Generation of cardiomyocyte-specific ADAM17-knockout mice

ADAM17flox/flox (A17fl/fl) mice with C57BL/6J background were obtained from the Jackson laboratory and α-myosin heavy chain (α-MHC)-Cre mice were derived from the Model Animal Research Center at Nanjing University. A17fl/fl mice were crossed with α-MHC-Cre mice to obtain cardiomyocyte-specific ADAM17-knockout mice (A17α-MHCKO) through excising specifically exon 2 of the ADAM17 gene in cardiomyocytes. PCR fragments isolated from mouse tails and amplified from genomic DNA were used to confirm genotypes of A17fl/fl mice and A17α-MHCKO mice. The primer sequences and cycling conditions were listed in Supplementary Table 1. The knockout efficiency was assessed by measuring ADAM17 expression with western blot and RT-PCR.

ADAM17 and TRAF3 overexpression by adeno-associated virus 9

To overexpress ADAM17 and TRAF3 in the cardiomyocyte of C57BL/6J mice, a cDNA encoding ADAM17 and TRAF3 sequence and cardiomyocyte-specific cTnT promoter were produced and inserted into AAV serotype 9 packaging vehicle, and AAV9 vehicle was applied as a negative control. AAV9-cTnTp-3Flag-ADAM17 (AAV9-oeA17), AAV9-cTnTp-3Flag-NC (AAV9-oeNC), AAV9-cTnTp-mcherry-TRAF3 (AAV9-oeTRAF3) and AAV9-cTnTp-mcherry-NC (AAV9-oeNC) were purchased from GeneChem Co., Ltd. (Shanghai, China) and injected to 8-week-old male mice through a tail vein with a dose of 200 μL virus at a titer of 5 × 1011 v.g./mL per mouse. The efficiency of virus transfection was measured by the expression of ADAM17 and TRAF3 by western blot and RT-PCR.

TRAF3 knockdown by adeno-associated virus 9

To knockdown cardiomyocyte TRAF3 in C57BL/6J mice and ADAM17 overexpression mice, a cDNA encoding TRAF3 sequence and cardiomyocyte-specific cTnT promoter was produced and inserted into AAV9 packaging vehicle, and AAV9 vehicle was applied as a negative control. AAV9-cTnTp-mcherry-TRAF3-shRNA (AAV9-shTRAF3) and AAV9-cTnTp-mcherry-NC-shRNA (AAV9-shNC) were purchased from GeneChem Co., Ltd. (Shanghai, China) and injected to C57BL/6J mice and ADAM17 overexpression mice as previously described. The efficiency of virus transfection was measured by the expression of TRAF3 by western blot and RT-PCR.

Animal model and grouping

Animal experiments consisted of eight proportions. In the first proportion of the in vivo experiments (Fig. 1a), male C57BL/6J mice aged 8 weeks were randomly divided into two groups (n = 10/group): normal saline (NS) group and doxorubicin (DOX) group. The mice in the DOX group were injected with an accumulative dose of 20 mg/kg doxorubicin (MCE, USA, cat# HY-15142) [5 mg/kg intraperitoneal (i.p.) injection at days 0, 7, 14 and 21], whereas an equivalent volume of normal saline was administered by i.p. injection to the NS group.64 Although doxorubicin is generally delivered through intravenous infusion in clinical setting, i.p. injection is widely used in mouse models of doxorubicin-induced cardiomyopathy for the advantages of stable serum drug concentration, high success rate and easy reproducibility.65 These mice were observed daily and finally euthanized with an overdose of sodium pentobarbital (200 mg/kg, i.p.) 4 weeks after the final doxorubicin or saline injection.

In the second proportion of the in vivo experiments (Fig. 2a), A17α-MHCKO mice and their littermates A17fl/fl mice were collected, who were randomly divided into four groups (n = 10 in each group): A17fl/fl + NS, A17fl/fl + DOX, A17α-MHCKO + NS, A17α-MHCKO + DOX. The injection method and doses of doxorubicin and normal saline were the same as those described earlier, and these mice were euthanized 4 weeks after the last doxorubicin or saline injection.

In the third proportion of the in vivo experiments (Fig. 3a), C57BL/6J male mice aged 8 weeks were selected and injected with ADAM17-overexpressing AAV9 and corresponding virus vehicle through the tail vein as previously described. The mice were randomly divided into four groups (n = 10 per group): AAV9-oeNC + NS, AAV9-oeNC + DOX, AAV9-oeA17 + NS and AAV9-oeA17 + DOX. The injection method and doses of normal saline and doxorubicin were the same as those described earlier. Four weeks after virus injection, the mice received i.p. injection of doxorubicin or equal volume of normal saline, as described earlier, and were euthanized 4 weeks after the last doxorubicin and saline injection.

In the fourth proportion of the in vivo experiments (Supplementary Fig. 9a), C57BL/6J male mice aged 8 weeks were collected and injected with TRAF3-knockdown AAV9 and corresponding virus vehicle through the tail vein as previously described. The mice were randomly divided into four groups (n = 10/group): AAV9-shNC + NS, AAV9-shNC + DOX, AAV9-shTRAF3 + NS and AAV9-shTRAF3 + DOX. The injection method and doses of normal saline and doxorubicin were the same as those described earlier. Four weeks after virus injection, the mice received i.p. injection of doxorubicin or equal volume of normal saline, as described earlier, and were euthanized 4 weeks after the last doxorubicin and saline injection.

In the fifth proportion of the in vivo experiments (Supplementary Fig. 10a), C57BL/6J male mice aged 8 weeks were applied and injected with TRAF3-overexpressing AAV9 and corresponding virus vehicle through the tail vein as previously described. The mice were randomly divided into four groups (n = 10/group): AAV9-oeNC + NS, AAV9-oeNC + DOX, AAV9-oeTRAF3 + NS and AAV9-oeTRAF3 + DOX. The injection method and doses of normal saline and doxorubicin were the same as those described earlier. Four weeks after virus injection, the mice received i.p. injection of doxorubicin or equal volume of normal saline, as described earlier, and were euthanized 4 weeks after the last doxorubicin and saline injection.

In the sixth proportion of the in vivo experiments (Supplementary Fig. 11a), the AAV9-oeNC and AAV9-oeA17 mice were injected with TRAF3-knockdown AAV9 and corresponding virus vehicle through the tail vein as previously described. The mice were randomly divided into four groups (n = 10/group): AAV9-oeNC + AAV9-shNC + DOX, AAV9-oeNC + AAV9-shTRAF3 + DOX, AAV9-oeA17 + AAV9-shNC + DOX, and AAV9-oeA17 + AAV9-shTRAF3 + DOX. The injection method and dose of doxorubicin were the same as those described earlier. Four weeks after virus injection, the mice received i.p. injection of doxorubicin as described earlier, and were euthanized 4 weeks after the last doxorubicin injection.

In the seventh proportion of the in vivo experiments (Supplementary Fig. 15a), C57BL/6J male mice at the age of 8 weeks were selected and randomly divided into three groups (n = 10/group): NS, DOX, DOX + infliximab, who received i.p. injection of normal saline, doxorubicin and infliximab, a monoclonal antibody of TNF-α,66 respectively. The injection method and doses of normal saline and doxorubicin were the same as those described earlier. Infliximab was injected after the last dose of doxorubicin, and 4 weeks later these mice were euthanized.

In the eighth proportion of the in vivo experiments (Supplementary Fig. 17a), 8-week-old C57BL/6J male mice were used and randomly divided into three groups (n = 10/group): NS, DOX, DOX + 5Z-7-ox, who received i.p. injection of normal saline, doxorubicin and 5Z-7-ox, an inhibitor of TAK1,67,68 respectively. The injection method and doses of normal saline and doxorubicin were the same as those described earlier. 5Z-7-ox was injected after the last dose of doxorubicin and 4 weeks later these mice were euthanized.

Different mouse groups were assigned in a randomized manner and investigators were blinded to the allocation of different groups when conducting drug treatments and outcome evaluations. Based on our preliminary experiments, we assume that type I error (α) and type II error (β) are 5% and 0.20%, respectively, with a power >0.80, and the sample size required is 6. By taking into account of accidental mouse death, we set n = 10 in each mouse group of this study. Exact animal numbers are shown in respective figure legends.

Mouse tumor models and tumor studies

To examine doxorubicin-induced cardiotoxicity in the tumor-bearing mice, mouse breast cancer cells E0771 were subcutaneously implanted into the mammary fat pads of 8-week-old female A17fl/fl and A17α-MHCKO mice, respectively (Fig. 8a). These breast cancer-bearing mice were randomly divided into four groups (n = 10/group): A17fl/fl + NS, A17fl/fl + DOX, A17α-MHCKO + NS, A17α-MHCKO + DOX. To avoid the effects of gender differences, mouse melanoma cells B16F10 were subcutaneously injected into the right flank of the 8-week-old male A17fl/fl and A17α-MHCKO mice, respectively (Fig. 8h). These melanoma-bearing mice were randomly divided into four groups (n = 10/group): A17fl/fl + NS, A17fl/fl + DOX, A17α-MHCKO + NS and A17α-MHCKO + DOX. The injection method and doses of doxorubicin and normal saline were the same as previously described and these mice were euthanized 1 week after the last doxorubicin or saline injection. Tumor volumes (mm3) were measured with a caliper and calculated as V(mm3) = (0.5 × length × width2).60 At the end of the experiment, all tumors were dissected and weighted for analysis.

RNA-sequencing analysis

The library construction and sequencing were performed by Sinotech Genomics Co., Ltd. (Shanghai, China). Total RNA was isolated using RNeasy mini kit (Qiagen, Germany) from mouse myocardial tissues, and the RNA concentration and quality were determined by the Qubit®3.0 Fluorometer (Life Technologies, USA) and the Nanodrop One spectrophotometer (Thermo Fisher Scientific Inc., USA). Paired-end libraries were synthesized by using the Stranded mRNA-seq Lib Prep Kit for Illumina (ABclonal, China) following preparation guide. Gene abundance was expressed as fragments per kilobase of exon per million reads mapped (FPKM). Stringtie software was used to count the fragments within each gene, and the TMM algorithm was used for normalization. Differential expression analysis for mRNA was performed using R package edgeR. Differentially expressed RNAs between A17fl/fl + NS and A17fl/fl + DOX, and between A17α-MHCKO + DOX and A17fl/fl + DOX mice with |log2(FC)| value > 1, p value < 0.05 and one group’s mean FPKM > 1, considered as significantly modulated, were retained for further analysis. We performed a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (http://www.genome.ad.jp/kegg) via enrich R package (version 3.4.3).

High-pressure liquid chromatography (HPLC) analysis

High-pressure liquid chromatography (HPLC) was performed as described previously.69,70 Fifty microliters mouse serum was transferred to 1.5 mL polypropylene centrifuge tube, and then 150 μL methanol was added, the supernatant was collected after mixing and centrifugation. In addition, an appropriate amount of doxorubicin standard (MCE, USA) was used to make a reserve solution of 2 mg/mL and diluted with methanol into a series of standard curve working solutions with concentration gradients. These working solutions were diluted ten times with blank plasma and transferred to 1.5 mL polypropylene centrifuge tube with 50 μL of each solution, and then 150 μL methanol was added, which was vortex-mixed for 1 min and centrifuged at 13,000 rpm for 10 min. The supernatant was collected and the concentration of doxorubicin was quantitatively determined by high-pressure liquid chromatography–mass spectrometry (HPLC–MS). Chromatogram acquisition and integration of the compounds were processed by a software Xcalibur 3.0 (Thermo, USA), and linear regression was performed with a weighting coefficient (1/X2).

Neonatal rat cardiomyocytes (NRCMs) isolation and culture

The ventricular muscle from 1- to 3-day-old Sprague Dawley rats was rapidly isolated and the heart tissue was minced to small pieces on ice, which were then digested with 0.75 mg/mL collagenase type II for 1 h at 37 °C.71 After repeated digestion for three times, the cells were harvested and centrifuged at 800 rpm for 5 min. The cell pellet was resuspended in the high-glucose Dulbecco’s modified Eagle medium (DMEM) supplemented with 8% horse serum, 5% newborn calf serum and 1% bromodeoxyuridine in plate for 2 h in a culture flask to let fibroblasts attach. Non-cardiomyocytes were removed by differential adherence and only non-attached cardiomyocytes were collected.

Cell treatment

Cell treatment consisted of eight proportions. In the first proportion of the in vitro experiment (Supplementary Fig. 4a), in order to determine the optimal concentration and duration of doxorubicin treatment to induce cardiomyocyte apoptosis in vitro, NRCMs were treated with different concentrations of doxorubicin (0, 0.25, 0.5, 1, 2 and 4 μM) for 24 h or 1 μM doxorubicin for different time periods (0, 6, 12, 18, 24 and 36 h). Previous studies often used 1 μM of doxorubicin to stimulate NRCMs for 24 h.72 In addition, by stimulating NRCMs with different time periods and concentrations, we found that cell apoptosis was obvious and cell viability was not <50% when NRCMs were treated with 1 μM doxorubicin for 24 h. Therefore, NRCMs were stimulated with this regimen in subsequent experiments.

In the second proportion of the in vitro experiments (Supplementary Fig. 4b), to elucidate the role of ADAM17 silencing in cardiomyocyte, siRNA targeting ADAM17 (siA17) was transfected into NRCMs to knockdown the expression of ADAM17 and the control group was transfected by a negative control siRNA (siNC). These NRCMs were treated with dimethyl sulfoxide (DMSO) or 1 μM DOX for 24 h before cell collection.

In the third proportion of the in vitro experiments (Supplementary Fig. 4c), to explore the role of ADAM17 overexpression in cardiomyocyte, plasmid overexpressing ADAM17 was transfected into NRCMs (oeA17) and a negative control plasmid vector was transfected into the negative control NRCMs (NC). These two groups of NRCMs were treated with DMSO or 1 μM DOX for 24 h before cell collection.

In the fourth proportion of the in vitro experiments (Supplementary Fig. 4d), in order to verify the upstream and downstream relationship between ADAM17 and TRAF3, plasmid overexpressing ADAM17 (oeA17) and negative control plasmid vectors (NC) were first transfected into NRCMs, respectively. After 24 h, these NRCMs were transfected with siRNA targeting TRAF3 (siTRAF3) and siNC, respectively. Finally, NRCMs were treated with 1 μM DOX for 24 h before cell collection.

In the fifth proportion of the in vitro experiment (Supplementary Fig. 13a), to determine the optimal dose of TNF-α treatment for NRCMs, we treated the cells with different concentrations of TNF-α for 24 h, and selected the optimal concentration for subsequent experiments.

In the sixth proportion of the in vitro experiments (Supplementary Fig. 13b), to verify whether TNF-α induces downstream TRAF3 expression via TNFR, infliximab (MCE, USA), a chimeric monoclonal IgG1 antibody that specifically binds to TNF-α, was used to block the interaction of TNF-α with TNFR, and NRCMs were treated with NS, 10 μM TNF-α or 10 μM TNF-α + infliximab, respectively.

In the seventh proportion of the in vitro experiments (Supplementary Fig. 13c), to ascertain whether doxorubicin induces cardiac injury via upregulating TNF-α–TNFR signaling, NRCMs were pretreated with infliximab for 24 h before doxorubicin treatment, and then treated with DMSO, 1 μM DOX or 1 μM DOX + infliximab, respectively, for 24 h before cell collection.

In the eighth proportion of the in vitro experiment (Supplementary Fig. 13d), in order to verify that the downstream MAPKs pathway was activated by TAK1, NRCMs were treated with DMSO, 1 μM DOX or 1 μM DOX + TAK1 inhibitor 5Z-7-ox (MCE, USA), respectively, for 24 h before cell collection.

Cell viability assay

NRCMs (4 × 103 cells/well) were cultured in a 96-well plate and exposed to different concentrations (0, 0.25, 0.5, 1, 2 and 4 μM) of DOX for 24 h and to 1 μM DOX for different durations (0, 6, 12, 18, 24 and 36 h). After cell treatment, the culture medium was changed with a fresh medium. Then, 10 μL of CCK-8 was added to the NRCMs and the cells were cultured for 1 h at 37 °C. Optical density (OD) at 450 nm was measured by a microplate reader (BioTek, VT, USA). Cell viability = (OD of DOX cells − OD of blank control)/(OD of normal cells − OD of blank control) × 100%.

Enzyme-linked immunosorbent assay (ELISA)

Serum TNF-α levels in mice were measured using a Mouse TNF-α ELISA Kit (ANRK, China). TNF-α levels in NRCMs culture supernatants were detected with a Rat TNF-α ELISA Kit (Proteintech, China).

ADAM17 activity assay

ADAM17 activity was measured in mouse heart tissue and NRCMs using SensoLyte 520 TACE activity assay kit (AnaSpec, Fremont, CA). The heart tissue sample and NRCMs sample were homogenized in Component C containing 0.1% Triton X-100, which were incubated on ice for 15 min followed by centrifugation at 2000 × g for 15 min, and then the supernatant was collected. The biological samples were put into the plate and incubated for 10 min, and then ADAM17 substrate solution was added and incubated for 30 min at 37 °C. Terminating solution was then added to each hole and the absorbance at 490 nm/520 nm was measured using a microplate reader. The activity of each sample was measured and calculated by the formula of the standard curve.

Statistical analysis

All data were presented as mean ± SEM. Shapiro–Wilk test was used to evaluate the normality distribution of the data. For the data with normal distribution, unpaired two-tailed Student’s t-test was used between two groups, and one-way ANOVA analysis followed by Tukey post-hoc test was performed among multiple groups. For data with a non-normal distribution, Mann–Whitney U test was used between two group comparisons, and Kruskal–Wallis test followed by Dunnett’s post-hoc test was used for multiple group comparisons. Kaplan–Meier curves and Log-Rank test were applied to assess animal survival data. In all statistical comparisons, p < 0.05 was considered statistically significant. All data analysis was performed using GraphPad Prism 8 (GraphPad, CA).

See Supplementary Materials for more details of experimental procedures.