Differential impact of tumor suppressor pathways on DNA damage response and therapy-induced transformation in a mouse primary cell model

Abstract

The RB and p53 tumor suppressors are mediators of DNA damage response, and compound inactivation of RB and p53 is a common occurrence in human cancers. Surprisingly, their cooperation in DNA damage signaling in relation to tumorigenesis and therapeutic response remains enigmatic. In the context of individuals with heritable retinoblastoma, there is a predilection for secondary tumor development, which has been associated with the use of radiation-therapy to treat the primary tumor. Furthermore, while germline mutations of the p53 gene are critical drivers for cancer predisposition syndromes, it is postulated that extrinsic stresses play a major role in promoting varying tumor spectrums and disease severities. In light of these studies, we examined the tumor suppressor functions of these proteins when challenged by exposure to therapeutic stress. To examine the cooperation of RB and p53 in tumorigenesis, and in response to therapy-induced DNA damage, a combination of genetic deletion and dominant negative strategies was employed. Results indicate that loss/inactivation of RB and p53 is not sufficient for cellular transformation. However, these proteins played distinct roles in response to therapy-induced DNA damage and subsequent tumorigenesis. Specifically, RB status was critical for cellular response to damage and senescence, irrespective of p53 function. Loss of RB resulted in a dramatic evolution of gene expression as a result of alterations in epigenetic programming. Critically, the observed changes in gene expression have been specifically associated with tumorigenesis, and RB-deficient, recurred cells displayed oncogenic characteristics, as well as increased resistance to subsequent challenge with discrete therapeutic agents. Taken together, these findings indicate that tumor suppressor functions of RB and p53 are particularly manifest when challenged by cellular stress. In the face of such challenge, RB is a critical suppressor of tumorigenesis beyond p53, and RB-deficiency could promote significant cellular evolution, ultimately contributing to a more aggressive disease.

Figure 1. Characterization of RB/p53 cooperation in cell growth and acute DNA damage response.

(A) Mouse adult fibroblasts isolated from Rb^loxP/loxP mice were infected with adenoviral GFP or GFP-Cre. Recombination was examined by PCR using primer sequences flanking Rb exon 19. MAFs were subsequently infected with retroviruses encoding LXSN or LXSN-p53DD. Cells were treated with 8 µM CDDP for 24 h, and total protein lysates were immunoblotted for p53 and p21. Lamin B served as a loading control. (B) Cells were cultured in normal growth media, pulsed with BrdU for 1 h prior to harvesting, stained with FITC-anti-BrdU and PI and analyzed by flow cytometry. Average percent BrdU incorporation is shown. **p<0.0054, ***p<0.0002 (C) Cells were cultured in normal growth media and viable cells were counted every 24 h. ***p<0.0001 (D) Cells were treated with 4 or 8 µM CDDP for 24 h, pulsed with BrdU for 1 h prior to harvesting, stained with FITC-anti-BrdU and PI, and analyzed by flow cytometry. Average percent BrdU incorporation relative to untreated controls is shown. ***p<0.0001.

Figure 2. RB status mediates cell growth vs. senescence in response to chronic therapy-induced damage, independent of p53 status.

(A) Cells were treated with 0.5 µM CDDP for 4 days. Cells were treated with 0.5 µM CDDP for 24 h, pulsed with BrdU for 1 h prior to harvesting, stained with FITC-anti-BrdU and PI, and analyzed by flow cytometry. Top, representative traces for both PI and BrdU staining are shown. Left, average percent BrdU incorporation relative to untreated controls is shown. ***p<0.0001 Right, total protein lysates were immunoblotted for the indicated proteins. Lamin B served as a loading control. (B) Left, cells were cultured in 0.5 µM CDDP, and viable cells were stained with crystal violet. Right, cells were treated with 0.5 µM CDDP for 2 days and stained for β-gal activity. Fold-increase in β-gal activity relative to untreated controls is shown. ***p<0.0001 (C) Four days post-CDDP exposure, RB-proficient cells were subjected to adenoviral infection to delete Rb. Left, four days post-infection, cells were pulsed with BrdU for 1 h prior to harvesting and processed as described in (A). Right, total protein lysates were prepared at each time point for immunoblot analysis. Lamin B served as a loading control.

Figure 3. RB loss promotes aggressive recovery from therapy-induced damage and significant alterations in gene expression.

(A) Cells were treated with 1 uM CDDP for 24 h, allowed to recover in fresh media, and stained with crystal violet two weeks post-treatment. Relative staining intensities from six independent experiments were quantified. ***p<0.0001 (B) Left, cells were cultured in normal growth media, and viable cells were counted over a course of 5 days. ***p<0.0001 Right, total protein lysates were immunoblotted for the indicated proteins. Lamin B served as a loading control. (C) Gene expression data was averaged for each cell line and normalized to the RB-proficient p53DD, untreated cell population (GFP Naïve). Left, changes in gene expression are displayed as a heat map. Right, those genes with specific involvement in transformation are highlighted in bar graphs. Gene expression values are displayed as fold-change in gene expression.

Figure 4. RB loss promotes alterations in epigenetic programming.

(A) Gene expression data for the RB-deficient p53DD, recurred cells demonstrated by microarray analysis were mapped along the mouse genome, indicating locations of up-regulated (Red) and down-regulated (Blue) genes. (B) Cells were cultured in the absence or presence of 5-aza-dC for 5 days, and RT-PCR was carried out for the indicated genes. (C) Cells were cultured in the absence or presence of 5-aza-dC for 5 days, and viable cells were counted. ***p<0.0001.

Figure 5. RB-deficiency and response to therapy-induced damage promotes tumorigenesis.

(A) Asynchronously growing cells were stained with phalloidin. Ras transformed cells are representative of transformed morphology. (B) Left, cells were cultured in non-coated Petri dishes, and viable cells were counted over a course of 6 days. ***p<0.0001 Right, total protein lysates were immunoblotted for the indicated proteins. Lamin B served as a loading control. (C) Top, Cells were injected into the flanks of nude mice, and tumor volumes were measured every 7 days post-palpable tumor formation. ***p = 0.0003 Bottom, representative H&E stained tissue sections of Cre p53DD, recurred xenografts are shown.

Figure 6. RB-deficiency promotes discrete therapeutic resistance after recovery from therapy-induced damage.

(A) Cells were treated with 8 uM CDDP for 24 h, pulsed with BrdU for 1 h prior to harvesting, and processed as described in Figure 1C. Average percent BrdU incorporation relative to untreated controls is shown. **p = 0.0024 (B) Cells were treated with 8 uM CDDP for 24 h, allowed to recover in fresh media for 4 days, and viable cells were counted at indicated time points. ***p<0.0001 (C) Cells were treated with 8 uM CDDP for 24 h and stained for CDDP-adduct formation. Red, CDDP adducts; Blue, DNA stained with DAPI. (D) Cells were treated with etoposide (Etop, 1 uM), camptothecin (Cpt, 5 uM), or mitomycin C (MMC, 2 uM) for 24 h, pulsed with BrdU 1 h prior to harvesting, and processed as described in Figure 1C. Average percent BrdU incorporation relative to untreated controls is shown. ***p<0.0001.