tumor tissues were compared via one-way ANOVA followed by Tukey’s test and paired two-tailed Student’s t test. Kaplan‒Meier survival curves of the recurrence-free survival (RFS) of patients with NMIBC based on lactate concentrations (g) and pan-Kla levels (h) in primary tumor tissues; the median lactate concentration and the median integral optical density of pan-Kla were employed as the cutoff values; the statistical significance of RFS was determined via the Kaplan‒Meier method, and the P value was calculated via the log-rank test. i Receiver operating characteristic (ROC) analysis for recurrence prediction based on lactate concentrations and pan-Kla levels in primary bladder tumor tissues; red circles indicate cutoff values of 34.69 and 40. j The parental and EPI-resistant (E-resistant) UM-UC-3 cell lines either treated with 0.2 μM epirubicin (EPI) for 24 h or untreated; representative confocal microscopy images (left) and quantitative analyses (right) show the formation and disappearance of γH2AX (green) foci and the merging with DAPI staining (blue) of nuclei; scale bar: 10 μm. k Western blot analysis examining RAD51 accumulation in chromatin fractions and γH2AX expression in whole-cell extracts of parental and E-resistant UM-UC-3 cells before or 24 h after treatment with 0.2 μM EPI. l Heatmap showing energy metabolism metabolites in tumor cells from EPI-treated CDX_E-resistant and CDX_parental models (n = 3). m Western blot analysis of LDHA and pan-Kla in whole-cell extracts from CDX_E-resistant and CDX_parental models. n Western blot showing RAD51 chromatin accumulation and γH2AX expression in whole-cell extracts from CDX_E-resistant and CDX_parental models. ***p < 0.001 represents a significant difference between two groups

Fig. 2

Lactate-derived lactylation induces HR and promotes EPI resistance. a Pan-Kla and β-actin levels were determined by western blot in E-resistant UM-UC-3 cells before or 24 h after treatment with 0.2 μM EPI, 20 mM sodium lactate, and/or 20 mM sodium oxamate as indicated. b Viability of E-resistant UM-UC-3 cells before or 24 h after treatment with 0.2 μM EPI, 20 mM sodium lactate, and/or 20 mM sodium oxamate; one-way ANOVA followed by Tukey’s test. c Representative images of comet assays performed with E-resistant UM-UC-3 cells treated with the same experimental setup as described in b; the percentage of DNA tails was compared by one-way ANOVA followed by Tukey’s test; scale bar: 100 μm. d AsiSI-ER U2OS system and qPCR-based quantification of single-stranded DNA (ssDNA) generated by DNA end resection; ssDNA levels were normalized to those of the control (-4OHT) group; one-way ANOVA followed by Tukey’s test. e The same experimental setup as in b, followed by western blot analysis of RAD51 accumulation in chromatin fractions and γH2AX expression in whole-cell extracts. f Same experimental setup as in b. Representative images (left) and quantitative analyses (right) showing the formation of EPI-induced γH2AX (green) foci merged with DAPI (blue); scale bar: 10 μm. g Nude mice were transplanted subcutaneously with E-resistant UM-UC-3 cells and treated with EPI (2 mg/kg), sodium lactate (100 mg/kg), or sodium oxamate (500 mg/kg); representative tumor images are shown (n = 5). h Western blot analysis of pan-Kla and γH2AX (whole-cell extracts) and RAD51 (chromatin fractions) from CDX models. Except for panels g and h, the graphs represent three replicates per condition (n = 3). ***p < 0.001, **p < 0.01, *p < 0.05 represent significant differences between two groups; ns represents no significant difference

Fig. 3

Hyperlactylation of BLM at Lys24 induces EPI resistance. a Scatterplot showing the quantification of lactylated sites in relation to peptide intensities in tumor cells separated from EPI-treated CDX_E-resistant and CDX_parental models, highlighting lactylated proteins and peptides in the DNA repair pathway. KEGG pathway (b) and GO biological process (c) analyses of upregulated lactylated proteins in EPI-treated CDX_E-resistant models. d Radar diagram showing the top 30 lactylated proteins enriched in the tumor resistance pathway in tumor cells separated from EPI-treated CDX_E-resistant models; the larger pink circles represent higher log₂FC values; blue and purple numbers represent the quantification of protein lactylation in the EPI-treated CDX_E-resistant and CDX_parental models, respectively. e BLM-knockout E-resistant cells (E-resistant-BKO) were transfected with HA-tagged wild-type (WT), K24R, K31R, or K38R BLM; after transfection, HA-tagged BLM proteins were immunoprecipitated with an anti-HA antibody or an IgG control and analyzed via western blotting with anti-HA and anti-pan-Kla antibodies. f Illustration of BLM-K24 lactylation identified by MS. g Representative immunofluorescence images showing K24-lactylation-specific antibodies (BLM-K24la, red) merged with DAPI (blue) of nuclei in NR-PT, R-PT, and R-RT bladder tumor tissues; scale bar: 50 μm. BLM-K24 lactylation levels were determined by western blot with BLM-K24la antibody in NR-PT and R-PT bladder tumor samples (h) and in parental and E-resistant UM-UC-3 cells (i). j Western blot showing BLM-K24 lactylation levels in E-resistant cells treated with 0.2 μM EPI, 20 mM sodium lactate, and/or 20 mM sodium oxamate, as indicated. k HA-tagged BLM proteins were immunoprecipitated with anti-HA from E-resistant-BKO cells transfected with HA-tagged WT and K24R BLM plasmids; western blot analysis was performed to determine BLM-K24 lactylation levels. l Cell viability assays detecting the viability of E-resistant and E-resistant-BKO UM-UC-3 cells transfected with WT, K24R, K31R, or K38R BLM plasmids after treatment with 0.2 μM EPI; one-way ANOVA followed by Tukey’s test. m The same experimental setup as in k but displaying cell proliferation by colony formation assays; one-way ANOVA followed by Tukey’s test. The graphs represent three replicates per condition (n = 3). ***p < 0.001 represents a significant difference between two groups

Fig. 4

BLM-K24 lactylation facilitates HR and increases BLM protein stability. a E-resistant BKO UM-UC-3 cells were transfected with WT or K24R BLM plasmids; representative images (top) and quantitative analyses (bottom) show the formation of EPI-induced γH2AX (green) foci merged with DAPI (blue) before and 24 h after treatment with 0.2 μM EPI and 20 mM sodium lactate; scale bar: 10 μm. b Western blot analysis was performed to examine RAD51-chromatin associations and γH2AX expression in E-resistant BKO UM-UC-3 cells treated as described in a. c AsiSI-ER U2OS system and qPCR-based quantification analysis of ssDNA generated by DNA end resection in cells transfected with WT or K24R BLM plasmids before and 24 h after treatment with 20 mM sodium lactate; the ssDNA level was normalized to that of the control (-4OHT) group; one-way ANOVA followed by Tukey’s test. d E-resistant BKO UM-UC-3 cells transfected with HA-tagged WT or K24R BLM plasmids were treated with 20 mM sodium lactate or 20 mM sodium oxamate for 24 h; HA-tagged BLM proteins were immunoprecipitated with an anti-HA antibody or an IgG control and analyzed via western blotting with anti-HA and target antibodies as indicated. e, f Half-life detection and quantitative analysis of the WT and K24R BLM proteins in E-resistant-BKO UM-UC-3 cells treated with 40 μM cycloheximide and/or 20 mM sodium lactate for the indicated times. g E-resistant BKO cells stably expressing HA-BLM were cotransfected with Flag-ubiquitin (Ub) and treated with 20 mM sodium lactate or 20 mM sodium oxamate for 24 h, followed by immunoprecipitation with an anti-HA antibody or an IgG control; the blots were probed with the indicated antibodies. h, i The interaction between BLM and MIB1 proteins was tested via surface plasmon resonance (SPR) assay; the BIAcore diagram shows a greater affinity between K24R BLM and MIB1 than between WT BLM and MIB1, with fast dissociation kinetics for WT BLM and slow dissociation kinetics for K24R BLM. j E-resistant BKO UM-UC-3 cells cotransfected with HA-WT, HA-K24R BLM, His-MIB1, and/or Flag-Ub plasmids were treated with 20 mM sodium lactate or left untreated; HA-BLM proteins were immunoprecipitated with an anti-HA antibody or an IgG control and analyzed via western blotting with anti-HA and target antibodies as indicated. The graphs represent three replicates per condition (n = 3). ***p < 0.001, **p < 0.01, *p < 0.05 represent significant differences between two groups

Fig. 5

AARS1 catalyzes BLM lactylation and promotes HR and EPI resistance. a Western blot analyses were performed to examine BLM-K24 lactylation levels via a K24-lactylation-specific antibody in E-resistant UM-UC-3 cells treated with different siRNAs, as indicated. b AARS1 and β-actin levels were determined by western blot in parental and E-resistant UM-UC-3 cells before and 24 h after treatment with 0.2 μM EPI. c Endogenous BLM and AARS1 proteins were coimmunoprecipitated with anti-BLM, anti-AARS1, or IgG control and analyzed via western blotting with the indicated antibodies. d Parental and E-resistant cells with stable overexpression or knockdown of AARS1 were subjected to immunoprecipitation with an anti-BLM antibody or an IgG control; the blots were probed with the indicated antibodies. e E-resistant BKO UM-UC-3 cells with stable overexpression or knockdown of AARS1 were transfected with WT or K24R BLM plasmids and treated with 0.2 μM EPI; representative images show cell proliferation levels determined via colony formation assays and quantitative assays; one-way ANOVA followed by Tukey’s test. f Same experimental setup as in e but displaying the cell viability via cell viability assays. g Same experimental setup as that in e but for determining RAD51 accumulation in chromatin fractions and γH2AX expression in whole-cell extracts by western blot. h Representative images (top) and quantitative analyses (bottom) showing the formation of EPI-induced foci for γH2AX (green) merged with DAPI (blue) of nuclei; scale bar: 10 μm. i The AsiSI-ER U2OS system and qPCR-based quantification analysis were used for detecting ssDNA generated by DNA end resection in cells with stable overexpression or knockdown of AARS1; the ssDNA level was normalized on the basis of the results in the control (-4OHT) group; one-way ANOVA followed by Tukey’s test. j E-resistant BKO cells stably expressing AARS1 or AARS1 shRNA were cotransfected with His-Ub, HA-WT BLM, and/or HA-K24R BLM plasmids; the cells were subsequently lysed under denaturing conditions, and Ni-NTA beads were used to pull down His-tagged ubiquitin. The graphs represent three replicates per condition (n = 3). ***p < 0.001, **p < 0.01, *p < 0.05 represent significant differences between two groups

Fig. 6

Irinotecan inhibits BLM-K24 lactylation and reverses anthracycline resistance. a Heatmap illustrating the relative affinities of the top 30 small-molecule drugs to the BLM-WT and BLM-K24R mutants determined via molecular docking; the left-side values represent the average affinities of five-time random docking between small-molecule drugs and the K24 regions of BLM-WT and the BLM-K24R mutant; the relative value was normalized to the average affinity of the small-molecule drug with the highest affinity to BLM-WT; the right-side log2FC and p values represent the changes in and significance of the average affinities of the small-molecule drugs toward WT BLM and the K24R BLM mutant. b Molecular docking diagram illustrating the docking pose of irinotecan small-molecule drug binding to the K24 region of the BLM protein; the dotted yellow lines represent two hydrogen bonds between irinotecan and the K24 residue with distances of 3.2 Å and 3.6 Å. c BIAcore diagram showing that the WT BLM protein bound to the irinotecan small-molecule drug with high affinity and slow dissociation kinetics. d The K24R BLM protein does not bind to the irinotecan small-molecule drug. Western blot analyses of BLM-K24 lactylation levels in E-resistant-BKO (e) and THP-resistant-BKO (f) UM-UC-3 cells transfected with the HA-BLM plasmid before and 24 h after treatment with 4 μM irinotecan. g, h Cell viability assays detecting the viability of E-resistant-BKO and THP-resistant-BKO UM-UC-3 cells transfected with WT or K24R BLM plasmids after treatment with 0.2 μM EPI, 0.3 μM THP, or 4 μM irinotecan; one-way ANOVA followed by Tukey’s test. i, j Representative images and quantitative analyses showing cell proliferation levels as assessed by colony formation assays in E-resistant-BKO and T-resistant-BKO UM-UC-3 cells treated with the same experimental setup described in g and h; one-way ANOVA followed by Tukey’s test. k The AsiSI-ER U2OS system and qPCR-based quantification analysis were used to detect ssDNA generated by DNA end resection in cells transfected with WT or K24R BLM plasmids after treatment with or without 4 μM irinotecan. l Same experimental setup as in (g) but determining γH2AX in whole-cell extracts by western blot. m E-resistant cells cotransfected with HA-BLM, His-MIB1, or Flag-Ub plasmids were treated with 4 μM irinotecan or left untreated; HA-BLM proteins were immunoprecipitated with an anti-HA antibody or an IgG control and analyzed via western blotting with antibodies as indicated. n, o Nude mice were transplanted subcutaneously with E-resistant BKO UM-UC-3 cells stably expressing WT BLM or K24R BLM and treated with EPI (2 mg/kg) and/or irinotecan (25 mg/kg) as indicated; tumor weights were measured as shown in M, and tumor images were acquired as shown in N (n = 5). p, q γH2AX and BLM-K24 lactylation in whole-cell extracts and RAD51 accumulation in chromatin fractions were examined by western blot in CDX models. Except for (o–q), the other graphs represent three replicates per condition (n = 3). ***p < 0.001, **p < 0.01, *p < 0.05 represent significant differences between two groups; ns represents no significant difference

Fig. 7

Combination therapy with irinotecan reverses EPI resistance by suppressing BLM lactylation and HR in pancancer PDX models. NSG mice were transplanted subcutaneously with NMIBC patient-derived xenografts (R-RT tumor tissues) and treated with EPI (2 mg/kg) and/or irinotecan (25 mg/kg) as indicated (n = 5); tumor images were acquired as shown in (a); tumor weights were measured (b), and volumes were calculated (c); one-way ANOVA followed by Tukey’s test. d Representative immunofluorescence images (left) and quantitative analyses (right) showing the nuclear expression and localization of γH2AX (green) colocalized with DAPI (blue) in PDX models; scale bar: 40 μm. e γH2AX expression and RAD51 chromatin accumulation were examined by western blot in PDX models. f Lactylated proteins were immunoprecipitated with anti-pan-Kla or IgG control in PDX models and analyzed by western blotting with anti-BLM. g, h Concentrations of ALT and AST in the serum of mice from different groups as indicated. NSG mice were transplanted subcutaneously with EPI-resistant breast cancer patient-derived xenografts and treated with EPI (2 mg/kg) and/or irinotecan (25 mg/kg) as indicated (n = 5); tumor images were acquired as shown in (i); tumor weights were measured (j), and volumes were calculated (k); one-way ANOVA followed by Tukey’s test. l γH2AX expression in whole-cell extracts and RAD51 accumulation in chromatin fractions were examined by western blot in PDX models. m Lactylated proteins were immunoprecipitated with anti-pan-Kla or IgG control in PDX models and analyzed via western blotting with anti-BLM. ***p < 0.001, **p < 0.01, *p < 0.05 represent significant differences between two groups; ns represents no significant difference

Fig. 8

Safety, feasibility, and clinical response of combination treatment with irinotecan liposomes and EPI. a Diagram of the clinical trial design for the combination of irinotecan liposomes plus EPI for patients with anthracycline-resistant recurrent NMIBC. Participants will complete immediate intravesical instillation of EPI (50 mg) within 24 h after TURBT surgery. During the first month after surgery, participants will receive two cycles of combination treatment with irinotecan liposomes (administered intravenously once every two weeks for 1 month, with increasing doses of 37.6 mg/m² and 56.5 mg/m²) and EPI (intravesical instillation, once a week for 1 month, 50 mg). All participants subsequently continued standard intravesical EPI chemotherapy (50 mg, once a month) until 6 months postsurgery. CE-MRI: contrast-enhanced magnetic resonance imaging, NMIBC: nonmuscle-invasive bladder cancer, TURBT: transurethral resection of a bladder tumor. b Representative BLM-K24la immunohistochemical staining of bladder tumor tissues from participants with anthracycline-resistant recurrent NMIBC; scale bar: 100 μm. c Representative immunofluorescence images showing nuclear expression and localization of γH2AX (green) and RAD51 (red) colocalized with DAPI (blue) in recurrent bladder tumor tissues; scale bar: 100 μm. d, e Frequency of adverse events (AEs) after each cycle of combination treatment. During the first cycle, the first three participants (Patients A, B, and C) received a reduced dose of irinotecan liposomes (37.6 mg/m²) without experiencing dose-limiting toxicity and subsequently escalated to the full dose (56.5 mg/m²) in the second cycle because of good tolerance; in addition, 3 participants (Patients D, E, and F) subsequently received the full dose of irinotecan liposomes during both treatment cycles. f, g Quality of life evaluation of participants after each cycle of combination treatment. h Representative CE-MRI images showing the bladder status of patients A and D during the recurrence period and after combination therapy; orange arrows indicate recurrent bladder tumors. i Model of how glycolytic reprogramming-induced BLM lactylation facilitates anthracycline resistance by activating HR repair (Drawn using GNU Image Manipulation Program)