NDRG1 links p53 with proliferation-mediated centrosome homeostasis and genome stability

Abstract

The tumor protein 53 (TP53) tumor suppressor gene is the most frequently somatically altered gene in human cancers. Here we show expression of N-Myc down-regulated gene 1 (NDRG1) is induced by p53 during physiologic low proliferative states, and mediates centrosome homeostasis, thus maintaining genome stability. When placed in physiologic low-proliferating conditions, human TP53 null cells fail to increase expression of NDRG1 compared with isogenic wild-type controls and TP53 R248W knockin cells. Overexpression and RNA interference studies demonstrate that NDRG1 regulates centrosome number and amplification. Mechanistically, NDRG1 physically associates with γ-tubulin, a key component of the centrosome, with reduced association in p53 null cells. Strikingly, TP53 homozygous loss was mutually exclusive of NDRG1 overexpression in over 96% of human cancers, supporting the broad applicability of these results. Our study elucidates a mechanism of how TP53 loss leads to abnormal centrosome numbers and genomic instability mediated by NDRG1.

Keywords: NDRG1; centrosomes; genomic instability; p53; proliferation.

Fig. 1.

TP53 KO cell lines exhibit an increased rate of genomic instability and supernumerary centrosomes. MCF10A, TP53 KI (KI1 and KI2), and TP53 KO (KO1 and KO2) clones were grown in physiologic EGF conditions (0.2 ng/mL) and prepared for chromosomal instability (CIN) analysis as described in SI Appendix, SI Materials and Methods. (A) Relative copy number (CN) deviation from the modal population was determined using fluorescence in situ hybridization (FISH) using gene-specific probes for cMYC, EGFR, and BCR. Although both clones exhibit CIN, only TP53 KO2 showed a statistically significant increase (**P < 0.01). Results shown are the average values of multiple independent experiments and represent the average CN deviation of all three probes counting 200 cells for each probe. Error bars represent the SEM. (B) Representative FISH images for each cell line. EGFR probe (red) and BCR probe (green) with nuclear DAPI staining are shown. (C) MCF10A, TP53 KI (KI1 and KI2) and TP53 KO (KO1 and KO2) clones were grown in physiologic EGF conditions (0.2 ng/mL) and prepared for centrosome analysis as described in SI Appendix, SI Materials and Methods. Shown is the relative increase in centrosome amplification averaged after counting 200 cells for each cell line and performed in triplicate. Results are normalized to parental MCF10A and amplification was scored as positive if cells had greater than two centrosomes. Error bars represent the SEM, **P < 0.01. (D) Representative images assaying for centrosome amplification as determined by immunofluorescent staining of γ-tubulin (green) and nuclear DAPI staining (blue) under physiologic EGF (0.2 ng/mL) culture. (E) Representative images of cells in 20 ng/mL EGF culture conditions.

Fig. 2.

NDRG1 gene expression is regulated by TP53 and is increased in physiologic low-proliferation culture conditions. (A) Reverse phase protein array (RPPA) was carried out using MCF10A, TP53 KI (KI1 and KI2) and TP53 KO (KO1 and KO2) cell line lysates after culture in physiologic EGF conditions. Phosphorylated NDRG1 (pNDRG1) exhibited a five- to sixfold increase in MCF10A and TP53 KI cell lines. (B) RPPA expression levels were confirmed with Western blot analysis using antibodies against pNDRG1 and total NDRG1 in 0.2 ng/mL and 20 ng/mL EGF conditions. GAPDH is shown as a loading control. (C) Cell lysates were harvested from MCF10A and TP53 derivative cell lines after culture in 0.2 ng/mL and 20 ng/mL EGF conditions as described in SI Appendix, SI Materials and Methods and used for Western blot with an anti-p53 antibody. GAPDH is shown as a loading control. (D) Chromatin immunoprecipitation (ChIP) of MCF10A and TP53-derivative cell lines cultured in 0.2 ng/mL and 20 ng/mL EGF concentrations using an anti-p53 antibody and quantitative real-time PCR primers was carried out as described in SI Appendix, SI Materials and Methods. Results show relative p53 binding of the NDRG1 promoter in 0.2 ng/mL and 20 ng/mL EGF conditions, and results are the average of three independent experiments with samples run in triplicate. Error bars represent SEM. ***P < 0.001.

Fig. 3.

TP53 KO and NDRG1 overexpression are mutually exclusive in human cancers. Representative data for 8 of the 24 datasets examined from the cBioPortal database are shown. Studies were queried and analyzed as reported in SI Appendix, SI Materials and Methods. Shown are patient tumor samples with homozygous deletion of TP53 (closed blue bars) and corresponding expression of NDRG1, with open red bars indicating samples with more than twofold overexpression of NDRG1.

Fig. 4.

NDRG1 expression is inversely correlated to centrosome amplification in MCF10A and HCT116 human cell lines. (A) NDRG1 expression was knocked down in parental MCF10A cells using a shRNA vector (NDRG1 KD) along with vector controls (KD Con), and NDRG1 was overexpressed in TP53 KO cells (KO2) using a transgene cDNA expression vector (NDRG1 OE) as described in SI Appendix, SI Materials and Methods. Cell lysates from these cell lines, MCF10A and TP53 derivative cell lines, were then harvested and used for Western blot. GAPDH was used as a loading control. (B) MCF10A, knockdown control (KD control), NDRG1 knockdown (NDRG1 KD), TP53 KO (KO2), and TP53 KO (KO2) NDRG1 overexpressing (NDRG1 OE) cells were grown in 0.2 ng/mL EGF conditions and prepared for centrosome analysis as described in SI Appendix, SI Materials and Methods. Shown is the relative increase in centrosome amplification averaged after counting 200 cells for each cell line and performed in triplicate. Results are normalized to parental MCF10A and amplification was scored as positive if cells had greater than two centrosomes. Error bars represent the SEM, ***P < 0.001, *P < 0.05. (C) NDRG1 expression was knocked down in parental HCT116 cells using a shRNA vector (NDRG1 KD) along with vector controls (KD Con), and also overexpressed in HCT116 TP53 KO cells using a transgene cDNA expression vector (NDRG1 OE). Cell lysates were then harvested and used for Western blot, with GAPDH used as a loading control. (D) HCT116, knockdown control (KD control), NDRG1 knockdown (NDRG1 KD), TP53 KO and TP53 KO NDRG1 overexpressing (NDRG1 OE) cells were prepared for centrosome analysis. Shown is the relative increase in centrosome amplification averaged after counting 200 cells for each cell line and performed in triplicate. Results are normalized to parental HCT116 and amplification was scored as positive if cells had greater than two centrosomes. Error bars represent the SEM, ***P < 0.001, *P < 0.05. (E) Representative images assaying for centrosome amplification as determined by immunofluorescent staining of γ-tubulin (green) and nuclear DAPI staining (blue) under physiologic EGF culture conditions.

Fig. 5.

NDRG1 associates with γ-tubulin, a principal component of centrosomes. A proximity ligation assay (PLA) was performed using antibodies against NDRG1 and γ-tubulin as described in SI Appendix, SI Materials and Methods. Shown are in situ signals (red) of NDRG1 and γ-tubulin association as determined by their proximity to each other within 40 nm. Nuclear DAPI staining is also shown (blue).