Cardioprotection by remote ischemic preconditioning of the rat heart is mediated by extracellular vesicles

DEV Elsevier Editorial System for Journal of Molecular and Cellular Cardiology Manuscript Draft Manuscript Number: JMCC7992R1 Title: Cardioprotection by remote ischemic preconditioning of the rat heart is mediated by extracellular vesicles Article Type: Rapid Communication Keywords: exosomes, microvesicles, microparticles, extracellular vesicles, remote conditioning, cardioprotection, ischemia-reperfusion Corresponding Author: Prof. Peter Ferdinandy, MD, PhD, MBA Corresponding Author's Institution: Semmelweis University First Author: Zoltán Giricz, PhD Order of Authors: Zoltán Giricz, PhD; Zoltán V Varga, MD; Péter Sipos, PhD; Krisztina Pálóczi, BSc; Ágnes Kittel, PhD, DSc; Edit Buzás, MD, PhD, DSc; Peter Ferdinandy, MD, PhD, DSc Abstract: Remote ischemic preconditioning (RIPC) of the heart is exerted by brief ischemic insults affected on a remote organ or a remote area of the heart before a sustained cardiac ischemia. To date, little is known about the inter-organ transfer mechanisms of cardioprotection by RIPC. Exosomes and microvesicles/microparticles are vesicles of 30-100 nm and 100-1000 nm in diameter, respectively (collectively termed extracellular vesicles [EVs]). Their content of proteins, mRNAs and microRNAs, render EVs ideal conveyors of inter-organ communication. However, whether EVs are involved in RIPC, is unknown. Therefore, here we investigated whether (1) IPC induces release of EVs from the heart, and (2) EVs are necessary for cardioprotection by RIPC. Hearts of male Wistar rats were isolated and perfused in Langendorff mode. A group of donor hearts was exposed to 3x5-5 min global ischemia and reperfusion (IPC) or 30 min aerobic perfusion, while coronary perfusates were collected. Coronary perfusates of these hearts were given to another set of recipient isolated hearts. A group of recipient hearts received IPC effluent depleted of EVs by differential ultracentrifugation. Infarct size was determined after 30 min global ischemia and 120 min reperfusion. The presence or absence of EVs in perfusates was confirmed by dynamic light scattering, the EV marker HSP60 Western blot, and electron microscopy. IPC markedly increased EV release from the heart as assessed by HSP60. Administration of coronary perfusate from IPC donor hearts attenuated infarct size in non-preconditioned recipient hearts (12.9±1,6% vs. 25.0±2.7%), similarly to cardioprotection afforded by IPC (7.3±2.7% vs. 22.1±2.9%) on the donor hearts. Perfusates of IPC hearts depleted of EVs failed to exert cardioprotection in recipient hearts (22.0±2.3%). This is the first demonstration that EVs released from the heart after IPC are necessary for cardioprotection by RIPC, evidencing the importance of vesicular transfer mechanisms in remote cardioprotection.

Answer to reviewers, JMCC7992 revision 1

Elizabeth Murphy, Ph.D. Associate Editor, JMCC RE: Answer to the Editor and the reviewers: Giricz et al, JMCC7992 Dear Dr. Murphy, Please find below itemized response to the editor and the reviewers. : Comment of the Editor: Your manuscript, "Remote ischemic preconditioning is mediated by extracellular vesicles," submitted for publication in the Journal of Molecular and Cellular Cardiology, has been read by expert reviewers. In its present form the manuscript is not acceptable for publication. Although the reviewers commented favourably on your manuscript, there were significant criticisms that preclude publication. In a revised manuscript it would be necessary to document (by immunoblot) the presence of exosomes in effluent collected from preconditioned hearts and the complete absence of exosomes in exosome-depleted and control perfusates. We would be willing to reconsider the manuscript after it has undergone a major revision that takes into account the criticisms of the reviewers. I will then return your revised manuscript to both reviewers for re-evaluation, with no assurance of acceptance. Answer to Editor: We greatly acknowledge the helpful comments of the editor and the reviewers. Accordingly, we have performed additional Western blots of the EV marker HSP60 and showed a marked increase of exosomes in effluent collected from preconditioned hearts as compared to the control perfusate where significantly lower amount of HSP60 marker was found. We have also confirmed the absence of exosomes in exosome-depleted perfusate. Moreover, we have supplemented the MS with online supplementary material to provide the readers of JMCC with more details of the methods, and revised the text of the MS, especially the discussion, as well as modified the title of the MS according to the comments of the reviewers. We indicated the major changes in the text by underlining. We hope that the revised MS will be accepted as a rapid communication. We wish the editor and the reviewers Happy Holidays. Best regards, Peter Ferdinandy

Covering letter and Author Response (for revisions)

Answer to reviewers, JMCC7992 revision 1

ANSWERS TO REVIEWER 1: Major Comment 1: Giricz and colleagues report that reduction of infarct size with remote ischemic preconditioning (RIPC), triggered by transfer of coronary effluent among isolated buffer-perfused rat hearts, is mediated by release and transport of extracellular vesicles. While these data are intriguing, two fundamental issues must be addressed: Is the release of exosomes unique to the isolated buffer-perfused heart model: i.e., possibly a consequence of the high flow conditions in this preparation? Efforts at clinical translation of RIPC involve in vivo skeletal muscle ischemia (rather than transfer of perfusate among hearts) as the remote stimulus; accordingly, the manuscript would be substantially strengthened by evidence of exosome release into blood following repeated brief episodes of hindlimb ischemia-reperfusion. Answer 1: We greatly acknowledge the reviewer to point out an important issue, which definitely needs to overcome many technical issues in EV research in general. Release of extracellular vesicles (EVs) has been confirmed from virtually any kind of cell types, including C2C12 myoblasts (Exp Cell Res. 2010;316(12):1977-84; FEBS Lett. 2013;587(9):1379-84), or cardiac myocytes (Am J Physiol Heart Circ Physiol. 2013;304(7):H954-65). Furthermore, various pathological conditions induce EV release from myocytes, amongst others, e.g. ischemia (J Mol Cell Cardiol. 2012;53(6):848-57). Similarly, in Fig 2A we show that although there is a basal release of EVs, preconditioning significantly increases it. Therefore, it is plausible that EV release is not uniquely characteristic to isolated heart preparations and that it is not induced by the elevated coronary flow. We agree that it would be more relevant to clinical settings to assess whether hindlimb ischemia induces EV release and whether released EVs convey cardioprotective signals. However, many technical difficulties would have to be tackled to answer these questions. The lack of technology so far to study the biological function of EVs in more details is possibly due to the fact that this field of research is relatively novel. For instance, identifying the EVs tissue of origin is not yet solved in in-vivo settings. Unfortunately, there are no any markers identified for muscle-originated EVs. For instance, Malik et al. (Am J Physiol Heart Circ Physiol. 2013;304(7):H954-65) reported several muscle-specific proteins in exosomes isolated from supernatant of carcdiac myocytes, however, the same proteins can be found in various, non-muscle derived exosomes as well (e.g. tropomyosin, http://microvesicles.org/gene_summary?gene_id=7168). Moreover, one can also speculate that from an ischemic limb EVs are released not only from ischemic myocytes but for example from platelets or endothelial cells etc. In this case, identifying the source of ischemia-induced EV release is even more complicated. Protein amount (specific or total) of the EV pellet as a surrogate measure for increased efflux could unlikely serve as a valuable parameter in an in-vivo model given that a high amount of EVs circulate in the blood and the expected release of EVs could be marginal compared to the total pool. Taken together, we believe that adding in-vivo data to this MS would highly exceed the scope of the current project

Answer to reviewers, JMCC7992 revision 1

where we provided the first evidence that transmission of RIPC requires EVs at least in the ex-vivo perfused heart. Major Comment 2. Please provide evidence of: (i) the presence of exosomes in effluent collected from preconditioned hearts and transferred to naïve hearts; and, of equal importance; (ii) the complete absence of exosomes in exosome-depleted perfusate; and (iii) absence of exosomes in control perfusate. This can be achieved by immunblotting for one or more of the protein constituents of exosomes as described by Malik et al, Am J Physiol 2013;304:H954-65. Answer 2. We thank for the reviewer to point out these important issues. Accordingly, we have performed a Western blot experiment on vesicular pellets isolated from preconditioned and control perfusates and the corresponding vesicle-depleted supernatants and detected the EV marker HSP60 signals as indicated in the revised MS at line 139-140. As the newly added Fig2A shows, HSP60 is present in the vesicular pellets of preconditioned effluent and in significantly lower amount in control samples, however, no HSP60 was detected in vesicle-depleted supernatants of these samples. These data together with EM images clearly demonstrate that (i) preconditioning induces vesicle release from the heart and that (ii) we were able to completely deplete vesicles from perfusate samples. Although we could not evidence a complete absence of vesicles in control samples, but since HSP60 signal was markedly weaker in control samples, together with the above mentioned results, it is highly plausible that the preconditioning stimulus is responsible for the release of cardioprotective signals which are transferred by EVs. Minor comments: Comment 3. The first report of remote conditioning via transfer of coronary effluent, published by Dickson et al, should be cited (Am J Physiol 1999; 277:H2451-7). Answer 3: This reference has already been included in the MS (ref. no. 8). Comment 4. Page 7: "One way analysis of variance (ANOVA) was used to evaluate differences in cell transfection experiments". As these experiments do not involve transfection, this is presumably a typo - please correct. Answer 4: We amended the MS accordingly.

Answer to reviewers, JMCC7992 revision 1

ANSWERS TO REVIEWER #2: Major Comment 1: Exosomes and micro-particles are indeed candidates of mediators of various types of remote preconditioning and the present study provides a piece of supportive evidence. However, there are a few problems in this study, which make the authors' argument unconvincing. Specific problems are as follows. Preparation of EVs for testing their effects on infarct size is not clearly described. The authors described how exosomes and microparticles were obtained as pellets after centrifugation at different conditions (i.e., 100,000xg vs, 12,200xg). However, how exosomes and microparticles were combined (?) and diluted is not stated in Methods. It is also unclear how "the EV depleted perfusate" was reconstituted for infarct size experiments. Answer 1: Thank you for pointing out these issues that was not clear enough in the original MS. We have not combined exosome and microvesicle pellets. For EM, pellets were used separately. However, for the additional experiments (i.e. EV marker HSP60 Western blots) we did not separate exosomes from microvesicles, as indicated in the Supplementary data, Methods section. EV-depleted perfusates were diluted to their original volume (as measured at the time of collection from the heart) by adding Krebs-Henseleit solution to the depleted perfusates (e.g. in case 200 mL perfusate was collected, 40 mL of depleted perfusate was combined with 160 mL Krebs-Henseleit solution). To clarify our methods, we amended the relevant section of Methods within the Supplementary data, due to restrictions on word count. Major Comment 2. It is not clear whether concentration of EVs in vivo is sufficient to afford protection. Since there is no information regarding concentration of EVs in circulating blood vivo and their concentration in the perfusate used for infarct size experiments, it is not possible to discuss whether EVs play roles in remote preconditioning in vivo. Answer 2: The reviewer pointed out an important issue that definitely needs to overcome many technical issues in EV research in general, as detailed in answer to Comment 1 of Reviewer 1 above. Given the lack of technology to identify muscle-originated EVs, we can speculate that we collected EVs from 1g of heart tissue for only 15 min, whereas in-vivo EVs might stay in the circulation for longer times (although the half life of EVs are also not known), thus forming a circulating pool. Given that a hindlimb ischemia in-vivo affects a larger amount of tissue including components of blood, we can speculate that the amount of cardioprotective EVs released from an ischemic hindlimb can be higher than what we detected in our isolated heart experiments. However, to prove this, enormous technical difficulties have to be solved, as detailed in our response to major question 1 of Reviewer 1. The lack of technology to study the biological function of EVs in more details is possibly due to the fact that this field of research is relatively novel. Nevertheless, we amended the Discussion with the following sentences to point out these limitations in lines 190 – 193 “Ischemia-induced release of EVs from cultured cardiomyocytes was reported by Malik et al [12] recently, which is in agreement with our current findings that EV-release of isolated hearts increases after brief ischemic episodes.”

Answer to reviewers, JMCC7992 revision 1

and in lines 203 – 204 “Although these findings might suggest that our present ex-vivo results might not translate to in-vivo models, however, it also suggests that exosomes rather than microvesicles are responsible for the propagation of cardioprotective signals.” Comment 3. It is not clear whether contents of EVs indeed matters for cardioprotection. In other words, EV depleted perfusate is not the only control. Do inert microparticles lack any cardioprotective effect? In this regards, it is important to present data on coronary flow and ventricular pressure during infusion of EVs. Answer 3. We agree with the reviewer that it is not known what substance in EVs may be responsible for cardioprotection as also discussed in the MS. Unfortunately, given the size range of the EVs (30-1000 nm), it would be difficult to use “inert microparticles” with the same distribution in size. Moreover, no technology is available for the depletion of the contents of EVs. The size of the EVs are a magnitude smaller than the diameter of the microvessels, therefore, EVs cannot produce e.g mechanical microembolisation that might interfere with ischemia and cardioprotection. Accordingly, there was no effect of EV containing perfusate on coronary flow (see online supplementary material). It has been evidenced is several reports that EVs of various origin bear cardioprotective effects (Biochem Biophys Res Commun. 2013;431(3):566-71.; Stem Cell Res. 2010;4(3):214-22.). Based on these data, together with our current results showing that EV-depleted perfusates fail to reduce infarct size, we can assume that EV’s content, or substances associated with EVs are at least in part responsible for cardioprotection. Therefore, we amended the Discussion section as follows on line 197 – 200: “Since in the latter two reports EVs from untreated cells induced pro-survival signals, based on our current findings, we cannot exclude the possibility that EVs released from the heart under basal conditions might be also cardioprotective, would their amount be as high as after preconditioning stimuli.” As requested, we included a supplementary table (Table S1) of coronary flow data showing that there is no significant difference between CON PERF, PRE PERF, or DEPL PERF groups in the given timepoints throughout the experimental protocol, only PRE (preconditioned perfusate-donor hearts) showed elevated coronary perfusion after preconditioning stimuli, which is in accordance with the literature (Circulation. 1995; 91: 2810-2818). However, in our experiment we did not employ LV balloons to document left ventricular pressure changes. We focused on measurement of infarct size, since this is the gold standard end-point to study ischemia/reperfusion injury and cardioprotection in ex-vivo isolated hearts or in-vivo models of infarction. Comment 4. A word, "remote preconditioning" should be used more specifically. There are several types of remote preconditioning (i.e., preconditioning with cardiac, renal, intestinal or limb ischemia), and kinds and amount of EVs may not be the same depending on the organ used for preconditioning. Answer 4. We agree that different organs may release different amount of EVs with different constituents. We chose the isolated heart as a source of EV-containing perfusion fluid since isolated perfused organs are more suitable for such experiments

Answer to reviewers, JMCC7992 revision 1

(as outlined in our response to major question 1 of Reviewer#1) and since we are experienced in isolated heart perfusions. We emphasized throughout the text that remote conditioning has been performed here by cardiac ischemia, and we modified the title of the revised MS accordingly. Minor comment: It is not clear which post hoc test was used when ANOVA indicated an overall difference.

Answer to minor comment:

We used Fisher LSD method as a post-hoc test in our ANOVA calculations, as

indicated in line 145.

Giricz et al, JMCC7992 revision 1

Cardioprotection by remote ischemic preconditioning of the rat heart is 1

mediated by extracellular vesicles 2

3

Zoltán Giricz1, Zoltán V. Varga1, Péter Sipos2, Krisztina Pálóczi3, Ágnes Kittel4, 4

Edit Buzás3, Péter Ferdinandy1,5 5

1Cardiometabolic Research Group, Department of Pharmacology and 6

Pharmacotherapy, Semmelweis University, Budapest, Hungary; 2Department of 7

Pharmaceutical Technology, University of Szeged, Szeged, Hungary, 3Department 8

of Genetics, Cell- and Immunobiology, Semmelweis University, Budapest, Hungary; 9

4Department of Pharmacology, Institute of Experimental Medicine, Hungarian 10

Academy of Sciences, Budapest, Hungary; 5Cardiovascular Research Group, 11

Department of Biochemistry, University of Szeged, Szeged, Hungary 12

13

14

15

Please address correspondence to: Péter Ferdinandy, MD, PhD 16

Cardiometabolic Research Group, Department of Pharmacology and 17

Pharmacotherapy, Semmelweis University, Nagyvárad tér 4, Budapest, H-1089, 18

Hungary, Fax: +36 1 2104412; Phone: +36 1 2104416 19

E-Mail: [email protected] 20

21

22

*ManuscriptClick here to view linked References

Giricz et al, JMCC7992 revision 1

2

ABSTRACT 23

24

Remote ischemic preconditioning (RIPC) of the heart is exerted by brief 25

ischemic insults affected on a remote organ or a remote area of the heart before a 26

sustained cardiac ischemia. To date, little is known about the inter-organ transfer 27

mechanisms of cardioprotection by RIPC. Exosomes and 28

microvesicles/microparticles are vesicles of 30-100 nm and 100-1000 nm in 29

diameter, respectively (collectively termed extracellular vesicles [EVs]). Their content 30

of proteins, mRNAs and microRNAs, render EVs ideal conveyors of inter-organ 31

communication. However, whether EVs are involved in RIPC, is unknown. 32

Therefore, here we investigated whether (1) IPC induces release of EVs from 33

the heart, and (2) EVs are necessary for cardioprotection by RIPC. 34

Hearts of male Wistar rats were isolated and perfused in Langendorff mode. A 35

group of donor hearts was exposed to 3x5-5 min global ischemia and reperfusion 36

(IPC) or 30 min aerobic perfusion, while coronary perfusates were collected. 37

Coronary perfusates of these hearts were given to another set of recipient isolated 38

hearts. A group of recipient hearts received IPC effluent depleted of EVs by 39

differential ultracentrifugation. Infarct size was determined after 30 min global 40

ischemia and 120 min reperfusion. The presence or absence of EVs in perfusates 41

was confirmed by dynamic light scattering, the EV marker HSP60 Western blot, and 42

electron microscopy. 43

IPC markedly increased EV release from the heart as assessed by HSP60. 44

Administration of coronary perfusate from IPC donor hearts attenuated infarct size in 45

non-preconditioned recipient hearts (12.9±1,6% vs. 25.0±2.7%), similarly to 46

cardioprotection afforded by IPC (7.3±2.7% vs. 22.1±2.9%) on the donor hearts. 47

Giricz et al, JMCC7992 revision 1

3

Perfusates of IPC hearts depleted of EVs failed to exert cardioprotection in recipient 48

hearts (22.0±2.3%). 49

This is the first demonstration that EVs released from the heart after IPC are 50

necessary for cardioprotection by RIPC, evidencing the importance of vesicular 51

transfer mechanisms in remote cardioprotection. 52

53

Keywords: exosomes, microvesicles, microparticles, extracellular vesicles, 54

remote conditioning, cardioprotection, ischemia-reperfusion 55

56

57

Highlights: 58

ischemic preconditioning (IPC) of hearts induces exosome & microvesicle 59

release 60

coronary perfusate collected during IPC reduced infarct size of recipient 61

hearts 62

perfusate depleted of exosomes & microvesicles does not protect recipient 63

hearts 64

this is the first demonstration that exosomes & microvesicles mediate remote 65

IPC 66

67

68

69

70

71

Giricz et al, JMCC7992 revision 1

4

1. INTRODUCTION 72

Remote ischemic conditioning (RIPC), where a remote area of the heart or 73

another organ is submitted to brief cycles of ischemia-reperfusion, protects the heart 74

against a lethal ischemic insult with efficiency comparable to that of classic in-situ 75

ischemic protocols [1,2]. Although effector pathways of RIPC have been well 76

described, it is currently unclear how cardioprotective signals are propagated 77

between organs [3]. Humoral and neuronal aspects have been hypothesized, but 78

vesicular transfer mechanisms have not been evidenced in inter- or intra-organ 79

communication of RIPC signals. 80

Exosomes and microvesicles/microparticles (collectively termed extracellular 81

vesicles, EVs) are membrane-bound structures secreted by a wide range of 82

mammalian cell types via distinct mechanisms [4,5]. Since EVs contain a high 83

concentration of RNAs and proteins, and since EVs can be secreted and specifically 84

taken up by other cells, they are prime medium for intercellular signal transfer 85

mechanisms [5]. Thus, it is not surprising that EVs have been shown to modulate 86

several essential cellular functions, including cell survival mechanisms [6,7]. 87

However, to date, it is not known whether EVs are involved in the transmission of 88

cardioprotective signals in ischemic conditioning maneuvers, particularly, their role in 89

the propagation of RIPC has never been studied. 90

Therefore, here we aimed to investigate whether the release of EVs from the 91

heart is induced by preconditioning stimuli; and to test if EVs are necessary for 92

RIPC-induced cardioprotection by assessing that RIPC can be exerted in the 93

presence and absence of EVs. 94

95

Giricz et al, JMCC7992 revision 1

5

2. MATERIALS AND METHODS 96

This investigation conforms to the Guide for the Care and Use of Laboratory 97

Animals published by the US National Institutes of Health (NIH publication No. 85-98

23, revised 1996) and was approved by the animal ethics committee of the 99

Semmelweis University, Budapest, Hungary. 100

101

2.1 Experimental setup, heart perfusion protocol, assessment of infarct size 102

Male Wistar rats (250-350g) were anesthetized by 85mg/kg ketamine and 103

10mg/kg xylazine and heparinized. Hearts were isolated and perfused in Langendorff 104

mode with 37°C Krebs-Henseleit solution for 20 min for stabilization; then hearts 105

were randomized to the following groups. Perfusate donor hearts either received 106

aerobic perfusion for an additional 30 min (CON) or were exposed to 3×5-5 min 107

ischemia and reperfusion (PRE). Perfusate recipient hearts were perfused with 108

collected perfusate from either CON or PRE hearts (CON PERF and PRE PERF, 109

respectively). Another group of hearts received perfusate of PRE hearts which had 110

been previously depleted of EVs (DEPL PERF). All hearts were then exposed to a 111

30 min global ischemia and 2h reperfusion (Figure 1). 112

Hearts then were cut into 6-8 slices, slices were weighed, and infarct size was 113

assessed by TTC-staining. Infarct size was expressed as a percentage of the total 114

heart weight. 115

116

2.2 Isolation of EVs, EV depletion 117

EVs were isolated from collected coronary perfusates by filtration and 118

differential centrifugation. Briefly, perfusates were dialyzed against 0.45% saline 119

containing 5mM EDTA for 4 h at room temperature then vacuum-distilled to 40mL. 120

Giricz et al, JMCC7992 revision 1

6

Concentrated perfusates were filtered through 800nm filter (Merck, Darmstadt, 121

Germany) and centrifuged at 12,200×g for 20 min at 4°C. Pellets were saved as 122

microvesicle/microparticle fraction. Then supernatants were filtered through 200nm 123

filter (Merck, Darmstadt, Germany) and centrifuged at 100,000×g for 90 min at 4°C. 124

Pellets were saved as exosome-rich pellet and the supernatant was saved as EV-125

depleted perfusate. EV-depleted perfusates were then reconstituted to their original 126

volume with Krebs-Henseleit solution and used in heart perfusion experiments. 127

128

2.3 Characterization and assessment of quantity and size distribution of EVs 129

Isolated vesicles were visualized by transmission electronmicroscopy. Vesicle 130

pellets were fixed with 4% formaldehyde, postfixed in 1% OsO4 (Taab Laboratories, 131

Aldermaston, UK). EVs were dehydrated in graded ethanol, block-stained with 1% 132

uranyl acetate in 50% ethanol, and embedded in Taab 812 (Taab Laboratories, 133

Aldermaston, UK). Ultrathin sections were cut and then analyzed with a Hitachi 7100 134

electron microscope. 135

Hydrodynamical average particle size of EVs in perfusates was measured by 136

Dynamic Light Scattering (DLS) apparatus Zetasizer Nano ZS (Malvern Instruments, 137

Malvern Hills, UK) (n = 3-4). 138

The presence and amount of EVs was assessed by HSP60 immunoblots from 139

vesicular pellets and EV-depleted perfusates. 140

For a detailed methods section please see Supplementary data online. 141

142

2.4 Statistical analysis 143

Giricz et al, JMCC7992 revision 1

7

Values are expressed as mean±SEM. One way analysis of variance (ANOVA) 144

followed by Fisher LSD post-hoc test was used to determine differences in infarct 145

size. 146

147

Giricz et al, JMCC7992 revision 1

3. RESULTS 148

3.1 Ischemic conditioning increase the amount of EVs released into coronary 149

perfusate 150

Coronary perfusates from preconditioned hearts (PRE) contained more EVs 151

than perfusates isolated from control hearts (CON) as evidenced by Western blot 152

against HSP60, a well-accepted marker of EVs (Figure 2A). Electron micrographs 153

revealed that these EVs can be classified as microvesicle/microparticles as defined 154

by a diameter of >100nm and light vesicular structure (Figure 2B upper panels), and 155

exosomes of <100nm in diameter (Figure 2B lower panel). 156

To further evidence that size of the particulate matter in coronary perfusate 157

conforms to the range of microvesicles and exosomes, we assessed their size 158

distribution by DLS in perfusates from PRE hearts. As the representative diagram 159

(Figure 2C) shows, three populations of particles could be distinguished in our 160

samples. Beside a small fraction of particles with hypothetical diameter of 161

approximately 10nm, the main particulate constituents fell into the size range of 162

exosomes (<100 nm) and microvesicles (100-1000 nm). 163

164

3.2 Perfusates depleted of EVs did not decrease infarct size 165

Isolated hearts that received 3×5-5 min ischemia and reperfusion (PRE) 166

before 30 min global ischemia demonstrated a significantly reduced infarct size as 167

compared to aerobically perfused hearts (CON). In hearts perfused with coronary 168

perfusates collected from PRE hearts (PRE PERF) infarct size was significantly 169

lower than in hearts perfused with coronary perfusate from CON hearts (CON 170

PERF). Perfusates of PERF hearts which had been depleted of EVs was also given 171

to recipient hearts (DEPL PERF). Infarct size after 30 min ischemia and 2h 172

Giricz et al, JMCC7992 revision 1

9

reperfusion in DEPL PERF hearts did not differ significantly from the extent of 173

infarction observed in CON PERF hearts (Figure 2D). 174

Giricz et al, JMCC7992 revision 1

4. DISCUSSION 175

We have shown here for the first time in the literature that the release of EVs 176

from the heart after preconditioning stimuli is increased and that EVs are responsible 177

for the transmission of remote conditioning signals for cardioprotection. 178

Previously several humoral and neuronal transmitter mechanisms have been 179

hypothesized to play a role in the propagation of remote ischemic conditioning, 180

however, to date none of them is generally accepted. Dickson et al. have proposed 181

first the involvement of humoral transmission pathways showing that transfusion of 182

blood from preconditioned rabbits confers cardioprotection in a naïve non-183

preconditioned animal against ischemia/reperfusion injury [8]. Later, Breivik et al. 184

showed that the soluble factor is likely to be hydrophobic [9]. The role of neuronal 185

pathways has also been studied, but results are also still controversial [10,11]. 186

Here we evidence a novel, vesicular mechanism for the transmission of 187

cardioprotective signals from a preconditioned heart to another heart subjected to 188

coronary occlusion and reperfusion, which might explain how the suspected humoral 189

and/or released neuronal factors of remote conditioning are transmitted. Ischemia-190

induced release of EVs from cultured cardiomyocytes was reported by Malik et al 191

[12] recently, which is in agreement with our current findings that EV-release of 192

isolated hearts increases after brief ischemic episodes. Elsewhere, exosomes 193

derived from mesenchymal stem cell cultures have been shown to exert 194

cardioprotection in mice [13], and microvesicles isolated from cell culture medium of 195

GATA-4-overexpressing bone marrow stem cells protected neonatal cardiomyocytes 196

from ischemic injuries [14]. Since in the latter two reports EVs from untreated cells 197

induced pro-survival signals, based on our current findings, we cannot exclude the 198

possibility that EVs released from the heart under basal conditions might be also 199

Giricz et al, JMCC7992 revision 1

11

cardioprotective, would their amount be as high as after preconditioning stimuli. 200

Seemingly controversial to our findings, microvesicles derived from blood of animals 201

underwent hind limb ischemia/reperfusion failed to decrease infarct size in rats [15]. 202

Although these findings might suggest that our present ex-vivo results might not 203

translate to in-vivo models, however, it also suggests that exosomes rather than 204

microvesicles are responsible for the propagation of cardioprotective signals. 205

In summary, this is the first demonstration that EVs are the carrier mechanism 206

of the cardioprotective effect of RIPC of the heart, although, further molecular and in-207

vivo experimentation is warranted to decipher the nature of actual effector factors 208

carried by these vesicles. 209

Giricz et al, JMCC7992 revision 1

210

5. SOURCES OF FUNDING 211

This work was supported by the following grants: Hungarian Scientific Research 212

Fund (OTKA PD 109051, OTKA ANN 107803 and OTKA NK 84043) and FP7-213

PEOPLE-2011-ITN-PITN-GA-2011-289033 “DYNANO”. ZG. holds a „János Bolyai 214

Fellowship” from the Hungarian Academy of Sciences and ZVV was supported by 215

the National Program of Excellence (TAMOP 4.2.4.A/1-11-1-2012-0001). 216

217

6. ACKNOWLEDGEMENT 218

We are indebted to Dorottya Gódor, Krisztina Dobosi and Ibolya Benkes Jenőné 219

(Semmelweis University) for their skillful technical assistance. 220

221

7. DISCLOSURES/CONFLICT OF INTEREST 222

None. 223

224

Giricz et al, JMCC7992 revision 1

13

FIGURE LEGENDS 225

Figure 1. Experimental protocol of Langendorff-perfused rat hearts. CON: control; 226

PRE: preconditioned; PERF: perfused; DEPL: depleted. 227

228

Figure 2. A: HSP60 immunoblots of depleted perfusates (DEPL PERF), EV pellets of 229

preconditioned (PRE) and control perfusates (CON). B: Representative electron 230

micrographs of EVs in the size ranges of microvesicles and exosomes isolated from 231

coronary perfusate of preconditioned hearts Scale bar represents 400 nm. C: DLS 232

analysis of perfusates showing distribution of different vesicle populations. D: Infarct 233

size indicated as percent of the total heart volume. Results are expressed as 234

mean±SEM; n=5-8. *p<0.05 vs. CON, #p<0.05 vs. CON PERF. 235

236

Giricz et al, JMCC7992 revision 1

14

REFERENCES 237

[1] Przyklenk K, Bauer B, Ovize M, Kloner RA, Whittaker P. Regional ischemic 238

'preconditioning' protects remote virgin myocardium from subsequent sustained 239

coronary occlusion. Circulation 1993;87:893-9. 240

[2] Kharbanda RK, Mortensen UM, White PA, Kristiansen SB, Schmidt MR, 241

Hoschtitzky JA et al. Transient limb ischemia induces remote ischemic 242

preconditioning in vivo. Circulation 2002;106:2881-3. 243

[3] Hirsch E, Hilfiker-Kleiner D, Balligand JL, Tarone G, De Windt L, Bauersachs J et 244

al. Interaction of the heart and its close and distant neighbours: report of the Meeting 245

of the ESC Working Groups Myocardial Function and Cellular Biology. Cardiovasc 246

Res 2013;99:595-9. 247

[4] Keller S, Sanderson MP, Stoeck A, Altevogt P. Exosomes: from biogenesis and 248

secretion to biological function. Immunol Lett 2006;107:102-8. 249

[5] Gyorgy B, Szabo TG, Pasztoi M, Pal Z, Misjak P, Aradi B et al. Membrane 250

vesicles, current state-of-the-art: emerging role of extracellular vesicles. Cell Mol Life 251

Sci 2011;68:2667-88. 252

[6] Waldenstrom A, Genneback N, Hellman U, Ronquist G. Cardiomyocyte 253

microvesicles contain DNA/RNA and convey biological messages to target cells. 254

PLoS One 2012;7:e34653. 255

Giricz et al, JMCC7992 revision 1

15

[7] Qu JL, Qu XJ, Zhao MF, Teng YE, Zhang Y, Hou KZ et al. Gastric cancer 256

exosomes promote tumour cell proliferation through PI3K/Akt and MAPK/ERK 257

activation. Dig Liver Dis 2009;41:875-80. 258

[8] Dickson EW, Reinhardt CP, Renzi FP, Becker RC, Porcaro WA, Heard SO. 259

Ischemic preconditioning may be transferable via whole blood transfusion: 260

preliminary evidence. J Thromb Thrombolysis 1999;8:123-9. 261

[9] Breivik L, Helgeland E, Aarnes EK, Mrdalj J, Jonassen AK. Remote 262

postconditioning by humoral factors in effluent from ischemic preconditioned rat 263

hearts is mediated via PI3K/Akt-dependent cell-survival signaling at reperfusion. 264

Basic Res Cardiol 2011;106:135-45. 265

[10] Gho BC, Schoemaker RG, van den Doel MA, Duncker DJ, Verdouw PD. 266

Myocardial protection by brief ischemia in noncardiac tissue. Circulation 267

1996;94:2193-200. 268

[11] Weinbrenner C, Nelles M, Herzog N, Sarvary L, Strasser RH. Remote 269

preconditioning by infrarenal occlusion of the aorta protects the heart from infarction: 270

a newly identified non-neuronal but PKC-dependent pathway. Cardiovasc Res 271

2002;55:590-601. 272

[12] Malik ZA, Kott KS, Poe AJ, Kuo T, Chen L, Ferrara KW et al. Cardiac myocyte 273

exosomes: stability, HSP60, and proteomics. Am J Physiol Heart Circ Physiol 274

2013;304:H954-65. 275

Giricz et al, JMCC7992 revision 1

16

[13] Lai RC, Arslan F, Lee MM, Sze NS, Choo A, Chen TS et al. Exosome secreted 276

by MSC reduces myocardial ischemia/reperfusion injury. Stem Cell Res 2010;4:214-277

22. 278

[14] Yu B, Gong M, Wang Y, Millard RW, Pasha Z, Yang Y et al. Cardiomyocyte 279

protection by GATA-4 gene engineered mesenchymal stem cells is partially mediated 280

by translocation of miR-221 in microvesicles. PLoS One 2013;8:e73304. 281

[15] Jeanneteau J, Hibert P, Martinez MC, Tual-Chalot S, Tamareille S, Furber A et 282

al. Microparticle release in remote ischemic conditioning mechanism. Am J Physiol 283

Heart Circ Physiol 2012;303:H871-7. 284

285

Figure 1Click here to download high resolution image

Figure 2aClick here to download high resolution image

Figure 2bClick here to download high resolution image

Figure 2cClick here to download high resolution image

Figure 2dClick here to download high resolution image

Giricz et al, JMCC7992 Revision 1

Supplementray data for „Cardioprotection by remote ischemic

preconditioning of the rat heart is mediated by extracellular vesicles”

by Giricz Z et al.

Methods

Coronary flow measurement

Coronary outflow (CF) was collected from Langendorff-perfused hearts for 1 min before the

first ischemia, after each ischemic period, or time matched, and at the end of protocol (n=5-

9). CF volume was measured by a graduated cylinder.

Isolation of EVs, EV depletion

EVs were isolated from collected coronary perfusates by filtration and differential

centrifugation. Freshly prepared CFs were dialyzed in cellulose dialysis tube (Sigma) against

0.45% saline containing 5mM EDTA for 4 h at room temperature. CFs were then vacuum-

distilled to 40mL (for approximately 40 min) at room temperature. Concentrated CFs were

filtered through 800nm filter (Merck, Darmstadt, Germany) by gravity, and filtrates were

centrifuged at 12,200×g for 20 min at 4°C. Pellets were saved as microvesicle/microparticle

fraction. Supernatants of the previous centrifugation were filtered through 200nm filter

(Merck, Darmstadt, Germany) by gravity and centrifuged at 100,000×g for 90 min at 4°C in

Beckman MLA-55 rotors. Pellets were saved as exosome-rich pellet and the supernatant

was saved as EV-depleted perfusate. EV-depleted perfusates were then reconstituted to

their original volume measured at the time of CF collection with Krebs-Henseleit solution

(e.g. in case 200 mL CF was collected, 40 mL of depleted perfusate was combined with 160

mL Krebs-Henseleit solution). Reconstituted perfusates were used in heart perfusion

experiments same day.

Supplementary Material PDF version

Giricz et al, JMCC7992 Revision 1

Western blots

In order to demonstrate presence or absence of EVs in the perfusate of Langendorff-

perfused hearts, CF collected from CON and PRE hearts (n=3) were dialyzed and

concentrated as described above. After passing concentrated samples through 800nm filters

by gravity, a total EV pellet was prepared by centrifuging samples at 100,000×g for 90 min at

4°C in Beckman MLA-55 rotors. Supernatants of the centrifugation were kept as EV-depleted

perfusates. Total EV pellets were resuspended in 35µL RIPA buffer (VWR International,

Radnor, PA) containing protease inhibitor cocktail (Roche, Basel, Switzerland). RIPA

containing protease inhibitor was added to EV-depleted perfusates as well in 1/10 ratio.

Twenty µL of each sample was mixed with 5× Laemmli-buffer (Bio-Rad, Hercules, CA) and

loaded on 4-20% Tris-glycine SDS-polyacrylamide gels (Bio-Rad, Hercules, CA). Page Ruler

prestained protein ladder was also loaded to identify molecular weights (Pierce, Rockford,

IL). Samples were electrophoresed at 100V for 1.5 h and transferred to a PVDF membrane

(Bio-Rad, Hercules, CA) at 300 mV for 16h at 4°C. Even transfer of proteins was confirmed

by Ponceau S-staining. The membrane was blocked in blocking buffer containing 5% non-fat

milk (Bio-Rad, Hercules, CA) in phosphate-buffered saline +0.05% Tween-20 (PBS-T) for 2h

and then incubated with HSP60 antibody (4870, Cell Signaling, Danvers, MA) diluted in

blocking buffer for 16h at 4°C. After 3×15 min wash in PBS-T, the membrane was incubated

with HRP-conjugates anti-rabbit IgG (Cell Signaling, Danvers, MA) in blocking buffer for 2h at

room temperature and washed again in PBS-T for 3×15 min. Signals were visualized after

incubation in enhanced chemiluminescence kit (Pierce, Rockford, IL) for 5 min by Chemidoc

XRS+ (Bio-Rad, Hercules, CA).

Statistical analysis

Values are expressed as mean±SEM. One way analysis of variance (ANOVA) followed by

Fisher LSD test was used as post-hoc test.

Giricz et al, JMCC7992 Revision 1

Results

Coronary flow

CF rate was significantly elevated in PRE samples after each ischemic period compared to

time-matched values of non-preconditioned CON hearts. CF of perfusate-recipient CON

PERF, PRE-PERF and DEPL-PERF did not differ significantly from each other in any given

timepoint (Table S1).

Tables

Group CF 19'-20' CF 29'-30' CF 39'-40' CF 49'-50' CF 199'-200'

CON (n=8) 10.65±0.75 10.74±0.78 10.51±0.87 10.68±0.89 5.70±0.25

PRE (n=5) 11.33±0.33 18.33±0.67* 18.00±0.08* 18.33±0.58* 7.00±0.00

CON PERF (n=9) 10.09±0.35 9.78±0.37 9.19±1.01 7.76±1.06 6.01±1.39

PRE PERF(n=7) 9.07±0.73 10.10±1.37 9.03±1.06 9.29±1.02 4.93±0.46

DEPL PERF (n=7) 12.00±0.49 11.29±0.42 10.86±0.63 10.00±0.71 6.79±0.51

Supplementary table 1. Coronary flow (CF, mL/min) was measured at the end of

equilibration period (19'-20'), after each cycles of preconditioning ischemia or time-matched

normoxic perfusion (29'-30', 39'-40' and 49'-50'), and at the end of the 120–min

reperperfusion (199'-200'). *p<0.05 vs. time-matched controls (CON).