NBS1 lactylation is required for efficient DNA repair and chemotherapy resistance

Abstract

The Warburg effect is a hallmark of cancer that refers to the preference of cancer cells to metabolize glucose anaerobically rather than aerobically^1,2. This results in substantial accumulation of lacate, the end product of anaerobic glycolysis, in cancer cells³. However, how cancer metabolism affects chemotherapy response and DNA repair in general remains incompletely understood. Here we report that lactate-driven lactylation of NBS1 promotes homologous recombination (HR)-mediated DNA repair. Lactylation of NBS1 at lysine 388 (K388) is essential for MRE11-RAD50-NBS1 (MRN) complex formation and the accumulation of HR repair proteins at the sites of DNA double-strand breaks. Furthermore, we identify TIP60 as the NBS1 lysine lactyltransferase and the 'writer' of NBS1 K388 lactylation, and HDAC3 as the NBS1 de-lactylase. High levels of NBS1 K388 lactylation predict poor patient outcome of neoadjuvant chemotherapy, and lactate reduction using either genetic depletion of lactate dehydrogenase A (LDHA) or stiripentol, a lactate dehydrogenase A inhibitor used clinically for anti-epileptic treatment, inhibited NBS1 K388 lactylation, decreased DNA repair efficacy and overcame resistance to chemotherapy. In summary, our work identifies NBS1 lactylation as a critical mechanism for genome stability that contributes to chemotherapy resistance and identifies inhibition of lactate production as a promising therapeutic cancer strategy.

Fig. 1. Lactate enhances DNA repair and resistance to DNA-damaging therapy.

a, Schematic describing quantitative proteomics and non-targeted metabolomics analyses of 24 postoperative tumour specimens. b,c, Volcanic map of proteomic (b) or metabolic (c) differences between NAC-resistant and NAC-sensitive gastric cancer tissues. Red dots indicate enzymes or metabolites from the glycolytic pathway. Vertical and horizontal dashed lines indicate cutoffs of log₂ fold change (0.58 or −0.58) and P value (0.05), respectively. d, Changes in the proteins and metabolites of glycolytic pathway between NAC-resistant and NAC-sensitive gastric cancer tissues. Black dots represent metabolites that were not detected by mass spectrometry. P, phosphate; TCA, tricarboxylic acid. e,f, The growth of MGC803-P (e) and PDX-1 (f) tumours was assessed in NSG mice. g, Survival curve analysis of NSG mice transplanted with PDX-1 tumours. h, The growth of PDX-1 tumours was assessed in NSG mice. i, The growth of with wild-type or MCT1-knockdown (MCT1 is encoded by SLC16A1) MGC803-P tumours was assessed in NSG mice treated with or without lactate. All mice were treated with cisplatin. shMCT1, knockdown with short hairpin RNA targeting SLC16A1; shNC, non-targeting short hairpin RNA. e–i, n = 6. j, Dose–response curves for cisplatin in PDO1 treated with or without lactate (20 mM). k,l, AGS-P cells were treated with lactate (20 mM) for 24 h and then treated with IR (2 Gy). Cells were collected at the indicated time after IR treatment. k, Cells were lysed for immunoblotting analyses. Gel source data are presented in Supplementary Fig. 1. l, Comet assays (left) and analysis of tail moment. n = 60. Scale bars, 25 μm. m, Schematic representation of the HR reporter (left). HR repair efficiency was measured in lactate-treated and control HeLa reporter cells (right). BRCA2 overexpression was used as a positive control. j,m,n, n = 3. Data are mean ± s.d. P values by two-sided t-test (b–d,l,m) or two-way ANOVA (e,f,h,i). Log-rank test (g). In all box plots, the centre line indicates the median, box edges delineate third and first quartiles and whiskers extend to 1.5 times the interquartile range above and below the box. Source Data

Fig. 2. Lactate induces NBS1 K388 lactylation and TIP60 mediates NBS1 K388 lactylation.

a, IHC staining with pan-Kla antibody in postoperative tumour specimens from patients with NAC-sensitive or NAC-resistant gastric cancer. Scale bars, 50 μm. b, AGS-P and AGS-R cells were lysed for immunoblotting. c, AGS-P cells were subjected to laser microirradiation and stained with anti-pan-Kla and anti-γH2AX. Representative of n = 20 cells. Scale bars, 5 μm. d, Volcano plot showing global lysine lactylation changes in AGS-R cells compared with AGS-P cells. e, Cell lysates of AGS-P and AGS-R were immunoprecipitated with anti-NBS1 or control IgG, followed by immunoblotting. f, Identification and quantification of NBS1 K388 lactylation. LC–MS/MS analysis of modified VS(Kla)MEQK is shown. g, AGS-P and AGS-NBS1(K388R) cells were treated with lactate (20 mM) for 24 h, and whole-cell extracts were collected for immunoprecipitation with anti-NBS1 antibody, followed by immunoblotting. h, The cofactor pocket of TIP60 (PDB: 2OU2) bound to acetyl-CoA (left) and lactyl-CoA (right). TIP60 is shown in cartoon representation. The transfer group in acetyl-CoA or lactyl-CoA is indicated with red circles. i, Endogenous co-immunoprecipitation assays in AGS-P cells. j, Purified NBS1 was incubated with GST-TIP60, followed by GST pull-down assay and immunoblotting with anti-NBS1. k, AGS-P cells were treated with cisplatin (2.5 μM) for 6 h, and whole-cell extracts were collected for immunoprecipitation with the indicated antibodies, followed by immunoblot analysis. l,m, AGS-P cells were transfected with Flag–TIP60 (l) or small interfering RNA (siRNA) targeting TIP60 (siTIP60) (TIP60 is encoded by KAT5) or non-targeting siRNA (NC) and analysed by immunoblotting. 1 and 2 represent two independent small interfering RNAs. n, Left, in vitro lactylation assay showing lysine lactyltransferase activity of TIP60. Right, TIP60-mediated histone H4 acetylation served as positive control. Data are mean ± s.d. P values by two-sided Mann–Whiney test (a) or two-sided t-test (d). Gel source data are presented in Supplementary Fig. 1. Source Data

Fig. 3. NBS1 K388 lactylation facilitates DNA repair by promoting MRN complex formation.

a, Dose–response curves for IR in AGS-P and AGS-NBS1(K388R) cells treated with or without lactate (20 mM). n = 3 biologically independent samples. b, AGS-P cells and AGS-NBS1(K388R) cells were collected for comet assay at indicated times after IR treatment (10 Gy). Scale bars, 25 μm. n = 60 biologically independent cells. c, HeLa DR-GFP reporter cells were transfected with I-SceI-DsRed and control vector, Flag-tagged wild-type NBS1 (NBS1-WT) or NBS1(K388R). HR efficiency was assessed by counting GFP-positive cells. n = 3 biologically independent samples. d,e, AGS-P and AGS-NBS1(K388R) cells were treated with IR (2 Gy), and stained with anti-BRCA1 (d) or anti-RAD51 (e) 60 min after IR treatment. A cell containing ten or more foci was considered as a foci-positive cell. n = 60 cells were examined over three independent experiments. f, AGS-P and AGS-NBS1(K388R) cells were treated with lactate (20 mM) for 24 h, and whole-cell extracts were immunoprecipitated with anti-NBS1 antibody, followed by immunoblotting. g, Fractions of AGS-P or AGS-NBS1(K388R) cells were separated by sucrose gradient centrifugation. Indicated proteins were quantified using ImageJ. h, AGS-P cells were treated with cisplatin (2.5 μM) for 0.5, 1, 2 or 4 h. Chromatin and soluble fractions were isolated for immunoblotting. i, eGFP–NBS1-WT or eGFP–NBS1(K388R) were transfected into AGS-P cells. At 18 h post-transfection, cells were treated with or without lactate for 12 h. Cells were then laser micro-irradiated and monitored by live-cell microscopy. Accumulation of eGFP–NBS1-WT or eGFP–NBS1(K388R) on the DNA damage tracks was quantified. n = 20 biologically independent cells. Data are mean ± s.d. P values by two-sided t-test (b–e). Gel source data are presented in Supplementary Fig. 1. Source Data

Fig. 4. Stiripentol overcomes resistance to DNA-damaging treatment and LDHA expression and NBS1 K388 lactylation were increased in NAC-resistant tumours.

a, Indicated PDOs were treated as indicated for 72 h and analysed for cell viability by ATPlite assay. Synergy graphs were generated with Combenefit (Loewe model). b,c, The growth of PDX-R tumours was assessed in mice treated with: control (saline), cisplatin (2 mg kg⁻¹, once a week) (b) or IR (2 Gy per fraction, once daily for consecutive 4 days per week) (c) alone or combined with stirtpentol (150 mg kg⁻¹, once daily for consecutive 5 days per week). d,e, Survival curve analysis of NSG mice transplanted with chemo-resistant PDX tumours. b–e, n = 6. f, Top, representative IHC staining of LDHA and NBS1 K388 lactylation in NAC-sensitive or NAC-resistant gastric cancer tissues. Bottom, quantification of IHC staining of LDHA and NBS1 K388la in NAC-sensitive and NAC-resistant gastric cancer tissues. g, Correlation between LDHA expression and NBS1 K388 lactylation in gastric cancer tissues from 94 patients who received NAC. Note that some dots represent more than one specimen. h, High lactylation of K388 NBS1 and high expression of LDHA are correlated with the lowest overall survival rate. Data are mean ± s.d. f–h, n = 94. P value by two-way ANOVA (b,c), log-rank test (d,e,h), two-sided Mann–Whiney test (f) or two-sided Pearson correlation test (g). Source Data

Extended Data Fig. 1. Lactate was upregulated in resistant cancer cells.

a, Characteristics of 24 patients with gastric cancer undergoing neoadjuvant chemotherapy. Green, patient classed as sensitive; dark red, patient classed as resistant. TP53, MYC and LDHA genomic aberration status: no detectable alteration (light green), pathogenic mutation (purple), amplification (red) and deletion (black) in baseline biopsies. Differentiation status: poor (dark brown), moderate (light brown), moderate & poor (brown). Clinical TNM stage before NAC: T4 (blue), T3 (yellow). Lauren type: 1 = intestinal type; 2 = diffuse type; 3 = mix type. b, Dose-response curves for cisplatin in resistant cells and the corresponding parental cells. c, ECAR was measured between resistant cells and the corresponding parental cells. Analysis was performed using the Seahorse XF96e Extracellular Flux Analyzer. d, Statistical analysis of glycolytic activity between resistant cells and the corresponding parental cells. e, The level of lactate in AGS-P, AGS-R, MGC803-P, MGC803-R, A549-P, and A549-R cells was measured. Data are presented as mean ± SD. n = 3 biologically independent samples. P value was determined by t-test (two-sided) for c, d, e. Source Data

Extended Data Fig. 2. Lactate facilitates DNA-damaging treatment resistance in vivo and in vitro.

a, The growth of PDX2 tumors was assessed in NSG mice treated with: (1) Control (saline); (2) lactate (100 mL of 1 mM, three times a week); (3) cisplatin (2 mg/kg, once a week); (4) the combination of both agents at the aforementioned doses; (n =  6 mice per group). b, Survival curve analysis of NSG mice transplanted with MGC803-P tumors and treated with: (1) Control; (2) lactate; (3) cisplatin; (4) the combination of both agents; (n =  6 mice per group). c, Immunoblot analysis of MCT1 knockdown in MGC803-P cells. d, Relative lactate levels in MGC803 and PDX tumors. e, Dose-response curves for cisplatin in PDO2-5 treated with or without lactate (20 mM). f, Dose-response curves for cisplatin in AGS-P, MGC83-P, and A549-P cells treated with or without lactate (20 mM). g, Cell viability was measured by colony formation assays. h, Cell lysates were immunoblotted for caspase-3. i, Cell apoptosis was determined by flow cytometry. Data are presented as mean ± SD. n = 6 biologically independent samples for d, n = 3 biologically independent samples for e-f, h-i. P value was determined by two-way ANOVA for a, log rank test for b, t-test (two-sided) for d, h, i. For gel source data, see Supplementary Fig. 1. Source Data

Extended Data Fig. 3. Lactate affects DNA repair.

a, Dose-response curves for etoposide and adriamycin in AGS-P cells treated with or without lactate (20 mM). b, Cell lysates were immunoblotted for caspase-3. c, Dose-response curves for IR in AGS-P cells treated with or without lactate (20 mM). d, Cell apoptosis was determined by flow cytometry. e, Cell lysates were immunoblotted for caspases and caspase substrates. f, MGC803-P cells were treated with lactate (20 mM) for 24 h and then treated with IR (2 Gy). Cells were harvested at the indicated time after IR treatment. Cells were lysed for immunoblotting analyses. g, AGS-P cells were treated with or without lactate (20 mM) for 24 h. Cells were harvested at the indicated time points after IR (2 Gy) treatment for IF assay. Scale bars, 10 μm. n = 60 cells examined three independent experiments. h, Schematic representation of NHEJ reporter (above). NHEJ repair efficiency was measured (below). Ku70 overexpression as a positive control for promoting NHEJ repair. Data are presented as mean ± SD. n = 3 biologically independent samples for a, c, d, h. P value is determined by t-test (two-sided) for d, g, h. For gel source data, see Supplementary Fig. 1. Source Data

Extended Data Fig. 4. Lactate or cisplatin inducs NBS1 lactylation.

a, Schematic view of the experimental workflow for quantification of Kla in AGS-P and AGS-R cells. b, Summary of identified and quantified peptides and proteins in lactylation proteome. c, Protein-protein interaction network analysis of the DNA damage repair-related Kla proteins based on the STRING database. d, Dose-response curves for cisplatin in AGS-P and AGS-sh-NBS1 cells treated with or without lactate (20 mM) (Left). NBS1 knockdown in AGS cells was confirmed by western bolt assay (Right). e, A549-P, A549-R, MGC803-P, and MGC803-R were lysed with RIPA buffer. Immunoprecipitations of cell lysates with anti-NBS1 or anti-immunoglobulin G (IgG), followed by immunoblotting with antibodies against the indicated proteins. f, AGS-P cells were treated with lactate for the indicated concentrations, and WCEs were collected for IP with anti-NBS1 antibody, followed by immunoblot analysis. g, HEK293T and Hela cells were treated with cisplatin for 6 h, and WCEs were collected for IP with anti-NBS1 antibody, followed by immunoblotting. Data are presented as mean ± SD. n = 3 biologically independent samples for d. For gel source data, see Supplementary Fig. 1. Source Data

Extended Data Fig. 5. TIP60 and HDAC3 regulates NBS1 K388 lactylation.

a, AGS-P, AGS-NBS1-K388 cells were treated with cisplatin (2.5 μM) for 6 h, and WCEs were collected for IP with anti-NBS1 antibody, followed by immunoblotting. b, Sanger DNA sequencing traces of the NBS1 target site from the genome isolated from AGS-P cells (unedited cells) and AGS-NBS1-K388R cells (edited cells). c, ddH₂O containing different peptides was added onto the nitrocellulose membrane, followed by immunoblotting using anti-NBS1-K388la antibody. d, NBS1 K388 lactylation-specific antibody was validated for IHC staining. Representative images showing gastric cancer tissues stained with the antibody of NBS1-K388la incubated in the presence or absence of lactylated NBS1 peptides. Scale bar = 50 μm. e, AGS-P, A549-P, and MGC803-P cells were treated with lactate for the indicated concentrations, followed by immunoblot analysis. f, AGS-P, AGS-R, A549-P, A549-R, MGC803-P and MGC803-R cells were immunoblotted for analysis. g, Table summarizes the vital proteins identified by mass spectrometry analysis. We marked the protein of interest in red font. h, Exogenous IP assays showed that TIP60 interacts with NBS1. i, Flag-HDAC1, Flag-HDAC2 or Flag-HDAC3 was transfected into AGS-P cells. Cells were lysed for immunoblotting analyses. j, HDAC3 knockdown enhanced NBS1 K388 lactylation in MGC803-P cells. k, Co-IP assays showed that HDAC3 interacted with NBS1. Data are presented as mean ± SD. For gel source data, see Supplementary Fig. 1.

Extended Data Fig. 6. NBS1 K388 lactylation promotes DNA-damaging treatment resistance via HR repair.

a-d, Cytotoxic effects of cisplatin on AGS-P and AGS-NBS1-K388R cells were examined by cell viability (a), western blot (b), flow cytometry (c), and colony formation assay (d). e-f, Cytotoxic effects of IR on AGS-P and AGS-NBS1-K388R cells were examined by cell viability (e), western bolt (f). g-h, The protective effects of lactate on cisplatin resistance were examined in AGS-P and AGS-NBS1-K388R cells by western blot (g) and flow cytometry (h). i-j, The protective effects of lactate on IR resistance were examined in AGS-P and AGS-NBS1-K388R cells by cell viability (i) and flow cytometry (j). k-l, The protective effects of LDHA on cisplatin resistance were examined in AGS-P and AGS-NBS1-K388R cells by flow cytometry (k) and cell viability (l). m, n, Immunoblotting of γ-H2AX in both AGS-P and AGS-NBS1-K388R cells after cisplatin (2.5 μM) (m) or IR (10 Gy) (n) treatment for indicated time. o-p, Immunostaining with anti-BRCA1 (o) and anti-RAD51 (p) was performed at 60 min after IR (2 Gy). Scale bars, 15 μm. Data are presented as mean ± SD. n = 3 biologically independent samples for a, c, d, e, h, i, j, k, l. P value was determined by t-test (two-sided) for c, d, h, j, k. For gel source data, see Supplementary Fig. 1. Source Data

Extended Data Fig. 7. NBS1 K388 lactylation promotes MRN complex formation.

a, b, AGS-P and AGS-NBS1-K388R cells were treated with cisplatin for indicated time, and WCEs were collected for IP with anti-MRE11 (a) or anti-NBS1 antibody (b), followed by immunoblotting. c, AGS-P and AGS-NBS1-K388R cells were treated with IR (10 Gy), and WCEs were collected for IP with anti-NBS1 antibody. Then, purified NBS1 was supplemented with DSSO for crosslinking assay. Subsequently, the samples were analyzed by mass spectrometry. d, Graphical overview of the chemical cross-linking mass spectrometry results (CLMS). The green lines represent the interlinks within NBS1. The black lines represent the exterlinks between NBS1 and MRE11. Colored portions indicate structurally characterized parts of each protein. e, AGS-P and AGS-NBS1-K388R cells were treated with lactate, and WCEs were collected for IP with anti-MRE11, followed by immunoblotting. f, BLI assays of the lactylated NBS1 and the naïve NBS1 binding to MRE11. The analysis was performed by Pall Fortebio Octet Red96. g, Increasing expression of TIP60 enhanced the MRE11/RAD50 and NBS1 interaction. h, AGS-P cells were transfected with indicated si-RNA and TIP60-expressing plasmid for 48 h. WCEs were collected for IP with anti-NBS1 antibody, followed by immunoblotting. i, AGS-P cells were transfected with TIP60 for 12 h, then cells were treated with KU-55933 (ATM inhibitor) for 12 h. WCEs were collected for IP with anti-NBS1 antibody, followed by immunoblotting. For gel source data, see Supplementary Fig. 1.

Extended Data Fig. 8. NBS1 lactylation regulates the recruitment of MRN complex to DSB sites.

a, b, Immunostaining with anti-MRE11 and anti-RAD50 was performed after IR. A cell containing 10 or more foci was considered as a foci-positive cell. The percentage of anti-MRE11 and anti-RAD50 foci-positive cells was plotted. Scale bars, 10 μm. c, IF staining was performed with anti-NBS1 and anti-γ-H2AX at 30 min following IR as indicated. Scale bars, 10 μm. d, EGFP-MRE11 or EGFP-RAD50 was transfected into AGS-P or AGS-NBS1-K388R cells. At 24 h post-transfection, cells were laser micro-irradiated and monitored using a live-cell imaging microscope. Scale bars, 5 μm. e, EGFP-NBS1-WT or EGFP-NBS1-K388R mutant was transfected into AGS-P cells. At 18 h post-transfection, cells were treated with or without lactate for 12 h. Cells were then laser micro-irradiated and monitored using a live-cell imaging microscope. Scale bars, 5 μm. f, AGS-P, AGS-R, A549-P, A549-R, HCT116-P, HCT116-R, HGC27-P, HGC27-R, MGC803-P and MGC803-R cells were immunoblotted for analysis. For gel source data, see Supplementary Fig. 1. Data are presented as mean ± SD. n = 60 cells examined three independent experiments for a-c, n = 20 biologically independent cells for d, e. P value is determined by t-test (two-sided) for a-c. Source Data

Extended Data Fig. 9. Inhibition of LDHA reverses DNA-damaging treatment resistance.

a, LDHA-Knockout MGC803-R cell line was generated using the CRISPR/ Cas9 system. b, MGC803-R cells were treated with sodium oxamate (10 mM). c, MGC803-R cells were treated with stiripentol (0.5 mM). a-c, Cells were divided into two parts. One part was lysed for immunoblotting and the other was used to detect lactate production. d, AGS-P cells and AGS-NBS1-K388R cells were transfected with the indicated siRNA for 2 days and then exposed to IR (8 Gy). Cells were harvested for comet assay at the indicated time after IR treatment. Scale bars, 25 μm. n = 60. e, Characteristics of PDOs from five gastric cancer patients. Green, patient classed as sensitive; dark red, patient classed as resistant. TP53, MYC and LDHA genomic aberration status: no detectable alteration (light green), pathogenic mutation (purple), amplification (red) and deletion (black) in baseline biopsies. Gender status: Male (blue), female (yellow). Differentiation status: poor (dark brown), moderate (light brown). Age: Years. Histology: Adenocarcinoma. f, Dose-response curves for cisplatin in PDOs. PDOs were cultured with various concentrations of cisplatin for 72 h. g, The growth of MGC803-R tumors was assessed in mice treated with: (1) Control (saline); (2) cisplatin (2 mg/kg, once a week); (3) stirtpentol (150 mg/kg, once daily for consecutive 5 days per week); (4) the combination of both agents at the aforementioned doses. h, Mouse body weights from each group in MGC803-R and PDX-R were measured. Data are presented as mean ± SD. n = 3 biologically independent samples for a-c, f. n = 6 individual mice per group for g, h. P value is determined by t-test (two-sided) for a-d, two-way ANOVA for g. For gel source data, see Supplementary Fig. 1. Source Data

Extended Data Fig. 10. NBS1 K388 lactylation and LDHA predicts poor survival.

a, Box plot showing LDHA RNA levels across biologically independent tissues from various cancer types in GEPIA database. The box plot indicates the median (central line), the third and first quartiles (box limits) and 1.5 times the interquartile range above and below the box (whiskers). b, High levels of LDHA and NBS1 K388 lactylation were associated with poor overall survival of gastric cancer patients. c, Representative images of IHC staining of LDHA and NBS1 K388 lactylation in tumor specimens obtained before and after NAC. IHC scores of LDHA and NBS1 K388 lactylation are shown. n = 19 biologically independent individuals (Sensitive groups). n = 36 biologically independent individuals (Resistant groups). Scale bar = 50 μm. We shot it with a Nikon microscope. d, Cartoon illustrating that LDHA induces lactate, which in turn signals via TIP60 to increase NBS1 K388 lactylation. Lactylated NBS1 promotes MRN complex formation and the recruitment of MRN complex to DSB sites and thus favors DNA repair and cell survival. P value was determined by One-way ANOVA (two-sided) for a. * indicates P < 0.05, log rank test for b, or paired Wilcoxon test (two-sided) for c. Source Data