Eight-week-old male C57/Bl6 mice were maintained in accordance with the guidelines published by the US National Institutes of Health (NIH). This study was conducted in compliance with the Guide for the Care and Use of Laboratory Animals published by the National Academy Press (NIH, revised in 1996). During the animal experiment, the investigator was blinded to the group allocation.

Eight-week-old male C57/Bl6 mice were treated with an InVivoPlus anti-mouse PD-1 inhibitor (Bioxcell, West Lebanon, USA), at a concentration of 5 mg/kg, intraperitoneally, on days 1 and 14 (cycle: 28 days). After the experiments, mice were sacrificed by CO₂ inhalation.

For echocardiography, 28 days after the first treatment of PD-1 inhibitor, mice were anesthetized with 1.5%–2% isoflurane and kept warm on a heated platform. During echo process, the vital signs were monitored. Vevo 2100 was applied to observe the echocardiographic parameters in mice to assess the cardiac functions (Vevo 2100, Visual Sonics, Canada). Short axis was applied to measure left ventricular internal diameter; diastole (LVID;d), left ventricular internal diameter; systole (LVID;s), left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV). The values of ejection fraction (EF) and fractional shortening (FS) were calculated.

RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, USA), and cDNA was synthesized using an Improm II reverse transcription kit (Promega, Wisconsin, USA). qRT-PCR was performed with SYBR Green to detect mRNA levels. The mRNA levels were calculated relative to the control Gapdh (for mRNA) or U6 (for miRNA) using the 2^-ΔΔCq method. The primers for target genes are listed in table 1.

Deparaffinized tissue sections were stained with antibodies for sarcomeric-α-actin (SAA; 1:30, Abcam, ab9465). Thereafter, the p16^INK4a was used to detect cellular senescence (1:50; Abcam, ab211542). Blocked using 10% goat serum, the sections were incubated with the antibody overnight at 4°C. Following primary incubation, the slides were incubated with Alexa Fluor 488 (1:50; Invitrogen, Cat No A32731) or Alexa Fluor 594 (1:50; Invitrogen, Cat No A32744) for 1 hour at room temperature. The sections were counterstained with 4,6-diamino-2-phenyl indole (DAPI). Fluorescence was detected under a microscope.

Mouse ventricular myocytes were isolated from C57BL/6 mice. The animals were euthanized by cervical dislocation. A digestion solution containing 0.25% trypsin and collagenase I was applied to digest heart. After 3 hours of differential sedimentation and adhesion, the myocytes were separated. They were then cultured in low-glucose Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (FBS).

The senescence-associated β-galactosidase (SA-β-gal) assay (Cell Signaling Technology, Massachusetts, USA) was applied to assess cellular senescence. Briefly, the cells at the density of 2×10⁴ were fixed with 2% paraformaldehyde and incubated with SA-β-gal staining solution as previously reported.22

Western blot analysis was conducted as previously described.23 Primary antibodies, including Flotillin-1 (ab41927), Tsg101 (ab125011), CD63 (ab217345), p21 (ab188224), p16^INK4a (ab211542), and β-actin (ab179467) were purchased from Abcam. PNUTS (14171) was purchased from Cell Signaling Technology.

Chemically modified antisense oligonucleotides were designed to target miR-34a-5p. Eight-week-old C57/Bl6 mice were subjected to sham or PD-1 inhibitor treatment. Then, the mice were injected (intraperitoneally) with antagomir-34a-5p (80 mg/kg) or control antagomir (80 mg/kg) for three consecutive days as previously reported.24

HL-1 murine cardiomyocytes were maintained in fibronectin-coated flasks, supplemented with 10% FBS, 100 U/mL penicillin, 100 mg/mL streptomycin, and 2 mM L-glutamine. The cell culture was exposed to 5 µg/mL of the treatment solution for short durations.25

The rate of cell proliferation was assessed using the cell counting kit-8 (CCK-8) assay, performed following the manufacturer’s protocol. Briefly, the cells grown in a 96-well plate were incubated with the CCK-8 solution for 1 hour at 37°C, and the absorbance of each well at 450 nm was recorded.

Cold anhydrous ethanol (70%) fixed cells were treated with propidium iodide (PI) (Sigma, St. Louis, Missouri, USA) and RNase A. A flow cytometer equipped with Cell Quest software was used to detect the cell cycle distribution.

Cellular apoptosis was detected using an Annexin V fluorescein isothiocyanate (FITC) and PI kit. Briefly, washed cells were resuspended in Annexin V binding buffer, then stained with Annexin V and PI at room temperature and then analyzed by bivariate flow cytometry using a BD FACS Canto II equipped with BD FACSDiva Software (Becton-Dickinson, San Jose, California, USA).

Cardiac macrophages were isolated as previously described.26 Briefly, the excised hearts were washed with warmed Hanks’ balanced salt solution. Then the hearts were placed in warm digestion buffer containing Liberase TH (5 mg/mL, Roche) and DNase1 (2000 units, Sigma) for 5 min, triturated with a pipette, and passed through a 40 µm filter. To isolate macrophages, the centrifuged cells were incubated with 50 µL of anti-CD45 magnetic beads for 30 min. The cells were then allowed to flow through the MACS magnetic column with a magnet in place.

The exosomes were isolated and purified from the supernatants of normal macrophages and 5 µg/mL PD-1 inhibitor–treated macrophages. Briefly, the supernatants were collected after 48 hours culture. The exosome quick extraction solution was added to the filtered solution at a 1:5 ratio and stored at 4°C for at least 12 hours.

The morphology of exosomes was observed by transmission electron microscopy (TEM, JEM-1400plus). The size and concentration were determined by nanoparticle tracking analysis (NTA). Western blot analysis was used for detecting exosomal markers Flotillin-1, Tsg101, and CD63. At meanwhile, β-actin in the whole-cell lysate was used as control.

Macrophages were seeded into six-well plates at a density of 1×10⁵ cells per well. To induce the inhibition of miR-34a-5p, the cells were transfected with an miR-34a-5p inhibitor or a negative control (NC) inhibitor (pre-miR miRNA Precursors, Life Technologies, Karlsruhe, Germany) using X-treme transfection reagent (Roche Applied Science, Penzberg, Germany). The cells were harvested for further analysis 48 hours after transfection, and qRT-PCR was applied to quantify the transfection efficiency.

Relative telomere length quantification in cardiomyocytes was performed using the PCR approach based on a previously established method,27 using Gapdh as the normalizing gene. The primer pairs used to detect the telomere length are listed in table 1.

The telomerase activity of cardiomyocytes was examined using a Telo TAGGG Telomerase PCR ELISA Plus kit according to the manufacturer’s protocols as described previously.28

The 3′-UTRs of PNUTS were synthesized, annealed, and inserted into the SacI and HindIII sites of the pmiR-reporter luciferase vector (Ambion), downstream of the luciferase stop codon to induce the mutagenesis of PNUTS. The cardiomyocytes were seeded into a 24-well plate for luciferase assay. After overnight culture, the cells were co-transfected with a wild-type (WT) or mutated plasmid and equal amounts of miR-34a-5p mimic or miR-NC mimic. Luciferase assays were performed using a dual-luciferase reporter assay system (Promega) 24 hours after transfection.

For the overexpression of PNUTS, the cardiomyocytes were transduced with adenoviral PNUTS (Ad-PNUTS) or adenoviral control (Ad-Ctrl) as described previously.29 The transfection efficiency was confirmed by qRT-PCR and western blot analysis.

Data were expressed as mean±SD. Statistical significance of differences among groups was tested by one-way analysis of variance (ANOVA) or repeated measures ANOVA. Comparisons between two groups were done using Student’s t-test or paired t-test. A value of p<0.05 was considered statistically significant.

Whether a PD-1 inhibitor impaired heart function by increasing heart senescence was investigated in a murine model. Echocardiography showed a significant decrease in left ventricular EF and FS in the PD-1 inhibitor group compared with the sham group (figure 1A–C). However, neither the ratio of heart weight to body weight nor the lung weight to body weight showed significant difference between control and PD-1 inhibitor treatment group (figure 1D,E).

Immunostaining was conducted to verify whether organ senescence involved cellular senescence of all cell types or a specific cell type. More p16^INK4a-positive cells were located on cardiomyocytes, which were also positive for cardiomyocyte marker SAA in the PD-1 inhibitor–treated mice (figure 1F). Cardiomyocytes were isolated from control and PD-1 inhibitor–treated mice (figure 1G). SA-β-gal staining revealed more positive cells in PD-1 inhibitor–treated mouse hearts compared with control mouse hearts (figure 1H,I). qRT-PCR and Western blot analysis confirmed increased expression levels of senescence-related genes p21 and p16 in PD-1 inhibitor–treated cardiomyocytes compared with controls (figure 1J–O).

Cardiomyocytes were isolated for the characterization of senescence (figure 2A). Microarray analysis was performed between cardiomyocytes isolated from the control mice and PD-1 inhibitor–treated mice (figure 2B). As shown in figure 2B, miR-34a-5p in heart treated with a PD-1 inhibitor was probably responsible for cardiac injury and cardiac senescence. QRT-PCR further confirmed that miR-34a-5p was more abundant in cardiomyocytes and macrophages isolated from PD-1 inhibitor–treated mouse hearts (figure 2C,D). Meanwhile, miR-34a-5p mRNA levels in the heart tissue from PD-1 inhibitor–treated mice increased in a time-dependent manner (figure 2E). MiR-34a-5p inhibitor transfection was performed in vivo to evaluate whether miR-34a-5p inhibition treatment influenced PD-1 inhibitor–modulated cardiac senescence and cardiac function (figure 2F). Echocardiography data revealed that PD-1 inhibitor–treated mice had significant impairment in cardiac function compared with control mice. However, mice treated with PD-1 inhibitor+miR-34a-5p inhibitor showed recovery in cardiac function (figure 3A–C). Meanwhile, more SA-β-gal-positive cardiomyocytes were found in PD-1 inhibitor–treated mice compared with control mice. However, the miR-34a-5p inhibitor impaired the increase in positive cells (figure 3D,E). As shown in figure 3F,G, the expression levels of senescence-related genes p21 and p16 increased in PD-1 inhibitor–treated mouse cardiomyocytes compared with control mouse cardiomyocytes. However, the expression of senescence-related genes p21 and p16 decreased in the PD-1 inhibitor+miR-34a-5p inhibitor group compared with the PD-1 inhibitor group, but no change was observed in the PD-1 inhibitor+miR NC inhibitor group.

The proliferation under PD-1 treatment was first examined to investigate the potential pro-senescence effects of a PD-1 inhibitor on cardiomyocytes. No hampering in proliferation was observed when HL-1 cardiomyocytes were exposed to a PD-1 inhibitor (online supplemental figure S1A). Meanwhile, the administration of the PD-1 inhibitor did not affect the cell cycle (online supplemental figure S1B,C). The PD-1 inhibitor did not induce SA-β-gal-positive cells (online supplemental figure SD,E). These results indicated that the PD-1 inhibitor only did not affect the biological activity of cardiomyocytes, thus leading to cellular senescence. Also, the PD-1 inhibitor did not induce cellular apoptosis (online supplemental figure S1F,G). Similarly, we found that miR-34a-5p expression was slightly increased in PD-1 inhibitor–treated cardiomyocytes, however, with no significant difference (online supplemental figure S1H).

Given the conditioned exosomes isolated from macrophages had a pro-senescent effect, whether exosomes derived from PD-1 inhibitor–pretreated macrophages affected cardiomyocyte senescence was determined. TEM and NTA showed that exosomes isolated from macrophages exhibited a round morphology and size of 50–100 nm. Moreover, the expression of exosome markers Flotillin-1, Tsg101, and CD63 was confirmed by western blot analysis (figure 4A–C). Also, the effects of PD-1 inhibitor–conditioned exosomes on cardiomyocyte senescence was determined. Exosome derived from PD-1 inhibitor–treated macrophages (exosome^{PD-1 inhibitor}) induced cardiomyocyte senescence, showing that more cells were trapped in the G0/G1 phase as measured using flow cytometric analysis cell scan (FACS) (figure 4D,E), with the increased expression levels of cellular senescence–related genes p21 (figure 4F) and p16 (figure 4G), decreased telomere length (figure 4H) and telomerase activity (figure 4I), and a higher percentage of SA-β-gal-positive cells (figure 4J,K). However, exosomes from untreated macrophages showed no significant pro-senescent effect, indicating that the observed pro-senescent effects were PD-1 inhibitor-specific.

Exosomal RNA was extracted from three PD-1 inhibitor–treated macrophages and three control macrophages, and the expression profile was analyzed using the miR array to identify exosomal miRs. As a result, the maximum difference was for miR-34a-5p, a cellular senescence–related miR, which was confirmed by qRT-PCR (figure 5A,B). Exosomes prepared from macrophages were labeled with DiI to confirm whether this miR could be transferred to cardiomyocytes through exosomes. Exosome uptake by cardiomyocytes was observed (figure 5C). Also, miR-34a-5p was detected in cardiomyocytes incubated with exosome^{PD-1 inhibitor} compared with cardiomyocytes without any treatment. As expected, miR-34a-5p had the strongest upregulation in cardiomyocytes incubated with exosome^{PD-1 inhibitor} (figure 5D). Furthermore, the expression of miR-34a-5p in macrophages was silenced. The qRT-PCR results confirmed a significant decrease in the expression level of miR-34a-5p in macrophages even if treated with a PD-1 inhibitor (figure 5E). Macrophages were transfected with an miR-34a-5p inhibitor or an miR-NC inhibitor and were then subjected to a PD-1 inhibitor. Then, exosomes were collected. The qRT-PCR results revealed that the PD-1 inhibitor induced miR-34a-5p accumulation in the exosomes, which was abolished by antago-miR (figure 5F). Exosomes derived from macrophages, which were transfected with an miR-34a-5p inhibitor or an miR-NC inhibitor and then treated with a PD-1 inhibitor or treated only with a PD-1 inhibitor, were added to cardiomyocytes. In parallel experiments, cardiomyocytes without any treatment were used as control. Then, the expression of miR-34a-5p in cardiomyocytes was examined using qRT-PCR. The results revealed that exosome^{PD-1 inhibitor} transferred miR-34a-5p to cardiomyocytes, while this transfer was destroyed by endogenous miR-34a-5p inhibition in macrophages (figure 5G).

Subsequently, exosomes derived from macrophages, which were transfected with miR-34a-5p inhibitor or an miR-NC inhibitor and then treated with a PD-1 inhibitor or treated only with a PD-1 inhibitor, were added to cardiomyocytes. Exosome^{PD-1 inhibitor} induced cardiac senescence, while miR-34a-5p inhibition in macrophages significantly reduced the percentage of G0/G1-phase cells (figure 6A), the number of SA-β-gal-positive cells (figure 6B,C), and the expression of classical senescence-associated genes (figure 6D,E), while recovered the telomere length and telomerase activity (figure 6F,G). These results strongly supported that macrophage-derived exosomes with a PD-1 inhibitor exerted a pro-senescent effect partially through direct miR-34a-5p transfer.

To identify the target genes of miR-34a-5p regulation, the bioinformatics database was used to speculate a putative binding site between miR-34a-5p and PNUTS (online supplemental figure S2A), which was confirmed by the dual-luciferase gene reporter assay. The relative luciferase activity was significantly weakened in the WT PNUTS+miR-34a-5p mimic group (online supplemental figure S2B). The cardiomyocytes were treated with exosome^{PD-1 inhibitor} to investigate the regulatory effect of exosomal-delivered miR-34a-5p on PNUTS expression in cardiomyocytes; the untreated cardiomyocytes were used as control. As expected, exosome^{PD-1 inhibitor} markedly inhibited PNUTS expression in cardiomyocytes, while the inhibition of miR-34a-5p in macrophages rescued the expression of PNUTS (online supplemental figure S2C,D). These data suggested that PNUTS were genuine targets of miR-34a-5p in cardiomyocytes.

Ad-PNUTS transfection enforced PNUTS overexpression in cardiomyocytes to explore the role of PNUTS in exosome^{PD-1 inhibitor} -induced senescence, which was confirmed by qRT-PCR (figure 7A) and western blot analysis (figure 7B,C). Cardiomyocytes were transfected with Ad- PNUTS or Ad-Ctrl as control, and treated with exosome^{PD-1 inhibitor}. In parallel, cardiomyocytes were treated with exosome^{PD-1 inhibitor}. The untreated cardiomyocytes were used as control. PNUTS overexpression restrained the pro-senescent effects of exosome^{PD-1 inhibitor}, decreasing the percentage of G0/G1-phase cells (figure 7D), the number of SA-β-gal-positive cells (figure 7E,F), and the expression of classical senescence-associated genes p21 and p16 (figure 7G,H). Meanwhile, it increased the telomere length and telomerase activity (figure 7I,J).