Evaluation of Senescence and Its Prevention in Doxorubicin-Induced Cardiotoxicity Using Dynamic Engineered Heart Tissues

Abstract

Background: Doxorubicin is an essential cancer treatment, but its usefulness is hampered by the occurrence of cardiotoxicity. Nevertheless, the pathophysiology underlying doxorubicin-induced cardiotoxicity and the respective molecular mechanisms are poorly understood. Recent studies have suggested involvement of cellular senescence.

Objectives: The aims of this study were to establish whether senescence is present in patients with doxorubicin-induced cardiotoxicity and to investigate if this could be used as a potential treatment target.

Methods: Biopsies from the left ventricles of patients with severe doxorubicin-induced cardiotoxicity were compared with control samples. Additionally, senescence-associated mechanisms were characterized in 3-dimensional dynamic engineered heart tissues (dyn-EHTs) and human pluripotent stem cell-derived cardiomyocytes. These were exposed to multiple, clinically relevant doses of doxorubicin to recapitulate patient treatment regimens. To prevent senescence, dyn-EHTs were cotreated with the senomorphic drugs 5-aminoimidazole-4-carboxamide ribonucleotide and resveratrol.

Results: Senescence-related markers were significantly up-regulated in the left ventricles of patients with doxorubicin-induced cardiotoxicity. Treatment of dyn-EHTs resulted in up-regulation of similar senescence markers as seen in the patients, accompanied by tissue dilatation, decreased force generation, and increased troponin release. Treatment with senomorphic drugs led to decreased expression of senescence-associated markers, but this was not accompanied by improved function.

Conclusions: Senescence was observed in the hearts of patients with severe doxorubicin-induced cardiotoxicity, and this phenotype can be modeled in vitro by exposing dyn-EHTs to repeated clinically relevant doses of doxorubicin. The senomorphic drugs 5-aminoimidazole-4-carboxamide ribonucleotide and resveratrol prevent senescence but do not result in functional improvements. These findings suggest that preventing senescence by using a senomorphic during doxorubicin administration might not prevent cardiotoxicity.

Keywords: cardiotoxicity; doxorubicin; hPSC-derived cardiomyocytes; senescence; senomorphic drugs.

Graphical abstract

Figure 1

Senescence Is Present in the Hearts of Patients With Doxorubicin-Induced Cardiotoxicity Gene expression of several senescence and apoptosis markers was assessed in patient hearts. No changes in apoptosis genes were observed, while senescence genes were up-regulated. (A) Western blot (WB) images of BAX, Bcl-2, caspase-3, and cleaved caspase-3. The positive control for caspase-3 consists of human pluripotent stem cell–derived cardiomyocytes treated with 5 μM doxorubicin for 24 hours. (B) WB analysis of BAX, Bcl-2, and caspase-3 expression. (C) Immunohistochemical p16 staining of control and patient cardiac tissue. p16-positive cells are indicated with black arrows. Scale bar: 100 μm. Quantification is normalized by tissue area. (D) WB analysis of p16, p21, (E) p53, phosphorylated p38MAPK (against total p38MAPK), and NF-κB expression in control and cardiotoxic hearts. (F) Messenger RNA (mRNA) expression of senescence-associated secretory phenotype factors in control subjects and patients (log-transformed fold changes [FCs]). Bar graphs show mean ± SEM and individual values. The heatmap shows means. ∗P < 0.05 and ∗∗P < 0.01. Statistical significance was determined using Student’s t-test, except for (B). Because of an outlier, the Mann-Whitney U test was used. CTR = control; ns = not significant.

Figure 2

p16 and p21 Follow Predicted Expression Patters in Senescent Dyn-EHTs Dynamic engineered heart tissues (dyn-EHTs) were exposed to multiple low doses of doxorubicin (Dox). Increases of CDKN2A (p16) and CDKN1A (p21) over time were observed. (A) Treatment scheme for the dyn-EHTs. Forty-eight hours of 0.1 μmol/L doxorubicin was followed by 5 days in fresh media, repeated 4 times (4-hit). (B) mRNA expression of senescence markers CDKN2A (p16) and CDKN1A (p21) over time. (C) Immunohistochemical p16 staining of CTR and 4-hit-treated tissues. p16-positive cells are indicated with black arrows. Scale bar: 300 μm. Linear graphs show mean ± SEM. Bar graphs show mean ± SEM and individual values. ∗P < 0.05. Statistical significance was determined using Student’s t-test or 1-way analysis of variance with the Bonferroni post hoc test for multiple pairwise comparisons. Abbreviations as in Figure 1.

Figure 3

Up-Regulation of Senescent Markers To confirm senescence in the 4-hit dyn-EHTs, expression of several markers was assessed, using WB and mRNA expression. (A) Representative WBs of p16, p21, p53, phosphorylated p38MAPK (against total p38MAPK), and NF-κB expression and (B) quantification. (C) mRNA expression of senescence-associated secretory phenotype factors in control and 4-hit dyn-EHTs (log-transformed FCs). (D) mRNA expression (log-transformed FC). (E) WB analysis of caspase-3 and cleaved caspase-3. The positive control consists of human pluripotent stem cell–derived cardiomyocytes treated with 5 μM doxorubicin for 24 hours. (F) Quantification of caspase-3 protein expression. Bar graphs show mean ± SEM and individual values. ∗P < 0.05 and ∗∗P < 0.01. Statistical significance was determined using Student’s t-test. Abbreviations as in Figures 1 and 2.

Figure 4

Function of Senescent Dyn-EHTs Is Impaired Senescent dyn-EHTs showed decreased functionality. (A) Treatment scheme with elongation of a representative tissue over time. Quantification of diastolic length (B), diameter (C), diastolic force (D), systolic force (E), diastolic stress (F), systolic stress (G), and contractile stress (H). Gray line denotes CTR (n = 12); red line denotes 4-hit (n = 17). Linear graphs show mean ± SEM. Bar graphs show mean ± SEM and individual values. ∗P < 0.05 and ∗∗P < 0.01 against control on the same day; ^#P < 0.05, 4-hit against its own baseline. Statistical significance was determined using 1-way analysis of variance with the Bonferroni post hoc test for multiple pairwise comparisons. Abbreviations as in Figure 2.

Figure 5

Mitochondrial Function Increases Upon Cotreatment With Doxorubicin and Senomorphic Drugs Mitochondrial function for doxorubicin- and senomorphic drug–treated cells was assessed using Seahorse. (A-C) Trace of Mito Stress tests. (D-F) Quantification of Mito Stress tests (basal respiration, adenosine triphosphate [ATP] production, and respiratory reserve). Linear graphs show mean ± SEM. Bar graphs show mean ± SEM and individual values. ∗P < 0.05 and ∗∗P < 0.01 against control. Statistical significance was determined using 1-way analysis of variance with the Bonferroni post hoc test for multiple pairwise comparisons. AICAR = 5-aminoimidazole-4-carboxamide ribonucleotide; OCR = oxygen consumption rate; other abbreviations in Figure 1.

Figure 6

Expression of Senescence Markers Is Decreased by Cotreatment With Senomorphic Drugs Prevention of senescence by senomorphic drugs on gene expression level was assessed in 4-hit + resveratrol and 4-hit + AICAR dyn-EHTs. (A) Treatment scheme showing senomorphic treatments. (B) Representative immunohistochemical p16 staining. Scale bar: 300 μM. (C) Quantification of p16 stainings. (D) WB for p16. (E) Quantification of WB for p16. (F) mRNA expression in control and 4-hit dyn-EHTs, cotreated with resveratrol or AICAR (log-transformed FCs). Bar graphs show mean ± SEM and individual values. ∗P < 0.05 against control and ^#P < 0.05 against 4-hit. Statistical significance was determined using 1-way analysis of variance with the Bonferroni post hoc test for multiple pairwise comparisons. Abbreviations as in Figures 1, 2, and 5.

Figure 7

Function of 4-Hit Dyn-EHTs Cotreated With Senomorphic Drugs Is Impaired To prevent senescence, 4-hit dyn-EHTs were cotreated with AICAR or resveratrol, and function was assessed. (A to F) Line graphs showing diastolic length, diastolic force, and systolic force measured in dyn-EHTs treated with doxorubicin and resveratrol (A to C) or AICAR (D to F). Linear graphs show mean ± SEM. Significance of the functional graphs was determined using repeated-measures analysis of variance (Supplemental Table 3) and post hoc tests. Abbreviations as in Figures 2 and 5.

Figure 8

Four-Hit Dyn-EHTs Cotreated With Senomorphic Drugs Show Increased Apoptosis and Fibrosis To explain functional decline, apoptosis and fibrosis markers were assessed in cotreated dyn-EHTs. (A) Masson’s staining, representative pictures. (B) Quantification of Masson’s staining. Scale bar: 200 μM. (C) mRNA expression of fibrosis markers in control and 4-hit dyn-EHTs, cotreated with resveratrol (Res) or AICAR (log-transformed FCs). (D) Total number of cells per tissue area. (E) WB analysis of protein expression of caspase-3, BAX, and Bcl-2. Bar graphs show mean ± SEM and individual values. Significant P values are shown in the figure. Significance in fibrosis were determined using 1-way analysis of variance. ∗P < 0.05 against control and ^#P < 0.05 against 4-hit. Abbreviations as in Figures 1, 2, and 5.

Central Illustration

Senescence Is Involved in Doxorubicin-Induced Cardiotoxicity, and Senomorphic Drugs Did Not Improve Function Senescence has been suggested to be involved in doxorubicin-induced cardiotoxicity. Here, we found senescence to be present in patient tissue. A similar phenotype was observed when dynamic engineered heart tissues (dyn-EHTs) were exposed to multiple low doses of doxorubicin, separated by recovery periods. We cotreated dyn-EHTs with senomorphic drugs to prevent the phenotype, which did not improve function. LVEF = left ventricular ejection fraction.