The p53 Transactivation Domain 1-Dependent Response to Acute DNA Damage in Endothelial Cells Protects against Radiation-Induced Cardiac Injury

Abstract

Thoracic radiation therapy can cause endothelial injury in the heart, leading to cardiac dysfunction and heart failure. Although it has been demonstrated that the tumor suppressor p53 functions in endothelial cells to prevent the development of radiation-induced myocardial injury, the key mechanism(s) by which p53 regulates the radiosensitivity of cardiac endothelial cells is not completely understood. Here, we utilized genetically engineered mice that express mutations in p53 transactivation domain 1 (TAD1) (p53^25,26) or mutations in p53 TAD1 and TAD2 (p53^25,26,53,54) specifically in endothelial cells to study the p53 transcriptional program that protects cardiac endothelial cells from ionizing radiation in vivo. p53^25,26,53,54 loses the ability to drive transactivation of p53 target genes after irradiation while p53^25,26 can induce transcription of a group of non-canonical p53 target genes, but not the majority of classic radiation-induced p53 targets critical for p53-mediated cell cycle arrest and apoptosis. After 12 Gy whole-heart irradiation, we found that both p53^25,26 and p53^25,26,53,54 sensitized mice to radiation-induced cardiac injury, in contrast to wild-type p53. Histopathological examination suggested that mutation of TAD1 contributes to myocardial necrosis after whole-heart irradiation, while mutation of both TAD1 and TAD2 abolishes the ability of p53 to prevent radiation-induced heart disease. Taken together, our results show that the transcriptional program downstream of p53 TAD1, which activates the acute DNA damage response after irradiation, is necessary to protect cardiac endothelial cells from radiation injury in vivo.

INTRODUCTION

Understanding the mechanisms regulating radiation-induced normal tissue injury can inform the design of improved approaches to treating cancer with radiation therapy. A better understanding of the mechanisms of normal tissue injury from radiation could also facilitate the development of therapeutic interventions for potential radiation accidents or disasters. The heart can be exposed to radiation therapy during treatment for breast cancers, lymphomas, lung cancers, or other tumors. Although the heart has been considered a relatively radioresistant organ, the cardiovascular system displays various pathologies after radiation exposure (1) and recent studies suggest that a mean radiation dose of 5 Gy to the heart can cause subsequent heart disease (2). A recently published analysis of 972 women who received radiation therapy for breast cancer and had no pre-existing cardiovascular disease showed that women who received left-sided radiation therapy had a higher risk of coronary artery disease (CAD) than those who received right-sided radiation therapy (HR: 2.5) (3, 4). The 27.5-year cumulative incidence of CAD was 10.5% in women treated with left-sided radiation therapy, which was sign ificantly higher incidence than the woman who received in right-sided radiation therapy (5.8%) (3). Moreover, cancer patients who receive thoracic radiotherapy have a 1.5 to 3 times higher risk of a fatal cardiac event than those who do not receive radiotherapy (1). Some of the atomic bomb survivors also developed an increased risk of cardiovascular disease years after the radiation exposure (5). Cardiovascular injury after radiation injury may be manifested as myocardial necrosis, vascular and valvular damage, and pericardial inflammation and fibrosis (1). Exposure of the heart to radiation can lead to systolic dysfunction, coronary artery disease, and heart failure (1, 5, 6), which may be a consequence of endothelial cell disruption and dysfunction (7). Endothelial cells are a critical cell type in the heart that responds to radiation by undergoing cell death, senescence, or alterations of cytokine production (7). These responses of endothelial cells to radiation lead to the disruption of microvasculature, hypoxia, myocardial necrosis, and tissue remodeling including myocardial fibrosis that can compromise cardiac contraction and cause arrhythmias and even death (8).

The DNA damage response (DDR) leads to many sequelae after exposure to ionizing radiation (9). One key protein that responds to DNA damage is the tumor suppressor p53 (10). Acute DNA damage activates p53 to induce apoptosis and/or cell cycle arrest in a cell type-dependent manner (10, 11). We have shown that p53 functions in cardiac endothelial cells to prevent the development of radiation-induced myocardial injury (6). Deletion of both alleles of p53 in endothelial cells by Cre recombinase expressed from either the Tie2 or vascular endothelial (VE)-cadherin promoter exacerbated cardiac dysfunction after 12 Gy whole-heart irradiation (6). Mice lacking p53 in endothelial cells (Tie2Cre; p53^FL/− mice) showed increased radiation-induced cell death in endothelial cells in vitro and in vivo (6). However, given that p53 controls a variety of signaling pathways including apoptosis, cell cycle arrest, DNA repair and metabolism, the key mechanism(s) by which p53 protects cardiac endothelial cells from radiation injury in vivo remains incompletely understood.

After exposure to ionizing radiation, p53 functions as a transcription factor to activate the expression of hundreds of genes (12, 13). p53-mediated target gene transcription requires two distinct domains in the N-terminus, namely transactivation domain 1 (TAD1) and 2 (TAD2) (11, 14). TAD1 and TAD2 work coordinately to interact with several proteins that facilitate the initiation of RNA polymerase II-mediated gene transcription or other transcription-related events (14). Specific amino acid residues within p53 TADs are critical for transactivation function. Mutation of the 25th and 26th residues of mouse p53 (22nd and 23rd residues in human p53) compromises TAD1 function on many p53 target genes while mutation of the 53rd and 54th residues (53rd and 54th residues in human p53) disrupts TAD2 function in the context of TAD1 mutation, leading to a transactivation defective mutant (14). Mutating p53 at these amino acids to hydrophobic residues alters transactivation, which provides a genetic tool to investigate the role of p53-induced gene transcription in tumor suppression and other functions in vivo. In p53^25,26/− cells, the mutated TAD1 domain is unable to drive the DNA damage-induced transcription of canonical p53 target genes such as PUMA and the cyclin-dependent kinase inhibitor p21, although the TAD2 domain in these cells remains functional to induce transcription of a small portion of non-canonical p53 target genes (15). In contrast, p53^{25,26,53,54/−} cells that harbor mutations in both TAD1 and TAD2 domains are deficient in activating all transcriptional targets of p53 and therefore resemble p53 null cells (15). These mutant alleles provide a sophisticated genetic tool to investigate the mechanism(s) by which p53 functions in endothelial cells to regulate radiation-induced heart disease by separating transactivation domain-dependent events into ones that require functional TAD1 and TAD2 with those that require TAD2 alone. Here, we performed experiments using mice with endothelial cells that lack p53 expression or express different TAD mutants to determine the role of p53 transcriptional targets induced by TAD1 (acute DNA damage response) vs. TAD1 and TAD2 (all p53 targets) in controlling the radiation response of cardiac endothelial cells in vivo.

MATERIALS AND METHODS

Mice

All procedures with mice were approved by the Institutional Animal Care and Use Committee (IACUC) of Duke University. All mouse strains used in this study have been described before, including VE-Cadherin-Cre (VE-Cre), p53^flox, p53^LSL−25,26, and p53^{LSL−25,26,53,54} mice (15–17). The VE-Cre mice were originally obtained from the Jackson Laboratory and then bred at Duke University. The p53^flox mice were originally provided by A. Berns (Netherlands Cancer Institute, Amsterdam, the Netherlands) and then bred at Duke University. Mice were at least 8 weeks old at the time of irradiation. Both sexes of mice were used. Comparisons were made using littermate controls to minimize genetic differences as these experiments were performed on mixed genetic backgrounds that included 129 Sv/J and C57BL/6. When mice developed severe radiation-induced cardiac injury, the mice developed hunched posture with limited movement and/or labored breathing.

Radiation Treatment

For whole-heart irradiations (6), a small-field image-guided irradiator, X-RAD 225Cx (Precision X-Ray Inc.), was used. Isoflurane was used for animal anesthesia. Mice were placed on the treatment stage in a prone position, and alignment was done by onboard fluoroscopic image guidance with 40 kVp (2.5 mA) X rays and a 2 mm aluminum filter. Radiation treatments were performed with parallel-opposed anterior and posterior fields using 225 kVp X rays at an average dose rate of 2.87 Gy per min as we previously described (6). A 15 mm circular collimator was inserted to create a circular radiation field at the isocenter where the heart was localized. Mice were either followed for the development of radiation-induced cardiac injury (n = 10–28 depending on the genotype) or sacrificed for tissue collections at day 55 after whole-heart irradiation (n=6–29 depending on the genotype).

Histological Analyses

Tissues were fixed with 10% neutralized formalin for paraffin-embedded tissue specimens. Paraffin-embedded tissue sections were stained with H&E or Masson’s trichrome staining. For immunohistochemistry, de-paraffinized slides underwent 3% hydrogen peroxide treatment, antigen retrieval with citrate-based solution, and blocking by serum. Slides were incubated with primary antibody against p53 protein (CM5; Leica Biosystems) and then with secondary antibody. The VECTASTAIN Elite ABC system (Vector Laboratories) was applied, and 3,3′-diaminobenzidine was used as chromogen. For p53/GS-IB₄ double staining, Alexa Fluor Plus highly cross-adsorbed secondary antibody (Invitrogen) was used to visualize p53 staining while the vasculature was detected by Isolectin GS-IB₄ Alexa Fluor 647 conjugate (Invitrogen).

For frozen tissue specimens, harvested tissues were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) at 4°C, and then placed in 30% sucrose in PBS overnight at 4°C. Processed tissues were then frozen in cryomolds of OCT compound (Sakura Finetek) and stored at −80°C. The heart of each mouse was cut in half before placing into a mold with the cut section facing the sectioning plane. Cryosections were used for EF5 hypoxia detection or were stained with H&E staining and then underwent blinded semiquantitative scoring.

For blinded semiquantitative scoring of myocardial necrosis, we provided a veterinary pathologist (JIE) a Masson’s trichrome stained and an H&E-stained slide from paraffin-embedded sections from the hearts or an H&E-stained slide from a frozen section from the hearts of other mice. In a manner blinded to genotype, the veterinary pathologist assigned a score from 0 to 4 to each sample according to the severity of myocardial injury with 0 being none and 4 being extensive.

EF5 Hypoxia Detection

EF5 solution (10 mM; Millipore Sigma) was prepared according to the product data sheet protocol. In brief, EF5 was dissolved in pure ethanol and was transferred to a 5% glucose solution for a final concentration of 10 mM EF5. The solution was administered through intraperitoneal injection at 26.5 μL/g body weight to mice. Hearts of the injected mice were harvested 3 h after the injection, and then fixed with 4% paraformaldehyde for frozen section preparations. The antibody Anti-EF5, clone ELK3–51 Cyanine 3 conjugate (Millipore Sigma) (75 μg/mL for staining) was used to detect EF5, and Isolectin GS-IB₄ Alexa Fluor 647 conjugate (Invitrogen) (10 μg/mL for staining) was used to detect vasculature.

Statistics

GraphPad Prism (GraphPad Software) was used for the statistics in this study. The Kaplan-Meier method was used to analyze the development of radiation-induced heart disease. Statistical comparison was performed using log-rank test. The Kruskal-Wallis test and the Dunn’s test of multiple comparisons were used to analyze statistical difference of the results of semiquantitative scoring.

RESULTS

To investigate the role of different p53 TADs in regulating the radiosensitivity of cardiac endothelial cells in vivo, VE-Cre was used to recombine the floxed allele of p53 (p53^FL) and remove the STOP cassette in the p53 TAD mutant alleles (Fig. 1A). We previously reported the distribution of VE-Cre activity within the epimyocardium and midmyocardium of the heart as measured by Cre-mediated recombination (6). The functional status of p53 in different transgenic mice used in this study is summarized in Fig. 1B.

FIG. 1.

p53 TAD1 in endothelial cells protects against radiation-induced cardiac injury in vivo. Panel A: A schematic showing the domain composition of p53 protein and the mutation sites of p53^25,26 and p53^25,26,53,54 (15, 23) (top) and a schematic showing VE-Cre-mediated recombination of the STOP cassette to express a p53 TAD mutant in the endothelial cell (bottom). Panel B: A table of the mouse genotypes used in this study. The p53 status of Cre-expressing and non Cre-expressing cells in each genotype is shown. Mice were treated with 12 Gy whole-heart irradiation (WHI). The rightmost column shows the median survival of each genotype. Panel C and D: Mice (n = 10–28) received 12 Gy wholeheart irradiation and were followed for the development of radiation-induced cardiac injury. The Kaplan-Meier curves of each genotype are shown in different colors. P values were calculated using log-rank test.

All mice that expressed various p53 mutant alleles specifically in the endothelial cells received 12 Gy wholeheart irradiation. Previously, we found that VE-Cre; p53^FL/− mice developed heart failure within 40 to 130 days after whole-heart irradiation (6). Here, we found similar results in which VE-Cre; p53^FL/FL mice that lack p53 in the endothelial cells were significantly more susceptible to the development of radiation-induced cardiac injury compared with VE-Cre; p53^FL/+ mice that retain one allele of p53 in the endothelial cells (Fig. 1C). These findings demonstrate that 12 Gy whole-heart irradiation reproducibly induces heart disease in mice lacking p53 in endothelial cells. Similar to VE-Cre; p53^FL/FL mice, VE-Cre; p53^{LSL−25,26,53,54/FL} mice succumbed to radiation-induced cardiac injury within 40 to 130 days after 12 Gy whole-heart irradiation (Fig. 1D), while VE-Cre; p53^{LSL−25,26,53,54/+} mice that retain one wild-type allele of p53 in the endothelial cells were not sensitized to radiation-induced cardiac injury (Fig. 1D). In addition, VE-Cre; p53^{LSL−25,26/FL} mice were also significantly more susceptible to the development of radiation-induced cardiac injury after 12 Gy whole-heart irradiation compared with VE-Cre; p53^{LSL−25,26/+} mice that retain one wild-type allele of p53 in the endothelial cells. Notably, the latency of radiation-induced cardiac injury was similar between VE-Cre; p53^{LSL−25,26/FL} mice and VE-Cre; p53^{LSL−25,26,53,54/FL} mice (Fig. 1D). Additionally, we found strong p53 protein staining in cardiac endothelial cells of VE-Cre; p53^{LSL−25,26/+} mice and VE-Cre; p53^{LSL−25,26,53,54/+} mice because TAD mutations lead to the accumulation of p53 protein in cells (Supplementary Fig. S1; https://doi.org/10.1667/RADE-22-00001.1.S1) at least in part because these mutants compromise the p53-MDM2 interaction (18). Nevertheless, the accumulation of p53 TAD mutant proteins did not impact the ability of the wild-type p53 allele to prevent the development of radiation-induced cardiac injury, and it might have potentially stimulated wild-type p53 activity (18). Collectively, our results from mice that have defects in p53 TAD1 (VE-Cre; p53^{LSL−25,26/FL}) or p53 TAD1/TAD2 (VE-Cre; p53^{LSL−25,26,53,54/FL}) in endothelial cells indicate that transcriptional targets of p53, especially the ones induced by p53 TAD1, are necessary to protect cardiac endothelial cells against radiation injury.

To investigate the causes leading to the development of cardiac injury in the p53 TAD mutants, we collected the hearts from different genotypes at day 55 after 12 Gy whole-heart irradiation. We found that both VE-Cre; p53^FL/FL and VE-Cre; p53^{LSL−25,26,53,54/FL} mice displayed extensive myocardial necrosis, while necrosis in the myocardium of VE-Cre; p53^{LSL−25,26/FL} mice was instead focal to multifocal (Fig. 2A). In contrast, VE-Cre; p53^FL/+, VE-Cre; p53^{LSL−25,26/+}, and VE-Cre; p53^{LSL−25,26,53,53/+} mice that retained a p53 wild-type allele in endothelial cells displayed no or minimal myocardial necrosis (Fig. 2A).

FIG. 2.

p53 TAD functions in endothelial cells regulate myocardial necrosis and fibrosis after whole-heart irradiation (WHI). Panel A: Hearts were harvested from mice of different genotypes at day 55 after 12 Gy whole-heart irradiation. Hematoxylin and eosin (H&E) staining and Masson’s trichrome staining were performed. Representative images of different genotypes are shown at two magnifications. The p53 status of Cre-expressing cells from each genotype is labeled at the left. Scale bars, 100 μm. Panel B: Semiquantitative scoring of myocardial injury of the hearts from mice of different genotypes (n = 6–29) at day 55 after 12 Gy whole-heart irradiation (bars represent medians with 95% confidence intervals). Blinded examination of the H&E stained or Masson’s trichrome stained slides was performed by a veterinary pathologist (JIE), and a score from 0 (no myocardial injury) to 4 (extensive myocardial injury) was assigned. Kruskal-Wallis test was performed, followed by Dunn’s multiple comparisons test. ****P < 0.0001.

We also examined fibrosis in the irradiated hearts by Masson’s trichrome staining. Similar to the H&E results, both VE-Cre; p53^FL/FL and VE-Cre; p53^{LSL−25,26,53,54/FL} showed prominent pathological changes in the myocardium, which included tissue fibrosis and disruption of the myocardium (Fig. 2A). The disrupted regions of the myocardium were replaced with fibrosis and they corresponded to the necrotic regions shown by H&E. Different levels of interstitial fibrosis were observed across the tissue sections. Similar to what we found with H&E staining, VE-Cre; p53^{LSL−25,26/FL} samples showed intermediate results with some of the samples showing minimal fibrosis and others showing more profound fibrosis (Fig. 2A). In addition, semiquantitative scoring of myocardial injury by a blinded examiner showed that both VE-Cre; p53^FL/FL and VE-Cre; p53^{LSL−25,26,53,54/FL} mice developed significantly higher levels of injury than VE-Cre; p53^FL/+ mice (Fig. 2B). The unbiased scoring also showed that the myocardial injury of VE-Cre; p53^{LSL−25,26/FL} samples was intermediate between VE-Cre; p53^FL/+ and VE-Cre; p53^FL/FL samples (Fig. 2B).

Taken together, loss of p53 transactivation (p53^25,26,53,54) led to multifocal myocardial necrosis, suggesting that the function of the TADs of p53 in the endothelial cells are essential to prevent radiation-induced myocardial necrosis. Aberrant p53 transactivation capability (p53^25,26) did not cause the same amount of profound myocardial necrosis, but still led to the development of cardiac injury that debilitated the animal, highlighting the critical importance of p53-mediated transactivation of TAD1 target genes. It is noteworthy that both p53^25,26 and p53^25,26,53,54 have been shown to compromise p53-mediated gene repression during hypoxia (19), so it is possible that derepression mediated by p53^25,26 and p53^25,26,53,54 might contribute to the phenotypes we observed in our model. However, whether p53 protein can directly repress target genes remains an area of controversy (20). There were almost no disrupted myocardial regions in the VE-Cre; p53^FL/+ samples 55 days after 12 Gy whole-heart irradiation. Similarly, there was none-to-minimal fibrosis in the VE-Cre; p53^FL/+, VE-Cre; p53^{LSL−25,26/+}, and the VE-Cre; p53^{LSL−25,26,53,53/+} samples (Fig. 2A). These results suggest that the presence of wild-type p53 in endothelial cells prevented the development of radiation-induced tissue fibrosis.

To understand if dysregulated endothelial response led to myocardial necrosis through tissue hypoxia, we utilized EF5 to label hypoxic regions as we previously reported (6). As expected, VE-Cre; p53^FL/FL samples showed multiple regions of EF5-labeled fluorescence signal while VE-Cre; p53^FL/+ samples showed none or minimal lesions (Fig. 3). The VE-Cre; p53^{LSL−25,26,53,54/FL} and the VE-Cre; p53^{LSL−25,26/FL} samples showed EF5-labeled hypoxic regions and/or disorganized vasculature detected by the glycoprotein GS-IB₄, though the VE-Cre; p53^{LSL−25,26/FL} samples were not as prominent. (Fig. 3). This suggests that in endothelial cells the non-canonical p53 targets from TAD2 may limit the extent of radiation-induced myocardial necrosis, possibly by preventing hypoxia. In contrast, there were no or fewer hypoxic regions and/or disorganized vasculature in the VE-Cre; p53^FL/+, VE-Cre; p53^{LSL−25,26/+}, and the VE-Cre; p53^{LSL−25,26,53,53/+} samples from mice that retain a wild-type p53 allele in endothelial cells (Fig. 3). The results from the histopathology indicate that extensive radiation-induced myocardial necrosis correlates with prominent hypoxia and/or disorganized vasculature. Taken together, this study provides evidence that inactivation of p53 TAD1 in endothelial cells is sufficient to sensitize mice to radiation-induced cardiac injury, while more severe damage to the myocardium occurs in mice where both p53 TAD1 and TAD2 are inactivated.

FIG. 3.

p53 TAD functions in endothelial cells contribute to limit hypoxia and/or a disorganized vasculature in the heart. Representative images of frozen sections of hearts stained with an antibody against EF5 (red) and counterstained with isolectin GS-IB₄ (white). The hearts were harvested from mice of different genotypes at day 55 after 12 Gy whole-heart irradiation (WHI) (n = 5–20). EF5 solution (10 mM) was administered through intraperitoneal injection at 26.5 μL/g body weight to the mice, and the hearts were harvested 3 h after the injection. Scale bars, 100 μm.

DISCUSSION

In the context of radiation-induced cardiac injury, we showed that both endothelial-specific loss of p53 (VE-Cre; p53^FL/FL mice with p53 null endothelial cells) and complete loss of p53 transactivation ability (p53^25,26,53,54) sensitized mice to radiation-induced injury to the heart. Compromised p53 TAD1 function (p53^25,26) also resulted in radiation-induced cardiac injury. This indicates that canonical p53 target genes that rely on functional p53 TAD1 in endothelial cells are critical to protect the heart after whole-heart irradiation. For example, p53^25,26 does not drive robust radiation-induced p21 expression after DNA damage (15), and p21 is known to protect irradiated hearts from injury (6). However, in mice lacking p53 TAD1 function in endothelial cells (VE-Cre; p53^{LSL−25,26/FL}), the extent of hypoxia and myocardial necrosis after whole-heart irradiation was not as severe as mice with endothelial cells lacking p53 (VE-Cre; p53^FL/FL mice). Therefore, non-canonical targets of p53 in endothelial cells may also contribute to limit the extent of radiation-induced pathological changes in the heart. Dissecting the transcriptomic changes in endothelial cells from irradiated hearts from VE-Cre; p53^{LSL−25,26/FL} mice and controls that retain wild-type p53 in the endothelial cells will be an interesting future direction to pursue. Moreover, comparing the transcriptional programs in endothelial cells from VE-Cre; p53^{LSL−25,26/FL} mice and VE-Cre; p53^FL/FL mice have the potential to identify candidate genes in addition to p21 that control DNA damage-induced late effects in general and radiation-induced cardiac injury in particular. While p53^25,26 fails to induce transcription of many canonical p53 target genes after DNA damage, it still retains the ability to activate a subset of non-canonical p53 target genes and express a basal level of p21 transcription (15). Therefore, it is possible that robust radiation-induced transcription of the non-canonical p53 targets and/or basal expression of p21 or other genes in endothelial cells function to prevent the most profound pathological changes in the heart after whole-heart irradiation. Regardless of the exact mechanism, our results show that this function of p53^25,26 fails to prevent radiation-induced cardiac injury. It is possible that the levels of necrosis in the VE-Cre; p53^{LSL−25,26/FL} mice were sufficient to induce heart failure. However, it is also possible that dysregulated p53 TAD functions might lead to cardiac injury through mechanisms other than necrosis. In future studies, it would be interesting to examine various forms of cell death and senescence in vitro in cardiac endothelial cells from VE-Cre mice with different p53 transactivation capacity, namely p53^FL/+, p53^FL/FL, p53^{LSL−25,26/+}, p53^{LSL−25,26/FL}, p53^{LSL−25,26,53,54/+}, and p53^{LSL−25,26,53,53/FL}. This may provide additional insights into the role of p53 TADs in the endothelial cells in response to radiation injury. Alternatively, the genotypes of cardiac endothelial cells could be generated using adenoviral infection to express Cre recombinase in the endothelial cells in vitro without the VE-Cre transgene.

A recently published article used Tie2Cre; p53^FL/FL mice to show that after whole-heart irradiation there were higher levels of nuclear L1 cell adhesion molecule (L1CAM) and γ-H2AX foci in p53-deficient endothelial cells in the heart (21). Interestingly, they showed that an anti-L1CAM antibody could reverse radiation-induced pathologies and improved animal survival compared to mice treated with an isotype control. Therefore, in future experiments it would also be interesting to look at nuclear L1CAM expression and γ-H2AX foci in the p53 TAD mutant mice.

One limitation of this study is selection bias. Only the hearts of mice that survived until the EF5 injection on day 55 after whole-heart irradiation could be subjected to histopathological examination. Some mice of the genotypes that were susceptible to radiation-induced cardiac injury (VE-Cre; p53^FL/FL, VE-Cre; p53^{LSL−25,26/FL}, and VE-Cre; p53^{LSL−25,26,53,54/FL}) died before or right at injection, preventing them from being included in the histological assays. This could lead to underrepresentation of the samples that developed more severe phenotypes. While this would shift the histological findings toward the milder side, this would affect all groups that lost mice prior to day 55. Although it is possible that studying the hearts from mice less than 55 days after whole-heart irradiation might reveal a more severe phenotype, the available data indicate a correlation between an intact p53 transcriptional program in endothelial cells and the control of radiation-induced cardiac injury.

The p53 protein functions by forming tetramers to transactivate target genes. Interestingly, p53^25,26,53,54 can form tetramers with wild-type p53 and the p53^{25,26,53,54/+} genotype results in upregulation of a subset of p53 target genes compared to p53^+/+ wild-type only, due to stabilization of wild-type p53 by p53^25,26,53,54 (18). Moreover, the p53^25,26/+ genotype also results in enhanced transactivation of some p53 target genes relative to p53 wild-type only (22). This indicates that with both a TAD mutant and wild-type p53 in the cell, the transactivation behavior of the p53 proteins should not be directly inferred from the genotype. Rather, abnormal p53 tetramers alter the target gene repertoire. In mouse embryos, many p53 target genes including p21 and Noxa are expressed at significantly higher levels in the p53^25,26/+ and p53^{25,26,53,54/+} genotypes, compared to wild-type p53, because of the stabilization of wild-type p53 (22). We found that a single wild-type p53 allele in endothelial cells was sufficient to protect mice from radiation-induced cardiac death, but it possessed a variable capacity to prevent hypoxia/necrosis depending on the presence of specific p53 TAD mutants within the same cell.

In conclusion, we identified a critical role for p53 transactivation functions in controlling radiation-induced late effects and that dysfunctional TAD1 alone in endothelial cells was sufficient to promote the development of cardiac injury after radiation exposure. This highlights the importance of p53 TAD1 function in endothelial cells to protect the heart against radiation-induced cardiac injury.

Supplementary Material

supplementary file 1

Supplementary Fig. S1. p53 proteins accumulate in endothelial cells of the heart from mice that express p53 TAD mutants. Panel A: Representative p53 protein immunohistochemistry (IHC) images of hearts harvested from mice of different genotypes at day 55 after 12 Gy whole-heart irradiation (n = 3–5). The images are shown at two magnifications. Scale bars, 100 μm. Panel B: Quantification of p53-positive endothelial cell in the hearts harvested at day 55 after 12 Gy whole-heart irradiation (n = 3–5). The area of GS-IB₄ vasculature was measured using Fiji (24), and the number of p53-positive cell per 10,000 μm² GS-IB₄ vasculature was calculated. Bars represent medians with 95% confidence intervals. *P < 0.05.