Cytoplasmic proliferating cell nuclear antigen connects glycolysis and cell survival in acute myeloid leukemia

Abstract

Cytosolic proliferating cell nuclear antigen (PCNA), a scaffolding protein involved in DNA replication, has been described as a key element in survival of mature neutrophil granulocytes, which are non-proliferating cells. Herein, we demonstrated an active export of PCNA involved in cell survival and chemotherapy resistance. Notably, daunorubicin-resistant HL-60 cells (HL-60R) have a prominent cytosolic PCNA localization due to increased nuclear export compared to daunorubicin-sensitive HL-60 cells (HL-60S). By interacting with nicotinamide phosphoribosyltransferase (NAMPT), a protein involved in NAD biosynthesis, PCNA coordinates glycolysis and survival, especially in HL-60R cells. These cells showed a dramatic increase in intracellular NAD+ concentration as well as glycolysis including increased expression and activity of hexokinase 1 and increased lactate production. Furthermore, this functional activity of cytoplasmic PCNA was also demonstrated in patients with acute myeloid leukemia (AML). Our data uncover a novel pathway of nuclear export of PCNA that drives cell survival by increasing metabolism flux.

Figure 1. Functional characterization and cytoplasmic PCNA localization in daunorubicin- sensitive (HL-60S) and – resistant (HL-60R) cells.

(a,b) HL-60 cells were treated with or without daunorubicin for 16 h. Data are mean ± SEM of six independent experiments performed in duplicates. (a) Viable cells were counted using flow cytometry. (b) Percentage of cells in the sub-G1 phase after propidium iodide labeling. (c,d) HL-60 cell lines were incubated with or without gliotoxin for 16 h to induce apoptosis. (c) Percentages of cells after DIOC6 labeling. Values are means ± SEM of four independent experiments. (d) Percentages of cells in sub-G1 phase showing DNA fragmentation after propidium iodide labeling. Values are means ± SEM of five independent experiments. (e) Effect of etoposide and cytarabine treatment in HL-60 cell lines for 16 h. Percentages of 7AAD + cells determined by flow cytometry. Values are means ± SEM of three independent experiments performed in duplicates. (f) Numbers of HL-60 cell lines counted after Trypan blue exclusion, where 1 × 10⁶ cells were cultivated in 5 ml of RPMI for 72 h in the presence of 10% FBS or 24 h in the presence of 0.5% FBS. Values are means ± SEM of six (at 72 h) and five (at 24 h) independent experiments. (g) Immunofluorescence analysis of PCNA localization by confocal microscopy (nuclei are visualized using Hoechst). This representative experiment was performed more than 10 times with identical results. Right panel represents quantification of nuclear fluorescence intensity by Image J. Data are means ± SEM of 15 values obtained after quantification of three independent experiments. (h,i) Western blot analysis of PCNA expression in HL-60 cell lines (h) in nuclear fractions using 20 μg protein/lane. Results were quantified by densitometry and normalized to lamin B1 loading control. The same experiments were performed with the whole lyzate using 10 μg protein/lane. Results were quantified by densitometry and normalized to β-actin loading control. For (a–f) multicomparison ANOVA test (indicated by*) was used to evaluate the difference between HL-60S and HL-60R and Student’s t test (indicated by #) was used to evaluate the effect of the drug on either HL-60S or HL-60R, *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2. The nuclear export of PCNA is increased in HL-60R cells and influenced their viability.

(a) Western blot analysis of CRM1 expression in whole cell lysates from HL-60S and HL-60R cells using anti-CRM1 rabbit polyclonal Ab (2 × 10⁶ cells/lane) and β-actin for the loading control. This experiment has been performed four times with identical results. Results were quantified by densitometry and normalized to β-actin loading control. The data are mean ± SEM of four independent experiments, *p < 0.05 (Student’s t test). (b) Effect of leptomycin B (LMB) on PCNA subcellular localization assessed by indirect immunofluorescence using anti-PCNA Ab5 rabbit polyclonal Ab. HL-60S and HL-60R cells were treated for 6 h with 20 ng/ml LMB. This experiment was performed four times with identical results. Right panel represents quantification of nuclear fluorescence intensity by Image J 1.48v soft©ware. Data are means ± SEM of at least 15 values obtained after quantification of four independent experiments. (c) Effect of LMB on the apoptosis of HL-60S and HL-60R. HL-60 cells cultured for 16 h at 37 °C in the absence or presence of LMB (5 or 10 or 20 ng/ml) or the pan caspase inhibitor Z-VAD(OMe)-FMK (50 μM) or the necroptosis inhibitor necrostatin-1 (Nec, 50 μM). The percentages of apoptotic cells were measured by annexin-V⁺ and 7-AAD⁻ using flow cytometry. The data are the mean ± SEM of six independent experiments, ***p < 0.001 (Student’s t test) for comparaison between HL-60R and HL-60S and ^###p < 0.001 (Student’s t test) for comparaison between untreated and Z-VAD(OMe)-FMK-treated HL-60R cells.

Figure 3. PCNA siRNA and p21-peptide-YIRS decreased the viability in HL-60R cells.

(a, left panel) Representative flow cytometry analysis plots of HL-60S and HL-60R treated with FITC-p21-peptide-YIRS. HL-60 cells exposed to increasing concentrations of the FITC-p21-peptide-YIRS (0, 10 (braun), 25 (red) and 50 (blue) μM) for 30 min. Experiment was performed four times with identical results. (a, right panel) Quantification of mean fluorescence intensity (MFI) of HL-60S and HL-60R treated with FITC-p21-peptide-YIRS. The data are mean ± SEM of four independent experiments, *p < 0.05 (ANOVA test). (b) Confocal microscopy analysis used to determine the localization of the FITC-p21-peptide-YIRS (1 h, 25 μM) in HL-60S and HL-60R. The FITC-p21-peptide-YIRS is visualized by green fluorescence and nucleus by Hoechst staining in blue. This experiment was performed three times with identical results. (c) Effect of p21-peptide-YIRS on DNA fragmentation. HL-60 cells were cultured for 16 h at 37 °C with or without 50 μM of p21-peptide-YIRS. Apoptosis was assessed by the percentage of cells in sub-G1 after propidium-iodide staining. The data are mean ± SEM of four independent experiments. **p < 0.01 (ANOVA test) (d) Representative Western blot analysis of HL-60R cells (50 μg/lane) showing decreased PCNA expression after PCNA siRNA treatment compared with control siRNA. Experiment was performed four times with identical results. (e) Apoptosis was induced by gliotoxin (1 μg/ml) for 4 h and mitochondria depolarization after DIOC6 labeling assessed. The data are mean ± SEM of four independent experiments, **p < 0.01 (ANOVA test).

Figure 4. Analysis of NAMPT expression in HL-60 cells.

(a) Co-localization experiments of NAMPT (green staining) and PCNA (red staining) in HL-60S and HL-60R using confocal immunofluorescence analysis. Hoechst was used for nuclear staining. This experiment has been performed three times with identical results. (b) Western-blot analysis of cytosolic NAMPT expression in HL-60 cell lines. Cytosols were obtained by nitrogen cavitation and 50 μg of protein loaded per lane. Results were quantified by densitometry of the top band and normalized to β-actin loading control. The data are mean ± SEM of five independent experiments, *p < 0.05 (Student’s t test). (c) Effect of the p21-peptide-YIRS on NAD concentrations in HL-60S and HL-60R. Measurement of NAD concentration after 6 h treatment with the p21-peptide-YIRS or with 3-bromopyruvate (3-BrP) used as a glycolysis inhibitor. The data are mean ± SEM of four independent experiments performed in duplicates. Multicomparison ANOVA test (indicated by*) was used to evaluate the difference between HL-60S and HL-60R and Student’s t test (indicated by #) was used to evaluate the effect of the drug on either HL-60S or HL-60R, *p < 0.05, **p < 0.01, ***p < 0.001. (d) Measurement of the interaction between PCNA and NAMPT using surface plasmon resonnance SPR analysis. A typical sensorgrams showing injection of NAMPT at the indicated concentrations over PCNA covalently immobilized on a CM5 sensor chip is shown (RU, response units). Arrows and stars indicate the injection start and end points for each concentration. A fit representing a typical experiment performed three times is shown, and was obtained by global fitting of the data using the 2-state model. The kinetic parameters of the interaction determined by analyzing response at varying NAMPT concentrations are listed in Supplementary Table 2. (e) P21-peptide binding to PCNA does not interfere with NAMPT–PCNA association. An overlay of the sensorgrams obtained with NAMPT, P21-peptide and a mixture of both on PCNA is showed (RU, response units). Binding assay was performed on a Biacore T200 system as described in Methods. This representative experiment has been performed twice giving identical results.

Figure 5. Metabolic characterization of HL-60S and HL-60R cells.

(a–c) Glycolytic flux measured as extracellular acidification rate (ECAR) of HL-60 cells at baseline and after adding D-glucose at two different concentrations, after oligomycin (OM), and after 2-DG with or without mutated p21-peptide-YIRS (a) or p21-peptide-YIRS (b). In (a,b), each panel represents one experiment in which the data plotted on the graph is the mean ± SEM (n = 5). Each experiment has been performed three times with identical results. (c) Quantification of the ECAR by measuring the total area under the curve. The data are mean ± SEM of three independent experiments performed in quintuplets. Multicomparison ANOVA test (indicated by*) was used to evaluate the difference between HL-60S and HL-60R and Student’s t test (indicated by #) was used to evaluate the effect of the p21 peptide on either HL-60S or HL-60R, **p < 0.01, ***p < 0.001. (d) Measurement of lactate in medium of HL-60S and HL-60R cells. The data are mean ± SEM of four independent experiments performed in duplicates, **p < 0.01 (Student’s t test). (e) Western blot analysis of hexokinase 1 in HL-60 cells (50 μg/lane). Quantification by densitometry of the top band is normalized to β-actin loading control. Data are mean ± SEM of three independent experiments, *p < 0.05 (Student’s t test). (f) Glucose-6-phosphate concentration in HL-60 cells. The data are mean ± SEM of three independent experiments, **p < 0.01 (Student’s t test). (g) Kinetic ECAR response of HL-60 cells after LMB treatment. The panel represents one experiment where the data plotted on the graph is the mean ± SEM of five values. (h) Results from (g) were quantified by total area under the curve. The data are mean ± SEM of five values. Multicomparison ANOVA test (indicated by*) was used to evaluate the difference between HL-60S and HL-60R and Student’s t test (indicated by #) was used to evaluate the effect of the LMB on either HL-60S or HL-60R, **p < 0.01, ***p < 0.001. (i) Confocal microscopy analysis of PCNA localization in HL-60 cells and Hoechst for nuclear labeling. Cells were treated for 6 h with 2-DG at 50 mM. Panel i shows a representative experiment of three all yielding the same results.

Figure 6. Analysis of PCNA subcellular localization in blast cells from patients with AML.

(a) Subcellular localization of PCNA in primary leukemic cells from patients with acute myeloid leukemia (AML). Immunofluorescence analysis by confocal microscopy of PCNA localization in blasts of AML patients with NK or CBF and in normal human CD34⁺ HPs using the rabbit polyclonal antibody Ab5 and Hoechst for nuclear labeling. Panel a shows a representative experiment of three, all yielding the same results. (b) Distribution of cytoplasmic or nuclear PCNA in blasts from AML patients according to the cytogenetic characterization. PCNA localization was determined in a cohort of 50 AML patients. The histogram depicts the percentages of AML patients showing either cytoplasmic or nuclear PCNA. (c) Immunofluorescence analysis by confocal microscopy of PCNA localization (performed as in a) in blasts from one AML patients after LMB treatment. Panel c shows a representative experiment of three yielding the same results. (d) Effect of the p21-peptide-YIRS on viability of blasts from 5 AML patients including 4 patients with cytosolic (open circle) and 1 patient with nuclear (dark triangle) PCNA. Blast cells were cultured for 16 h at 37 °C with or without the p21-peptide-YIRS (50 μM) and annexinV-7AAD-positive cells measured by flow cytometry.

Figure 7. Analysis of NAMPT expression in blast from AML patients.

(a) Colocalization between NAMPT (green) and PCNA (red) in blasts cells of AML patients after immunofluorescence labeling. This panel shows a representative experiment of three yielding the same results. (b) Western blot analysis of NAMPT expression in the cytosol of blasts from AML patient including three patients with cytoplasmic PCNA localization and two patients. (c) Results were quantified by densitometry and normalized to β-actin loading control for 15 AML patients with cytoplasmic PCNA versus 6 AML patients with nuclear PCNA. The data are mean ± SEM, *p < 0.05 (Mann-Whitney test).