NAA60 facilitates LRRC8A- and LRRC8D-mediated platinum drug uptake

Abstract

The platinum-based drugs cis- and carboplatin, which are crucial for treating cancers with DNA repair defects, like those caused by BRCA1/2 mutations, rely on the volume-regulated anion channel subunits LRRC8A and LRRC8D for about 50% of cellular drug uptake. Yet, the precise mechanisms of how LRRC8A and LRRC8D mediate this function are largely unknown. Here, we identify NAA60, an N-terminal acetyltransferase, which localizes to the Golgi apparatus to affect LRRC8A and LRRC8D function. Our data suggest that NAA60 acetylates the LRRC8A/D N-termini, and its loss decreases cis- and carboplatin uptake resulting in drug resistance of otherwise hypersensitive BRCA1;p53-deficient cells and tumors. Furthermore, we mimicked the absence of the neutralizing acetyl moiety that is observed after loss of NAA60 by introducing positively charged amino acids at the N-termini of LRRC8A/D, which indeed decreased cis- and carboplatin sensitivity. Our findings highlight the importance of N-terminal acetylation by NAA60 for effective platinum drug uptake, offering new insights into overcoming drug resistance.

Fig. 1. CRISPR/Cas9 screen identifies Naa60 as mediator of Pt drug sensitivity in BRCA1/2;p53-deficient mouse mammary tumor cells.

a CRISPR/Cas9 screen layout, where KB2P (K14cre;Brca2^F/F;Trp53^F/F) mouse mammary tumor cells were transduced (1) using the GeCKo V2 pool B library and subsequently selected using 0.2 µM or 0.25 µM of cisplatin over the course of 7 days (2). gDNA was harvested, sequenced and subjected to enrichment analysis using MAGeCK (3). The top hits were verified using CRISPR/Cas9 mediated gene knockout (4). Created in BioRender. Widmer (2025) https://BioRender.com/ynvrayk. b, c Volcano plots of the results from screens performed with either 0.2 µM or 0.25 µM cisplatin over the course of 7 days. Size of the points was scaled to −log₁₀ of FDR. d Representative big gene deletion knockout PCR control of wild-type (ntgB1 and ntgC1) and monoclonal Naa60 knockout KB1PM5 cell lines (Naa60_24A4, Naa60_12A6, and Naa60_13C3) used in the validation experiments. e, f RTqPCR and gel electrophoresis of the resulting product of control and Naa60 knockout lines. g, j Clonogenic survival assays of KB1PM5 wild-type and NAA60-deficient cell lines treated with cisplatin or carboplatin for 24 h with the indicated drug concentrations. Representative images of selected lines and concentrations are shown. h, k Quantification of clonogenic growth assays in the presence of cisplatin or carboplatin. Data represent mean ± SD of five independent replicates for cisplatin and three independent replicates for carboplatin and were fitted to a non-linear regression dose-response curve (log(inhibitor) vs. normalized response -Variable slope). p values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test for the log(IC50) values of the survival curves. Kaplan–Meyer overall survival curves of mice transplanted with NAA60-deficient (Naa60_24A4) or wild-type control (ntg) KB1PM5 organoids treated with either vehicle or 6 mg/kg of cisplatin for (i) and 50 mg/kg of carboplatin for (l). N = 5 mice per group. Statistical analysis was performed with the log-rank test (Mantel–Cox).

Fig. 2. Loss of Naa60 reduces cisplatin uptake and subsequent accumulation of DNA damage.

a CyTOF-based measurement of Pt drug uptake over time using 0.5 µM cisplatin in wild-type control, Lrrc8a-, and Naa60-knockout cell lines. The data represents the median ± SD of three independent replicates where approximately 10,000 cells per condition and cell lines were acquired (two-way ANOVA followed by Tukey’s multiple comparisons test). b Representative images of yH2AX immunofluorescence staining of Naa60 knockout and control cell lines following 2 µM cisplatin treatment. The scale bar represents 20 µm. c Quantification of yH2AX foci in the nucleus of Lrrc8a and Naa60 knockout and control cell lines in response to cisplatin treatment. Per cell line and condition, approximately 150 nuclei were quantified for each replicate. Lines at median and quartiles of three independent replicates are shown (ordinary one-way ANOVA followed by Tukey’s multiple comparisons test). d Western blot expression confirmation of the pOZ-HA-Naa60 or pOZ-Naa60-HA constructs in the Naa60 knockout cell line Naa60_12A6. e CyTOF-based Pt drug uptake measurement of wild-type, Naa60 knockout and the Naa60-HA C10 rescue cell lines after 24 h of treatment with 0.5 µM of cisplatin. The data represents the median Pt counts of three independent replicates where approximately 50,000 cells per condition and cell lines were acquired (two-way ANOVA followed by Tukey’s multiple comparisons test). f Quantification of yH2AX foci in the nucleus of Naa60 knockout and rescue cell lines in response to cisplatin treatment. Per cell line and condition, approximately 150 nuclei were quantified for each replicate. Lines at median and quartiles of three independent replicates are shown (ordinary one-way ANOVA followed by Tukey’s multiple comparisons test). g Representative images of yH2AX immunofluorescence imaging of empty vector or Naa60 cDNA transduced Naa60_12A6 cell line upon 0.5 µM cisplatin treatment over the course of 24 h. The scale bar equals 10 µm.

Fig. 3. LRRC8A and LRRC8D N-termini are a substrate for NAA60.

a Mass spectrometry analysis of the Nt peptide of LRRC8A after enrichment by immunoprecipitation using anti MYC-magnetic beads. Displayed are the peptide intensity values of three IP replicates for the detection of the acetylated peptide and two for the detection of the non-acetylated peptide. b The Nt peptide of LRRC8D was identified and analyzed by molecular weight focused gel band cutting and subsequent proteomics analysis. The results display the measurement of four gel bands for each group. The Nt peptide could however, not be identified in all analyzed samples. Data represent mean ± SD. c Proximity ligation assay (PLA) between the N-terminally HA tagged NAA60 and the C-terminally MYC tagged LRRC8A, shown together with a co-staining for the Golgi Marker 130 (GM130). The scale bars for both full size and cropped images equal 10 µm. The PLA was performed three independent times with similar results. Representative images of one of the replicates are shown.

Fig. 4. NAA60 and LRRC8A are epistatic.

a Western blot rescue control, where either HA-NAA60 or LRRC8A-MYC or both proteins were reintroduced into a double knockout line DbKO_A (Naa60^−/−; Lrrc8a^−/−). b Immunofluorescence imaging of the HA or MYC tags on the proteins, which were rescued into the DbKO_A line. The scale bar equals 10 µm. c–f Clonogenic survival assays of the different DbKO_A rescue cell lines treated with cisplatin, carboplatin, oxaliplatin, or blasticidin S for 24 h with the indicated drug concentrations. Representative images of selected concentrations are shown. g–j Quantification of clonogenic growth assays using the DbKO_A-based rescue cell lines in the presence of cisplatin, carboplatin, oxaliplatin, or blasticidin S. Data represent mean ± SD of four independent replicates and were fitted to a non-linear regression dose-response curve (log(inhibitor) vs. normalized response - Variable slope). p values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test for the log(IC50) values of the survival curves.

Fig. 5. LRRC8A is enriched at the plasma membrane of NAA60-deficient cells.

a Representative Western Blot of the LRRC8A and LRRC8D expression in wild-type and NAA60-deficient cell lines. b Quantification of LRRC8A and LRRC8D protein levels in wild-type and NAA60-deficient cell lines. The protein intensities were corrected according to the tubulin signal. Three independent replicates were quantified. The floating bar plot displays the line at mean. p values are calculated by ordinary one-way ANOVA followed by Tukey’s multiple comparisons test. c Representative Western Blot of the cycloheximide chase assay for LRRC8A, LRRC8D, and tubulin loading control. Results of ntgC1 and Naa60_12A6 are shown. The cells were treated with 0.1 mg/ml of cycloheximide for the indicated time periods (0 h, 4 h, 8 h, 20 h) followed by immunoblot analysis. As a positive control for the assay, a total ubiquitin antibody staining was included. d Quantification of the cycloheximide chase assay of LRRC8A or LRRC8D in wild-type and NAA60-deficient cell lines. LRRC8A and LRRC8D signals were quantified and corrected according to the tubulin signal intensity. The resulting value at the 0 h timepoint was set as 100%. The data was fit to a one-phase decay curve. Data represent mean ± SD of five independent replicates for LRRC8A and three independent replicates for LRRC8D. e Venn diagram displaying the overlaps of all significantly (adj. p < 0.01) upregulated proteins in the NAA60-deficient cell lines compared to the parental KB1PM5 cell line. f Differential plasma membrane enriched protein expression analysis of Naa60 knockout cell lines (Naa60_24A4, Naa60_12A6, and Naa60_13C3) in comparison to the parental KB1PM5 cell line. Data from three replicates of the plasma membrane protein enrichment and analysis are shown. The dotted line indicates the –log(pValue) significance cutoff of 2. g Immunofluorescence analysis of LRRC8A-GFP (8A-GFP) expressed in wild-type (ntgB1, ntgC1) and Naa60 knockout cell lines. Two replicates with three independent transfections of the LRRC8A-GFP construct were performed. Similar results were observed for all. Representative images of the co-localization with CellMask membrane stain are shown. The linear intensity profiles of each channel were generated using the Plot profile plugin of ImageJ. The white dot marks the start of the analysis along the selected linear ROI. The scale bar on the full-size images represents 20 µm and on the cropped images 5 µm.

Fig. 6. A positive charge at the N-termini of LRRC8A or LRRC8D reduces Pt drug uptake.

a, b Western blot rescue control of the cell lines, where either wild-type (wt) or N-terminally mutated LRRC8A (p.I2R) or LRRC8D (p.F2R) constructs were reintroduced into LRRC8A or LRRC8D-deficient cells. c Immunofluorescence imaging of the clonal rescue cell lines, expressing the C-terminally MYC tagged LRRC8 subunits, the scale bar represents 20 µm. Clonogenic survival assays of the different LRRC8A or LRRC8D rescue cell lines treated with cisplatin (d–f), carboplatin (g–i), and blasticidin S (j–l) for 24 h with the indicated drug concentrations. Representative images of selected lines and concentrations are shown. The data represent mean ± SD of three independent replicates and were fitted to a non-linear regression dose-response curve (log(inhibitor) vs. normalized response -Variable slope). p values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test for the log(IC50) values of the survival curves. m, n Representative yH2AX immunofluorescence images of the LRRC8A or LRRC8D-deficient cells expressing either the wt or mutated Lrrc8a or Lrrc8d rescue constructs treated with 2 µM cisplatin for 24 h. The scale bar represents 10 µm. o, p Quantification of yH2AX foci in the nucleus of LRRC8A or LRRC8D-deficient cells expressing either the wt or mutated Lrrc8a or Lrrc8d rescue constructs treated with 2 µM cisplatin for 24 h. Per cell line and condition, approximately 100 nuclei were quantified. Lines at median and quartiles of three independent replicates are shown (ordinary one-way ANOVA followed by Tukey’s multiple comparisons test).

Fig. 7. Illustration of how NAA60 facilitates LRRC8A- and LRRC8D-mediated platinum drug uptake.

Upon loss of NAA60, a positive charge at the N-terminus of the LRRC8 subunits remains non-neutralized and potentially leads to channel constriction, which might prevent passing of larger molecules such as cisplatin and carboplatin. This leads to decreased cellular Pt-drug accumulation and lower DNA damage levels, resulting in the resistance phenotype of NAA60-deficient cells. Created in BioRender. Widmer (2025) https://BioRender.com/9t5q7sr.