Introduction
Radiation remains indispensable for the treatment of solid tumors, with increasing insight into the properties of high-energy electromagnetic waves continuing to reveal new possibilities for its application in oncological diseases [
1]. Although more than 50% of all solid tumors are treated with irradiation during cancer therapy [
2‐
4], the healthy tissue surrounding the tumor is often also harmed. A comparison of the effects induced by exposure of the survivors of the Hiroshima and Nagasaki nuclear attacks with patients who received ionizing irradiation for peptic ulcers or breast cancer has revealed an increased relative risk of developing cardiac diseases with increasing radiation doses, irrespective of the radiation source [
5]. The standard postoperative radiotherapy dose to the heart in women with breast cancer—the most common cancer in women worldwide [
6]—ranges between 10 and 40 Gy [
7]. It is now considered that acute and chronic cardiac pathologies such as coronary artery diseases, cardiomyopathy, and myocardial fibrosis that remain asymptomatic for nearly a decade [
8‐
10] might be initiated by partial heart irradiation, which initiates irreversible heart failure [
8,
9]. These findings are supported by randomized trials demonstrating a significantly increased risk for developing ischemic heart failure after postoperative radiotherapy in breast cancer patients [
11]. Moreover, it has been reported that cardiac mortality is approximately 60% and 20% in patients with left- and right-sided breast cancer 10 years after irradiation, respectively [
12], with the estimated mean dose to the heart after radiotherapy being 6.6 Gy and 2.9 Gy, respectively [
13]. The linear increase in the risk of developing a major coronary event starting 5 years after the radiotherapy up to 20 years has been calculated to be 7.4% per Gy [
13]. According to the cancer register, 22% of all deaths 10 years after thoracic irradiation are caused by heart diseases such as myocardial infarction and ischemic heart diseases [
14], and of these 894 patients, 535 suffered from left-sided breast cancer.
The lung is also considered an organ at risk after irradiation of the thorax in breast cancer patients. Irrespective of the location of the breast cancer, several lobes of the lung are inevitably exposed to radiation. To avoid life-threatening radiation-induced lung diseases, the volume of the lung receiving up to 20 Gy (V20) should always remain below 10% and the mean dose to the lung should be lower than 6 Gy [
15]. The close proximity of lung and heart cause reciprocal effects and multiorgan damage causes complications in both organs [
16]. Radiation of one organ leads to a lower tolerated dose in the other [
16], and co-irradiation of the heart and lung leads to drastically lower tolerated doses of both organs. Therefore, in this study, we investigated both primary heart and lung endothelial cells (ECs) derived from mice receiving complete local heart and partial lung irradiation.
Since recent studies have demonstrated that radiotherapy reduces the rate of relapses and breast cancer-mediated mortality rates after breast-conserving surgery by approximately 50%, radiotherapy is necessary and indispensable [
17]. Previous data have demonstrated that vascular inflammation is an acute adverse radiation effect which contributes to normal tissue toxicity in the heart and lung [
18]. The aim of this study was to identify and characterize long-lasting and late irradiation-induced effects in the microvasculature which might contribute to chronic heart diseases, including cardiac infarction. The CT-guided Small Animal Irradiation Platform (SARRP; Xstrahl, Camberley, UK) and a newly developed radiation plan enabled high-precision local irradiation of the heart, which spares large parts of the lung and thereby permits analysis of chronic, long-lasting irradiation effects (20 up to 50 weeks) in mice [
18]. Herein, the chronic pathomechanisms of radiation on the microvasculature are compared to the acute effects induced by myocardial infarction.
Materials and methods
Computed tomography-guided irradiation of the heart and lung
Ten-week-old female C57Bl/6 mice (Charles River, Sulzfeld, Germany) were anesthetized by isoflurane/oxygen inhalation and randomly allocated to different treatment groups. Irradiation was performed using the high-precision image-guided Small Animal Radiation Research Platform (SARRP, Xstrahl, Camberley, UK). Briefly, cone beam computed tomography (CT; 60 kV, 0.8 mA) was performed to visualize the thorax in each mouse. For quantitative CT analysis, the region of interest (ROI) was manually inserted into the CT images of mice 20, 30, 40, and 50 weeks after irradiation with 0 (sham), 8, and 16 Gy. The ROI included the area of the heart exposed to the prescribed doses and excluded most parts of the lung to enable long-term survival of the mice up to at least 50 weeks. The heart was irradiated with 8 or 16 Gy (220 kV, 13 mA) using a lateral 6 × 8 mm2 X‑ray beam. Control mice were sham irradiated with 0 Gy but received a CT scan. The SARRP software and MuriPlan treatment planning system (Xstrahl) were used to precisely target the heart position and irradiation dose. The left lung lobe received an irradiation of less than 6 Gy and the right lung lobe remained unirradiated. The co-irradiation of the lung volume in total was 18% and the lung volume exposed to the maximum irradiation dose was 7%.
All animals were housed in individually ventilated cages (IVC) under specific pathogen-free conditions. All experiments were approved by the Government of Upper Bavaria and were performed in accordance with institutional guidelines of the Klinikum rechts der Isar, TUM, Germany.
Immunohistochemical γH2AX staining
Irradiation delivery to the complete heart and to parts of the lung was determined with the DNA double-strand break marker γH2AX (Cell Signaling Technology, Danvers, MA, USA) on a Bond Rx staining machine (Leica Biosystems, Nussloch, Germany) using a Polymer Refine detection system without post-primary reagent. The heart and lung of one animal was removed 1 h after irradiation with 16 Gy, fixed in formalin overnight and embedded in paraffin. 2 µm sections were also stained with hematoxylin (Mayer’s hematoxylin) and eosin (0.5% aquaeous eosin γ‑solution).
Left anterior descending coronary artery ligation
Ten-week-old female C57Bl/6 mice were anesthetized by intraperitoneal injection of midazolam (5 mg/kg body weight), medetomidine (0.5 mg/kg body weight), and fentanyl (0.05 mg/kg body weight). Mice were intubated and ventilated with oxygen-enriched air using a small animal respirator (MiniVent, Hans Sachs Elektronik, Germany). Buprenorphine (0.05 mg/kg body weight) was injected subcutaneously for analgesia. The thorax was opened by a 0.8 cm cut at the third intercostal space between the third and fourth ribs. The pericardium was removed and the left anterior descending artery (LAD) was permanently ligated for 15 weeks using an 8/0 monofilament polypropylene suture (PROLENE®, Ethicon, Norderstedt, Germany). Narcotics were antagonized using atipamezole (2.5 mg/kg body weight) and flumazenil (0.5 mg/kg body weight) via subcutaneous injection. Primary heart endothelial cells (ECs) were isolated from the infracted left heart ventricle and the noninfarcted right ventricle of the same mouse. The isolated primary ECs from these areas were screened for the indicated cell surface markers 15 weeks after the artificially induced heart infarction.
Isolation of primary microvascular ECs from heart and lung
The isolation of primary ECs of mouse organs was performed as described previously [
19]. Briefly, following craniocervical dislocation, the heart and left and right lung lobes of mice were collected under aseptic conditions. The left and right atria were surgically removed from the heart to reduce contamination with macrovascular ECs. After rinsing in ice-cold phosphate-buffered saline (PBS; Gibco/Thermo Fisher Scientific, Darmstadt, Germany) the tissues were minced with a sterile scalpel blade into cubes with a size of 1 mm
3 and incubated in 10 ml of prewarmed (37 °C) digestion solution consisting of collagenase A (0.5 units/ml; Roche Penzberg, Germany) diluted in Hanks’ balanced salt solution (HBSS, Gibco)/10% v/v fetal bovine serum (FBS, PAA Laboratories GmbH, Freiburg i. Breisgau, Germany) for 45 min under gentle rotation (2 rpm). The suspension was forced 10 times through a sterile 18 G injection needle and filtered through a 70 µm cell strainer (BD Biosciences, Heidelberg, Germany). After two washing steps in 50 ml HBSS/10% v/v FBS solution (400 × g for 10 min), the final cell pellet was resuspended in 600 µl ice-cold isolation buffer (Invitrogen) with DSB‑X biotin-labeled (Molecular Probes) rat anti-mouse CD31 antibody (25 µl; 0.5 mg/ml; BD Biosciences, Heidelberg, Germany) for 10 min at 4 °C under gentle rotation (3 rpm). After another washing step in ice-cold isolation buffer, FlowComp™ beads coated with streptavidin (75 µl; Invitrogen/Thermo Fisher Scientific, Darmstadt, Germany) were added and incubated for another 15 min at 4 °C under gentle rotation. CD31-positive cells were isolated using a magnetic bead separation method (Invitrogen™ DynaMag™ magnet), after which CD31-positive cells immobilized to the DSB-X–streptavidin bead complex were released using biotin–streptavidin competition by incubating beads in 1 ml FlowComp™ release buffer (Invitrogen) for 2 min at 21 °C and pipetting 10 times. Isolated CD31-positive cells were used for flow cytometric analysis.
Immunophenotypic characterization of primary ECs
Freshly isolated primary microvascular ECs were phenotypically characterized by flow cytometry on a BD FACSCalibur™ instrument (BD Biosciences, Heidelberg, Germany) using the following fluorescein (FITC)-, phycoerythrin (PE)-, or allophycocyanin (APC)- conjugated antibodies: integrin β3 (CD61, BD Biosciences, clone 2C9.G2), endoglin (CD105, eBioscience, clone MJ7/18), VE-cadherin (CD144, BD Bioscience, clone 11D4.1), mucosialin (CD34, eBioscience, clone RAM34), FAT (CD36, Invitrogen, 12-0362-82), PECAM‑1 (CD31, BD Biosciences, clone MEC 13.3), HCAM (CD44, Santa Cruz, sc-9960), ICAM‑1 (CD54, BD Biosciences, clone 3E2), ICAM‑2 (CD102, BD Biosciences, clone 3C4), VCAM‑1 (CD106, Santa Cruz, sc-18864), and common leukocyte marker (CD45, BD Biosciences, clone 30-F11) as a negative control. Appropriately labeled isotype-matched immunoglobulins were used as nonspecific binding controls. Briefly, 0.1 × 106 viable cells were incubated with the indicated antibodies for 30 min at 4 °C in the dark. Following a washing step in PBS/FBS (10% v/v), cells were analyzed on a BD FACSCalibur™ instrument. Dead cells were excluded from the analysis by propidium iodide (PI) co-staining and a negative gating strategy.
Statistical analysis
Data were analyzed using BD CellQuest™ Pro software (BD Biosciences) and the statistical significance of differences between experimental groups determined using Tukey’s test. Differences were considered significant at *
p < 0.05 (5%), **
p < 0.01 (1%), and ***
p < 0.001 (0.1%). Data are presented as means of the number (
n) of each experiment [
19].
Discussion
Although irradiation is an indispensable treatment option for many solid tumors, chronic toxicity to normal tissues can limit the therapeutic success. High-dose irradiation of breast, lung, or esophageal cancer can cause chronic heart failure approximately a decade after radiation therapy [
22]. In order to better understand the long-term consequences of irradiation in this context, we have studied the chronic effects (20 to 50 weeks) of local heart irradiation and partial irradiation of the left lung lobe on the biology of primary ECs isolated from these organs in mice. The parameters that were examined—proliferation, stemness, lipid metabolism, adhesion, and inflammation—are known to increase the risk for developing chronic cardiac diseases such as myocardial infarction after irradiation of the left thorax in breast cancer patients [
5]. Although previous studies have described acute inflammatory responses of normal tissues of mice after irradiation [
23], chronic effects were not studied due to the limited overall survival of mice after complete chest irradiation [
20]. Our laboratory has developed a radiation plan for a high-precision local heart irradiation that spares large parts of the lung tissue and allows much longer survival of the mice [
18]. In a previous study, inflammatory and fibrotic changes in the irradiated heart and lung tissue 50 weeks after local heart irradiation of mice was demonstrated by the existence of intra-alveolar foam cells, heart hyaline deposition, lung interstitial fibrosis, lung perivascular infiltrations, and an increased lung density as determined by cone beam CT [
20]. Herein, we have demonstrated for the first time a chronic and long-lasting upregulated expression of major inflammatory markers such as CD44 (HCAM), CD54 (ICAM-1), and CD106 (VCAM-1) on primary heart ECs up to 50 weeks after local heart irradiation with clinically relevant doses of 8 and 16 Gy. These data indicate an ongoing chronic inflammatory process in the heart microvasculature as a late side effect after heart irradiation which might attract leukocytes to the endothelium and thereby promote further damage of the vessels [
24]. Chronic microvascular dysfunction induce complications in vascular tone and blood hemostasis, and promote chronic inflammatory processes [
25] which, in turn, induce radiation-induced cardiovascular injuries [
9]. Ischemic heart disease is the most common cause of death after radiotherapy [
26]. A long-term and permanent upregulation of inflammatory markers generates a chronic inflammatory micromilieu which accelerates atherosclerotic lesions in coronary vessels, pericardial fibrosis, myocardial fibrosis, and vessel calcination [
8].
Local irradiation of the heart with 8 Gy also caused permanent upregulation of CD36 (FAT), an integral membrane glycoprotein regulating lipid metabolism. CD36 is responsible for the uptake of cholesterol by macrophages and supports their transformation into lipid-loaded foam cells [
27] which are involved in the formation of atherosclerosis-like lesions in the microvasculature [
28,
29]. The removal of cholesterol from foam cells is regulated by the transcription factor peroxisome proliferator-activated receptor alpha and gamma (PPAR‑α, PPAR-γ), which inhibits lipid-metabolizing enzymes [
30,
31]. As the activity of PPAR‑α after radiotherapy appears to be drastically decreased [
32] 8 and 40 weeks after irradiation [
33], cholesterol removal is prevented and the synthesis as well as uptake of fatty acids might therefore be increased. An increased formation of foam cells can induce chronic inflammation which promotes the onset of atherosclerosis [
34,
35], but the disease remains asymptomatic for nearly a decade before the cardiovascular disease-related mortality in patients increases significantly [
36,
37] after radiation of left-sided breast tumors [
38]. This implies that several diseases like obstructive coronary artery disease, myocardial fibrosis, pericardial disease, arrhythmias, and valvular abnormalities are most likely related to a previous thoracic radiotherapy [
9]. The apex and the anterior wall of the heart are the anatomic sites that obtain the highest radiation doses. Therefore, the risk of developing radiation-induced diseases such as atherosclerosis leading to myocardial infarction years later is very high, especially in this part of the heart [
39,
40]. Depending on dose and duration of the radiation therapy, the risk of developing coronary artery diseases increases significantly [
36].
In the present study, some of the main late occurring side effects of irradiation of the heart, such as EC dysfunction leading to chronic inflammation and the development of atherosclerosis, could be documented on primary ECs of the heart and lung in a mouse model.
Ligation of the LAD is used as a model for myocardial infarction in mice [
41]. Interestingly, an artificially induced heart infarction in mice also causes a significant increase in the expression of prominent inflammatory markers CD54 (ICAM-1), CD102 (ICAM-2), and CD106 (VCAM-1) on primary heart ECs of the infarcted area compared to those of the noninfarcted area of the same heart. This may suggest that chronically induced inflammatory effects caused by a local heart irradiation on primary microvascular ECs also play a role in response to a heart infarction. Hence, acute inflammatory processes after myocardial infarction are similar to long-term inflammatory effects after radiotherapy. Therefore, we speculate that strategies preventing acute and chronic inflammation of the microvasculature in the heart might help to avoid cardiovascular complications. One strategy to avoid inflammation involving lipid metabolism might be the re-activation of PPAR‑α by fenofibrate, which decreases foam cell formation and might thereby help to decrease cardiac inflammation [
30].
It is known that the interaction between heart and lung influences the total tolerated radiation dose. The radiation of one organ leads to a lower tolerated dose in the other organ [
16]. This effect is responsible for tachypnoea and right ventricular hypertrophy after radiation with doses above 20 Gy for the heart and 5 Gy for the lungs [
42‐
44]. As the highest radiation dose in the described experiments was 16 Gy for the heart and 3 Gy for the lung, the reciprocal effect between heart and lung tissue after radiation appears to be less relevant under these conditions and, therefore, long-term survival of the mice was maintained, which enabled us to study chronic irradiation effects. Our data might suggest that chronic negative abscopal effects induced by complete heart and partial lung irradiation with 16 Gy are unlikely, since in our mouse model, the nonirradiated lung ECs remained unaffected up to 50 weeks after irradiation.