TAT‐dextran–mediated mitochondrial transfer enhances recovery from models of reperfusion injury in cultured cardiomyocytes

Abstract Acute myocardial infarction is a leading cause of death among single organ diseases. Despite successful reperfusion therapy, ischaemia reperfusion injury (IRI) can induce oxidative stress (OS), cardiomyocyte apoptosis, autophagy and release of inflammatory cytokines, resulting in increased infarct size. In IRI, mitochondrial dysfunction is a key factor, which involves the production of reactive oxygen species, activation of inflammatory signalling cascades or innate immune responses, and apoptosis. Therefore, intercellular mitochondrial transfer could be considered as a promising treatment strategy for ischaemic heart disease. However, low transfer efficiency is a challenge in clinical settings. We previously reported uptake of isolated exogenous mitochondria into cultured cells through co‐incubation, mediated by macropinocytosis. Here, we report the use of transactivator of transcription dextran complexes (TAT‐dextran) to enhance cellular uptake of exogenous mitochondria and improve the protective effect of mitochondrial replenishment in neonatal rat cardiomyocytes (NRCMs) against OS. TAT‐dextran–modified mitochondria (TAT‐Mito) showed a significantly higher level of cellular uptake. Mitochondrial transfer into NRCMs resulted in anti‐apoptotic capability and prevented the suppression of oxidative phosphorylation in mitochondria after OS. Furthermore, TAT‐Mito significantly reduced the apoptotic rates of cardiomyocytes after OS, compared to simple mitochondrial transfer. These results indicate the potential of mitochondrial replenishment therapy in OS‐induced myocardial IRI.


| INTRODUC TI ON
Prokaryotes can horizontally transfer genetic material, such as drug resistance genes, 1 while it has long been considered that eukaryotes could only vertically transfer their genetic information.
The concept was challenged by the observation that subcellular organelles, such as mitochondria, could intercellularly move through the transportation system, such as neural axons mainly constituted by actin filaments, and these are known as tunnelling nanotubes. 2 These nanotube structures have been recognized in various in vitro and in vivo conditions by independent researchers. [3][4][5][6] In organs harvested for transplantation, mitochondria that might be encircled by exosomes in the donor body and resident in the donor organ, which could evoke significant inflammation in the recipient following reperfusion. This might occasionally result in primary graft nonfunction and death during surgery. 7 Mitochondria possess several mechanisms to protect the host such as activating the innate immune response and inducing apoptosis of recipient cells through cytochrome c, which functions as an electron transporter within the respiratory super-complex. 8 Reactive oxygen species (ROS) generated by mitochondrial respiratory chains activate mitochondrial antiviral-signalling protein (MAVS), relaying the signal to NF-κB and NACHT, LRR and PYD domain-containing protein 3 (NLRP3) to release IL-1β and IL-18. 9 Upon mitochondrial membrane depolarization, cardiolipin expressed on the outer mitochondrial membrane reinforces NLRP3 activation. 10 Leakage of mitochondrial DNA from the mitochondrial matrices is a strong ligand of Toll-like receptor 9 (TLR9), which stimulates NF-κB. 11 Given that mitochondria have potential therapeutic effects in certain disease conditions, persistence of exogenous mitochondria in the vasculature or extracellular spaces might be a risk factor for inflammation. It has been reported that direct injection of isolated mitochondria into a damaged heart supported recovery of organ function. 12 To mitigate the adverse effects of exogenous mitochondria as an inducer of inflammation, it is necessary to develop a method for effective and rapid uptake by target cells.
We previously reported that isolated mitochondria were taken up by cultured cells through macropinocytosis. 4 Macropinocytosis could occur stochastically, depending upon mutual densities of cells and materials to be engulfed. One strategy to increase contacts between the cells and the material is through their enrichment in a limited space by centrifugation. 13 The transfer efficiency of isolated mitochondria in several cell types, including mesenchymal stem cells, human dermal fibroblasts and cancer cell lines, is significantly increased by centrifugation. Although the transfer of mitochondrial content was enhanced with centrifugation, every respiratory functional parameter declined with the mitochondrial content, suggesting that it was not an ideal protocol to transfer exogenous mitochondria into host cells. Magnetomitotransfer using magnetic bead-labelled mitochondria has been described using a human embryonic fibroblast cell line. 14 As macropinocytosis is initiated by receptor tyrosine kinase (RTK) activation, which is the intracytoplasmic domain of the epidermal growth factor (EGF) receptor, EGF was utilized to enhance the transfer efficiency of exogenous mitochondria into human osteosarcoma cells, resulting in a significant increase in efficiency. 15 The cell-penetrating peptide (CPP) family has been intensively investigated to deliver small chemicals or genetic materials into cells. 16 Among more than hundreds of CPPs, HIV TAT peptide demonstrated superior transfer characteristics of payloads into cells, such as nucleotides, 17 proteins 18 and small molecules, 19 and also enabled crossing the blood brain barrier. 20 Pep-1 shows less cytotoxicity among CPP family members, and the mode of action is independent of the endosomal pathway, including macropinocytosis. Pep-1 with isolated mitochondria enabled the functional recovery of fibroblasts derived from patients with MERRF (myoclonic epilepsy with ragged red fibres) syndrome. 21 Pep-1-mediated mitochondrial transfer was applied to osteosarcoma-derived cybrid with mitochondrial diseases and showed a significant improvement in transfer efficiency. 22,23 There have been no reports of a mitochondrial transfer protocol to enhance the efficiency in primary differentiated cells, such as cardiomyocytes and neurons. The transfer efficiency could highly depend upon the cell type. To investigate protocols to achieve a high transfer rate, the recipient cells for isolated mitochondrial uptake are required to be of the disease cell type.
The prevalence and mortality as a result of coronary heart disease was 16.5 million in the United States between 2011 and 2014 and 0.37 million in 2015 and is the leading cause of death. 24 Despite advancements in treatment of acute myocardial infarction (AMI), there is a considerable occurrence of heart failure, even after timely and effective reperfusion and, therefore, there are still opportunities for development of novel therapeutics. 25 The mortality and morbidity for ST-segment elevation myocardial infarction (STEMI) is approximately 7% and 22%, respectively, despite myocardial reperfusion treatment by primary percutaneous coronary intervention (pPCI) to reduce acute ischaemic injury and infarct size. 26 The unsatisfactory outcomes in patients receiving pPCI are mainly attributed to acute myocardial reperfusion injury that generates nearly half of the final myocardial infarct size. 27 Recently, some molecular targets, such as protein kinase G (PKG), reperfusion injury salvage kinase (RISK) and survivor activating factor enhancement (SAFE) signalling, have been identified by analysing the mechanism of ischaemic preconditioning. 28 Mitochondria are involved in almost all molecular aspects of reperfusion injury either directly or indirectly. To date, direct pharmacological interventions targeting mitochondria to ameliorate reperfusion injury, and several successful animal experiments for cardio-protection have been reported. 29 However, there is still no treatment for reperfusion injury, even in combination or by using multiple interventions at different time-points. 29 We suggested that exogenous mitochondrial transfer to cardiomyocytes could relieve oxidative stress (OS). Herein, we demonstrated that mitochondrial transfer could be effectively enhanced using a cell-penetrating protein TAT. In addition, OS-induced apoptosis resulting in the detrimental release of ROS which occurs at the onset of reperfusion could be ameliorated through mitochondrial transfer in hydrogen peroxide-treated rat neonatal cardiomyocytes. 30

| Cell culture
Human uterine endometrial gland-derived mesenchymal cells (EMCs) were kindly provided by Dr Umezawa. 31 The H9c2 cardiomyoblasts were obtained from the American Type Culture Collection.
Stably expressing DsRed-Mito or GFP-Mito cells were generated to visualize the isolated mitochondria, as described previously. 4

| Isolation of neonatal rat cardiomyocytes
The experimental procedures and protocols were approved by the

Animal Experiment Ethics Committee of the Kyoto Prefectural
University of Medicine and were performed in accordance with the US Animal Welfare Act. Neonatal rat cardiomyocytes (NRCMs) were isolated from 1-day-old Wistar/ST rats and cultured as previously described. 32 Briefly, the isolated hearts were digested in 0.2% collagenase type II in three cycles. The enzymatically dissociated ventricular myocytes were pre-plated twice for 30 minutes each for cardiomyocyte enrichment. The isolated cardiomyocytes were seeded into 6-well culture plates at a density of 1 × 10 5 cells per well or 24-well culture plates at a density of 4 × 10 4 cells per well and maintained in DMEM-Ham's F-12 nutrient mixture (Thermo Fisher Scientific) supplemented with 5% newborn calf serum (Thermo Fisher Scientific) and 1% penicillin/streptomycin (Thermo Fisher Scientific).

| In vitro model of oxidative stress
Oxidative stress was induced using hydrogen peroxide (H 2 O 2 ).
NRCMs were washed twice with PBS and incubated in culture medium containing H 2 O 2 (Wako, Tokyo, Japan) for 2 hours, 33 followed by reperfusion with fresh medium for 6 hours. The ideal concentration of H 2 O 2 was determined to 200 µmol/L, based upon cell death under the exposure of H 2 O 2 at various concentration ( Figure S1).

| Isolation of mitochondria
DsRed-labelled mitochondria were isolated from EMCs or H9c2 cells expressing Mito-DsRed immediately before use by mechanical disruption of cells, followed by differential centrifugation as described previously. 4 Briefly, cells were ruptured by 10-20 strokes using a syringe with 27-gauge needle in homogenization buffer [HB; 20 mmol/L HEPES-KOH (pH 7.4), 220 mmol/L mannitol and 70 mmol/L sucrose] containing a protease inhibitor mixture (Sigma-Aldrich, St. Louis, MO, USA). The homogenate was centrifuged at 400 × g for 5 minutes to remove cellular debris, and the supernatant was further centrifuged at 6000 × g for 5 minutes to pellet the isolated mitochondria-enriched fraction. The amount of isolated mitochondria was expressed as protein concentration using the Lowry method (Bio-Rad). The isolated mitochondria were resuspended in 1 mL of HB for their characterization. First, the hydrodynamic size and surface charge (zeta potential: electrostatic potential generated by the accumulation of ions at the surface of colloidal particles) of isolated mitochondria were determined by dynamic light scattering and electrophoretic light scattering measured using a Zetasizer Nano ZS (Malvern Instruments). 34 Isolated mitochondria were morphologically examined by electron microscopy to ensure enrichment of healthy mitochondria, based on the our previous report. 4 µL) was added to stop the reaction and stirred overnight. The reactant was purified using a PD-10 column to obtain the TAT-dextran complex. The concentration of TAT-dextran was measured using the conventional phenol-sulphonic acid method. 35

| Modification of isolated mitochondria and mitochondrial transfer
To obtain TAT-dextran-modified mitochondria, 2 µg of isolated mitochondria was incubated with 2, 6 or 10 µg of 1 µg/µL TAT-dextran solution at room temperature for 5 minutes in HB containing a proteinase inhibitor mixture up to 20 µL, immediately before use. Mitochondrial transfer was performed by co-incubating isolated mitochondria with the cells, as described previously. 4 Briefly, 2 µg of isolated mitochondria was added to 2 × 10 4 cells per well of 24-well culture plates in 400 µL of culture medium containing 116.6 mg/L CaCl 2 by reverse method and co-incubated at 37°C in a humidified atmosphere with 5% CO 2 . In the experimental model of OS, 2 µg of control or modified mitochondria was added to each well after induction of OS.

| Flow cytometric analysis
Mitochondrial transfer was confirmed by fluorescence microscopy and fluorescence-activated cell sorting (FACS) analyses as described previously. 4 Briefly, after washing the cells extensively with PBS, fluorescent images were captured using an IX71 fluorescence microscope (Olympus, Olympus, Tokyo, Japan) or BIOREVO BZ-9000 fluorescence microscope (Keyence, Osaka, Japan). The DsRedpositive cell population was evaluated by FACS analysis using 488 and 561 nm laser lines. Fluorescence intensity data were collected using SH800 (Sony) and analysed using FlowJo software (TreeStar).

| Comparison of mitochondrial transfer efficiency between co-culture and co-incubation methods
In the co-culture method, GFP-Mito EMCs were treated with H 2 O 2 for 6 hours and washed twice with PBS. Following this, fresh medium was added containing DsRed-Mito MSCs. The two types of cells, GFP-Mito EMCs and DsRed-Mito MSCs, were co-cultured with each cell number of 5 × 10 5 in 10-cm culture dish, whereas, in the co-incubation method, mitochondria isolated from 5 × 10 5 cells of DsRed-Mito MSC were added to 5 × 10 5 of GFP-Mito EMCs in a culture dish. Subsequently, 24 hours after co-culture or co-incubation, the percentage of MSCs containing DsRed-Mito was calculated.

| Statistical analysis
All data are expressed as mean ± SEM. Statistical significance between two groups was determined by unpaired Student's t test using the Microsoft Office Excel analysis tool and GraphPad Prism 6 software (GraphPad Software, San Diego, CA, USA). Results with a value of P < .05 were considered as statistically significant.  Figure   S2A). Conversely, dextran is a water-soluble, naturally occurring polysaccharide with multiple hydroxyl groups applicable for chemical modification. In this study, the conventional reductive amination method was performed to conjugate TAT peptides to dextran, named as TAT-dextran ( Figure 1A). We previously reported that isolated exogenous DsRed-labelled mitochondria could be taken up by other cells through co-incubation. 4 To increase the efficacy of isolated mitochondrial uptake into cells via co-incubation, we utilized TAT-dextran to coat isolated mitochondria ( Figure S2B). To determine whether TAT-dextran adversely affected mitochondrial respiratory function, respirometry was used to examine isolated mitochondria mixed with TAT-dextran. State3, State4 and the ratio of State3 to State4 were almost the same between isolated mitochondria and TAT-dextran-coated mitochondria ( Figure S3). First, we evaluated the mitochondrial uptake efficacy in H9c2 rat cardiomyoblast cells expressing mitochondria-targeted DsRed. We modi-  Figure 1G). Therefore, we determined the concentration ratio of TAT-dextran and mitochondria to be 5:1. It was demonstrated that TAT-dextran-coated mitochondria fused with endogenous mitochondria in the combination of EMC with genetically GFP-marked mitochondria as recipient cells and DsRed-labelled mitochondria at 6 hours after co-incubation ( Figure S4). As isolated mitochondria fused with endogenous mitochondria, it suggested that TAT-dextran might not inhibit fusions, which might be a mechanism to rescue damaged mitochondria in addition to replenishing respiratory chain complexes. Using the same labelling combination to estimate the amount of exogenous mitochondria replacing endogenous mitochondria in recipient cells, the ratio of DsRed versus GFP was measured ( Figure S5A,B). Internalization in TAT-dextran-coated mitochondria reached to about 30% of the endogenous mitochondria, which was significantly higher than that in isolated mitochondria, based on quantification of fluorescent areas ( Figure S5C).

| Enhanced cellular uptake of isolated mitochondria into NRCMs by TAT-dextran
We examined whether TAT-dextran modification of isolated mitochondria enhanced mitochondrial uptake in primary cultured neonatal rat cardiomyocytes (NRCMs). We demonstrated that 24 hours of coincubation of NRCMs and DsRed-labelled isolated mitochondria from EMCs resulted in efficient uptake by using TAT-dextran (Figure 2A).
We confirmed that TAT-dextran modification significantly increased the internalization of exogenous mitochondria into NRCMs by flow cytometric analysis (182.8 ± 18.4%, P < .01 versus Mito group, n = 3) ( Figure 2B,C). Additionally, we determined that DsRed2-labelled mitochondria were localized in the cytoplasm of NRCMs using threedimensional reconstructed images. The number of DsRed2-labelled mitochondria in the cytoplasm seemed to be higher in the TAT-Mito group of cells than in the Mito group of cells ( Figure 2D). Although cytotoxicity is a concern using TAT-dextran, there were no significant cytotoxic effects observed in NRCMs as a result of TAT-dextranmodified isolated mitochondria in this study ( Figure 2E).

| Mitochondrial transfer efficiency is higher in co-incubation method compared to cell-to-cell interaction
Tunnelling nanotube (TNT) is known as a bridge to transport intracellular organelles, including mitochondria, to other cells. 2 Indeed, it has been previously reported that mitochondrial transfer via TNT recovered damaged cells when co-cultured with healthy cells. 6,38 Studies have suggested that mitochondrial transfer via TNT was probably induced by damage-associated molecular pattern molecules (DAMPs) such as mitochondrial DNA (mtDNA). 39,40 Therefore, to evaluate the efficiency of mitochondrial transfer between co-culture method using two cell types and co-incubation of cells with isolated mitochondria, two fluorescent markers GFP and DsRed that are genetically expressed in mitochondria were used for comparison. In co-culture conditions, we cultured mesenchymal cells (MSCs) that were genetically modified with mitochondria-targeted DsRed and damaged EMCs that were genetically modified with mitochondria-targeted GFP in an OS model

| Bidirectional intracellular mitochondrial transfer and isolated mitochondria concomitantly with DAMPs induce necroptosis-like cell death
Necroptotic cell death mediated via receptor-interacting protein kinase 3 is a form of regulated necrosis, 41 which plays an essential role in IRI. In contrast to apoptotic cell death, necroptosis appears as a cellular burst releasing cellular constituents into the extracellular space, which was observed in damaged EMCs either with or without DAMPs ( Figure 4A,B, Video S1). Surprisingly, necroptosis-like cellular bursts of MSCs, which had been initially considered as donor cells of healthy mitochondria, were observed after contact with damaged EMCs with DAMPs ( Figure 4C,D, Video S2). We detected that reverse transport of damaged mitochondria from unhealthy EMCs to healthy MSCs with DAMPs via TNT, resulted in a necroptosis-like appearance ( Figure 4E

| Mitochondrial transfer significantly suppresses expression of apoptotic-related genes in NRCMs after in vitro OS induction
We examined the effect of TAT-dextran and isolated mitochondria on apoptosis in an in vitro NRCM model of OS ( Figure 5A). After induction of OS, the expression levels of apoptosis-promoting gene

| Mitochondrial transfer might rescue mitochondrial metabolic function in NRCMs after in vitro OS induction
To evaluate the effect of TAT-dextran along with isolated mitochondria on mitochondrial metabolic function in NRCMs after induction of OS, we measured mitochondrial oxygen consumption using the Oroboros Oxygraph-2k ( Figure 6A). After induction of OS, routine, ETS and free routine activity of NRCMs were significantly reduced compared with the control group (P < .05, P < .01 and P < .05, respectively, n = 3) ( Figure 6B). Mito significantly improved routine, ETS and free routine activity compared to the OS group (P < .05, P < .05 and P < .01, respectively, n = 3) ( Figure 6B). Specifically, routine and free routine activities in the Mito group showed levels comparable to the control group. Proton leakage in the Mito group also increased significantly, compared with the OS group (P < .05, n = 3), which could be due to restoration of ATP production ( Figure 6B).
TAT-Mito group also showed similar results as the Mito group. Mito and TAT-Mito cells recovered from OS-induced damages; however, they did not show significant changes. Only TAT-Mito significantly improved R/E compared to the OS group ( Figure 6C).

| Significant rescue of cardiomyocytes in NRCMs from OS-induced apoptosis by TAT-Mito
Transactivator of transcription-dextran with isolated mitochondria did not significantly enhance the anti-apoptotic pathway and metabolic recovery, compared to simple co-incubation with isolated mitochondria (Figures 5 and 6). We speculated that it could attrib-  Mitochondria play roles as not only energy generators, but also regulators of apoptosis and hubs for innate immunity. 48 A key mediator of the intrinsic apoptotic pathway is cytochrome c, which is re- Although it has been reported that intercellular mitochondrial transport via tunnelling nanotubes (TNT) and exosomes could rescue the damages in several disease models, 6,38 our study suggested that direct mitochondrial transfer might be more efficient for cellular uptake ( Figure 3C) and that TAT-dextran modification significantly enhanced mitochondrial transfer (Figure 2A-C). There is a recent This study demonstrated that efficient mitochondrial transfer was feasible in primary cardiomyocytes and exogenous mitochondria could function so as to alter the physiology of recipient cells.

| D ISCUSS I ON
Although studies in cell lines have reported that direct mitochondrial transfer even with poor transfer efficiency has the potential to rescue damaged cells, the methodology described in this study to transfer mitochondria into primary cells with high efficacy could provide the basis for animal studies, and probably for clinical evaluation. In conclusion, our data support direct transfer of TAT-Mito as a promising approach for the treatment of IRI.

ACK N OWLED G EM ENTS
We would like to express our sincere thanks to Ms Sayuri Shikata for her technical assistance. This research was supported by Grant-in-Aid for Challenging Exploratory Research 26670586, from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

AUTH O R CO NTR I B UTI O N S
HM, DK, RM, YM, JJ, TK and SG designed and performed the experiments. HM analysed and interpreted the data. HM, DK and SG prepared the manuscript. SM, YT and SG provided technical support, discussion and reviewed the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.