Studying the Innate Immune Response to Myocardial Infarction in a Highly Efficient Experimental Animal Model

Contact address: Razvan Gheorghita MARES, 38 Gheorghe Marinescu Street, 540142, Targu Mures, Romania. E-mail: razvan.mares@umfst.ro 1 „George Emil Palade” University of Medicine, Pharmacy, Science and Technology, Targu Mures, Romania 2 Center for Advanced Medical and Pharmaceutical Research, Targu Mures, Romania 3 Clinical County Hospital, Targu Mures, Romania 4 Emergency Clinical County Hospital, Targu Mures, Romania 5 Department of Clinical Sciences Malmö, Lund University, Sweden ABSTRACT The reduction in mortality following acute myocardial infarction (AMI) is an important achievement of modern medicine. Despite this progress, AMI remains the most common cause of heart failure (HF) and HF-related morbidity and mortality. The involvement of the innate immune response in different stages after AMI has at tracted important attention in recent years. With the increasing range of potential therapeutic compounds and delivery vectors, the need of highly effi cient experimental AMI models is increasing, to support further advancement in this fi eld. Here, we present a high-throughput model for the assessment of the innate immune response to AMI. The model is based on permanent surgical ligation of the left descending coronary artery (LAD) in mice, followed by complex fl ow-cytometry and histological analyses of immune cellular populations in blood and myocardium. We are presenting time-dependent qualitative and quantitative analysis results, demonstrating intense accumulation of Ly6Ghi neutrophils and Ly6Chi monocytes in the infarcted myocardium on days 1 and 3 post-AMI, followed by successive accumulation of reparatory Ly6CloMerTKhi macrophages, neovascularization and fi brosis development by day 7.


INTRODUCTION
Acute myocardial infarction (AMI) is the most severe manifestation of coronary artery disease, causing more than a third of deaths in developed nations annually 1 . Although advances in critical care, pharmacological interventions and emergency management of AMI by timely reperfusion therapy have signifi cantly reduced acute mortality in AMI, surviving patients remain at high risk to develop heart failure (HF) 2 . Thus, there is an urgent need for a better understanding of the pathogenesis of cardiomyocyte death, cardiac dysfunction and ventricular remodeling, in view of further development of novel therapeutic approaches.
Animal models of ischemic heart disease hold great importance, both in exploring novel mechanisms and for the development of future successful therapeutic interventions that can be clinically translated. Despite alternative AMI models that have been proposed and studied during recent years, including chemical 3 , radiofrequency 4 , thermal 5 or cryogenic injuries 6 , permanent left descending coronary artery (LAD) occlusion remains the most common model used by researchers 7 .
Classically, a ventilation-based thoracotomy to induce AMI by left descending coronary artery (LAD) ligation in mice was fi rst introduced by Johns and Olson in 1954, in which substantial animal trauma and tissue damage are involved in order to reach the ischemic event 8 . This procedure is extremely labor intensive and time-consuming, and is associated with high surgery-related mortality, due at least in part to the trauma caused by intubation and mechanical ventilation, and to the large intercostal thoracotomy. To alleviate these problems, Gao et al. substantially improved the method by eliminating the need for mechanical ventilation and orotracheal intubation, using a minimally invasive approach 9 .
In this paper, we will detail the adaptation and implementation in our laboratory of the murine AMI model proposed by Gao et al. To our knowledge, we are at the moment the only research team in Romania th at successfully uses this technique. Further, we are presenting an immunological analysis platform for the study of the innate immune response to AMI in this mouse model, as well as histological and immunohistochemical techniques for the study of revascularization and fi brosis. The mouse model and the associated techniques presented here offer a competitive platform for the future study of pathologic processes and potential therapeutic interventions aiming to improve cardiac function and recovery post-AMI.

Animals
Female wild-type (C57BL6) mice, 8-12 weeks of age, 20-25 g of body weight, were purchased from the Cantacuzino National Research and Development Institute (Bucharest, RO). All animal studies were conducted at the Experimental Station of "George Emil Palade" University of Medicine, Pharmacy, Science and Technology of Târgu Mureş according to protocols approved by the Scientifi c Research Ethics Commission of the University. The mice were allowed to acclimatize for at least one week before being included into an experimental group. Adequate environmental conditions (temperature, humidity and ventilation) were provided according to the University's guidelines for the accommodation and care of laboratory animals. Mice were housed in cages that ensured adequate space with a 12-hour light-dark cycle and free access to water and regular mouse diet.

Protocol of permanent coronary artery occlusion
2.2.1. The surgical instruments were sterilized and placed on the operating table so that they were easily and quickly accessible at any time during the operation. Keeping the same location of the instruments for all subsequent operations to develop certain refl exes is of vital importance during the intervention ( Figure  1A). 2.2.2. The mouse (generally 8-12 weeks of age and at least 20 grams body weight) was anaesthetized with 5% isofl urane combined with 1 L/minute oxygen inhalation in an inducing sealed chamber made of polycarbonate plastic to which the gaseous anesthetic was delivered by an EZ-SA800 Single Animal System (Philadelphia, USA) ( Figure 1B). 2.2.3. Once anaesthetized, the mouse was moved from the inducing chamber to a heated surgical board, immobilized with tape on all 4 paws and continuously anaesthetized with 2-3% isofl urane combined with 0.5 L/minute oxygen administered through a non-invasive mask placed over the nose and mouth of the animal, connected to the same anaesthesia system ( Figure  1C). 2.2.4. The fur was removed with a standard trimmer for small animals (an electric beard trimmer can also be used) and the skin was cleaned with water and thereafter with betadine and alcohol pads. 2.2.5. A small skin cut (1-1.2 cm) was made on the left side of the chest, followed by careful dissection and Razvan Gheorghita MARES et al. Studying the Innate Immune Response to MI in an Experimental Animal Model 2.2.7. In the shortest possible time, the LAD was visually located (in the region between the terminal branches of the left cardiac vein -left and lower limit, left atrial appendage -upper limit, right ventricle -right limit) and ligated at approximately 2-3 mm from the origin by using a 6.0 silk suture ( Figure 1H-1J). The ligation was deemed successful when the anterior wall of the left ventricle turned pale. Important care was taken not to puncture the right ventricle as it causes a massive hemorrhage. Silk thread is preferable because it better prevents the loosening of the ligature compared to the synthetic thread. This surgical step should be performed as quickly as possible, as increasing the retraction of the pectoral major and minor muscles to expose the 4 th intercostal space as shown in Figure 1F. It is very important that the pectoral muscles remain intact, because they are necessary to cover the opening (the mini-thoracotomy) once the heart has been returned into the thorax. 2.2.6. A small window was made at the 4th intercostal space level with a curved mosquito clamp to open the pleural membrane and pericardium ( Figure 1G). With the clamp slightly open and by gently pressing on the opposite side of the chest, the heart was smoothly "popped out" of the thorax through the small intercostal opening. rin solution was also added to the collection tubes, to prevent blood coagulation. After blood collection, the coronary and systemic vasculature were perfused with 5-10mL of PBS or 0.9% NaCl by left ventricle apical puncture (after fi rst cutting a liver lobe to allow the extravasation of blood and perfusion fl uid) and the hearts were harvested and fi xed in 4% formalin for 24 hours at room temperature. The perfusion step is important in order to remove blood cells from the coronaries. The formalin-fi xed tissues were processed and embedded in paraffi n using routine histological procedures. Hematoxylin and eosin (H&E) and Masson's trichrome kits were purchased from Sigma-Aldrich (MO, USA) and used as described in the protocol of the manufacturer. The hearts were serially sectioned in 4μm-thick sections at several levels along the transversal axis, starting at the apex and ending at the level of the ligature. Up to 10 slides (3 sections per slide) were collected from each level. The fi rst slide of the second level of sectioning was started 300μm from the fi rst slide of the fi rst level, and so on. Generally, 5-6 levels were obtained from each heart, covering approximately 1.5-2mm. For each marker, we stained one section from all 5-6 levels.
Hematoxilin eosin staining was used to visualize the infl ammatory infi ltrate and cardiomyocyte loss secondary to acute ischemia. Masson's Trichrome staining was used to stain collagen fi bers, in order to evaluate the extension of fi brosis in the infarcted myocardium. Primary antibodies against S100A8/A9, CD31 and alfa-smooth muscle actin (-SMA) were purchased from Cell Signaling Technology (Danvers, MA, USA). S100A8/A9 is a pro-infl ammatory alarmin that increases rapidly in the infarcted myocardium. S100A8/ A9 is produced in large quantities by neutrophils but also by activated monocytes and macrophages 10 . We time required for this step to over 30 seconds significantly decreased the survival rate in our experiments. 2.2.8. The heart was re-inserted into the thorax, the pneumothorax was evacuated by gentle bilateral side pressure, the pectoral muscles were layered again to cover the incision, and the skin was closed with 6.0 Prolene suture ( Figure 1K-1L). 2.2.9. The mouse was disconnected from the isofl uorane inhalator and allowed to breathe room air and monitored on a heating blanket until recovery. No invasive artifi cial respiratory aid was required. Parenteral analgesia with one dose of buprenorphine (0.1 mg/kg) was administered subcutaneously immediately after the incision was closed. The mouse usually regained consciousness within 3 minutes. 2.2.10. After completion of the surgical procedure, the mice were connected to a continuous three-lead electrocardiogram (ECG) monitoring device (Small Animal Physiological Monitoring System, Harvard Apparatus, United States) and the success of the intervention was determined by changes in the morphology of the QRS complexes ( Figure 2). 2.2.11. After the surgical intervention, the surgical tools were cleaned with water and alcohol.

Histological and immunohistochemical analysis
Histological analyses of heart sections were performed at 1, 3, 7 and 14 days after AMI induction. To collect the heart, the mouse was anesthetized by intraperitoneal injection of ketamine at a dose of 80 mg/ kg. If needed for biochemical or fl ow cytometry measurements, blood was collected by opening the thorax and puncturing the left and then the right ventricle with a heparinized 1mL syringe with a 26G needle. To heparinize the syringe, we drew and ejected a heparin solution once. A volume of 10mL of the same hepa- Chlorophyll-Protein-Cyanine 5.5 (PerCP-Cy 5.5)-conjugated anti-mouse Ly6-C antibody, Allophycocyanin (APC)-conjugated anti-mouse F4/80 antibody, PE-conjugated anti-mouse MerTK antibody (all reagents from BioLegend , San Diego, CA, USA). After staining of blood cells, erythrocyte lysis was performed with BD FACS™ Lysing Solution (BD Biosciences, San Jose, CA, USA). Cells were washed three times, resuspended in 500μL of staining buffer and immediately analyzed. All reagents were purchased from BD PharMingen (San Diego, CA, USA). Identifi cation of cellular populations in the antibody-stained cell suspensions was performed with a FACSAria II Flow Cytometer (BD Biosciences, San Jose, CA, USA). Data were acquired until 100.000 events were collected in the live cell gate, defi ned by using the forward/side scatter plots. The live cell gate was set to exclude events with low forward scatter signals, which are most likely cellular debris ( Figure 5 Aa, Ba and Ca). The FlowJo software (BD Biosciences, San Jose, CA, USA) was used for data analysis and fl ow cytometry plot design.

AMI induction was successfully validated by ECG
We examined the changes in the morphology of the QRS complexes by continuous three-lead ECG monitoring. As shown in Figure 2, the S-T segments were markedly elevated in mice after LAD ligation. These changes appeared within the fi rst 3 minutes of ischemia, thereby confi rming AMI induction. A successful surgery is characterized by a total surgical procedure time of up to 3 minutes, with a heart exposure time for LAD ligation of up to 20 seconds. The time to full recovery of the animals after cessation of inhalatory anesthesia was up to 5 minutes. In the hands of an experienced operator, the short-term post-operative animal survival reached up to 80%, and the long-term survival rate (from 1 hour after surgery until harvest) was up to 95%. The overall survival rates (including surgery-related death) were about 65-75%, which is in accordance with the existing literature 9 .

The histological features and timeline characterization of the infarcted myocardium
We developed histological and immunohistochemical methods to assess the timeline of the local changes induced by ischemia to the local environment of the used immunohistochemical staining of the endothelial cell marker CD31 to visualize capillaries in the infarcted and remote myocardium. -SMA is a marker used to identify mature myofi broblasts. BrightVision Poly-HRP-Anti Rabbit Biotin-free (ready-to-use) from Immunologic (a WellMed Company, Duiven, The Netherlands) was used as secondary antibody. Images were acquired on an Axio Imager Z2 microscope with a color Axiocam 506 camera and processed using the ZenPro 3.2 software (all Zeiss, Germany) and QuPath (https://qupath.github.io).

Cell isolation for fl ow cytometry
In separate experiments, after the mice were sacrificed and perfused as above, the heart was rapidly removed and stored in PBS tubes kept on ice. The heart was minced in small parts with scissors (or scalpel) and enzymatically digested by a solution containing collagenase type I (2 mg/ml), collagenase type XI (0.2 mg/ml), deoxyribonuclease (DNase) type I (0.16 mg/ ml), and hyaluronidase type I-S (0.14 mg/ml), all reagents from Sigma-Aldrich. The incubation time was one hour at 37°C, shaken at 230 rotations per minute. Cell suspensions were sequentially fi ltered through 70μm and 50μm cell strainers to remove debris. For blood analysis, the samples were collected in pre-heparinized tubes by left and right ventricle apical puncture and kept on ice, as described above.

Flow cytometry
The protocol for cell isolation and fl ow cytometry staining is an adapted version of the protocol used in our previous publications 11,12 . Cells from blood and heart were collected at 1, 3 and 7 days after AMI and analyzed by fl ow cytometry. After tissue digestion, cells from individual hearts were resuspended in 50mL of staining buffer (1% fetal calf serum, 0.05% sodium azide in PBS), centrifuged at 1500 RPM for 5 minutes, resuspended in 250μL of staining buffer, and then split into fi ve samples of 50μL from each heart (of which 3 samples were used to stain the leukocyte populations mentioned below). Circulating cells were stained in tubes with 50μL blood. Leukocytes (neutrophils, monocytes and macrophages) were stained by adding the antibody mix into the tubes containing 50μL heart cell suspension or 50μL blood, followed by incubation in the dark at 4°C for 40 minutes, and 3 washing steps with staining buffer. The following antibodies were used, diluted 1:200 before being added into the samples: Pacifi c Blue (PB)-conjugated anti-mouse CD45 antibody, Allophycocyanin/Cyanine 7 (APC/ Cy7)-conjugated anti-mouse CD11b antibody, Phyco-infl ammatory cells that progresses from day 1 to day 3. Interestingly, the infl ammatory infi ltration does not only occur in the infarcted area (central), but also in the nearby and remote myocardium ( Figure 3B, up-myocardium. Myocardial necrosis triggers an acute infl ammatory reaction that is maximal during the fi rst 3 days. In the hematoxylin-eosin staining presented in Figure 3A-3B, we can observe a massive infi ltration of Figure 3. Infl ammatory infi ltration post-AMI determined by histology and immunohistochemistry A, H&E at day 1 after AMI. B, H&E at day 3 after AMI. C, S100A8/A9 positive cells at day 1 after AMI. D, S100A8/A9 positive cells at day 3 after AMI. E, S100A8/A9 positive cells at day 7 after AMI. F, S100A8/A9 positive cells at day 14 after AMI. All 10X magnifi cation.

Local and systemic analysis of the innate immune cell dynamics by fl ow cytometry
We analysed by fl ow-cytometry the immune cells from blood and heart at various time-points during the fi rst week post-AMI. CD45, a pan-leukocyte marker, was used to identify the immune cells and to estimate the percentages of circulating and heart-infi ltrating leukocytes. The surface marker CD11b was used to further discriminate cells of myeloid origin ( Figure 5 Ac, Bb, Cb). Neutrophils were characterized by the presence of high levels of the lymphocyte antigen 6G (Ly6G), as previously described 14,15 . Monocytes were identifi ed by the expression of surface CD115 and were divided into two distinct populations according to the expression levels of the lymphocyte antigen 6C (Ly6C) on their surface. The classical or infl ammatory monocyte subset have high levels of Ly6C (Ly6C hi ) and the nonclassical or patrolling subset are Ly6C low 16 . Macrophages were identifi ed as F4/80+ cells, as described in the literature 17 . In addition, according to Wan E et al 18 , we identifi ed reparatory macrophages expressing the myeloid-epithelial-reproductive tyrosine kinase receptor (MerTK) on their membrane. MerTK is an important efferocytosis receptor that mediates the uptake and clearance of dead or dying cardiomyocytes by the reparatory macrophages that infi ltrate the myocardium. Based on surface markers, reparatory macrophages are defi ned as Ly6C low MerTK hi , and their Ly6C hi MerTK low counterparts are considered to be infl ammatory.
The serial gating presented in Figure 5 shows a massive infi ltration of neutrophils (CD45 + CD11b + Ly6G + cells) into the heart in the fi rst 3 days post-AMI. The percentage of neutrophils out of the total myeloid cell population (CD45 + CD11b + ) peaked on day 1 after infarction ( Figure 5A-d) and massively decreased by day 7 (71.9% on day 1 vs. 5.57 % on day 7, Figure 6A), revealing the importance of neutrophils in the immediate infl ammatory phase post-AMI.
Further, the ischemic injury led to a progressive increase in the number of monocytes (CD45 + CD11b + CD115 + cells) in the blood stream, with a peak on day 3 ( Figure 5B, Figure 6B). Importantly, the infl ammatory Ly6C hi sub-population constituted the overwhelming majority of monocytes at this stage ( Figure 5 B-d, Figure 6C). As shown in other stu-dies19, the infl ammatory Ly6C hi monocytes are recruited from the blood stream into the heart post-AMI and differentiate into monocyte-derived macrophages, involved in both the infl ammatory and the reparatory phases of the cardiac immune response to AMI. per left corner), which may have important negative effects on the regions of the heart which have not been affected by the infarction (Figure 3A-B). The detailed identifi cation of the different infl ammatory cell populations by fl ow cytometry is presented below.
Besides the infl ammatory infi ltrate, at this stage we can observe microscopic aspects of interstitial oedema as blank spaces between the cells, even among the remote cardiomyocytes. We also observed nuclear pycnosis, as well as loss of cardiomyocyte nuclei and striations, all features of cardiomyocyte necrosis. In addition, we found an intense infi ltration of S100A8/ A9 in the infarcted area and in the remote myocardium ( Figure 3C-F). The staining was both cellular and extracellular, suggesting S100A8/A9 secretion from the cells into the surrounding tissue. The S100A8/A9 infi ltration reached a maximum on day 3 ( Figure 3D), and then progressively decreased until day 14. On day 14 post-MI, S100A8/A9 was no longer detectable in the myocardium ( Figure 3F). Myocardial thinning already occurs on day 7 after the onset of ischemia, due to removal of dead cardiomyocytes, which are replaced by fi brous tissue with a lower volume. The extent of myocardial fi brosis was detected by Trichrome Masson staining on days 7 and 14 post-AMI ( Figure 4A). The preserved myocardium is colored in red, and the fi brous tissue in blue. Here, we present the border zone between infarcted and healthy myocardium (Figure 4A).
Additionally, on days 7 and 14 we could observe large fully endothelialised blood vessels forming in the border zone between the infarcted myocardium and the remote myocardium ( Figure 4B). We also detected an important infi ltration of mature myofi broblasts expressing -SMA in the infarction zone and in the border zone ( Figure 4C). Myofi broblasts synthesize type I collagen that strengthens the infarcted area and prevents rupture, forming the post-AMI myocardial scar. Both bone marrow-derived fi broblasts and myofi broblasts, as well as a TGF1-driven conversion of epicardial-derived cells (resident fi broblasts) into collagen-secreting myofi broblasts have been described 13 . By using the above techniques, the effects of therapeutic or experimental interventions on the infi ltration of infl ammatory cells, neovascularization, myofi broblasts, and fi brosis can be individually assessed in the infarcted area, remote myocardium and in the border zone. al reperfusion strategies and antithrombotic therapies, innovative therapies are still needed in order to maximize the survival and prognosis of AMI patients. Early studies suggesting that infl ammation is deleterious and has a considerable impact on long-term prognosis have stimulated great interest in developing benefi cial anti-infl ammatory strategies in AMI patients 20,21 . The reliance on animal models of cardiovascular disease is a mandatory path to follow for a better understanding of disease pathophysiology and for testing of any kind of therapeutic approach. Experimental ischemia can be induced by ligating the coronary arteries, and since the fi rst mouse AMI model was published 8 , LAD ligation has been considered as a standard model to induce myocardial ischemia and infarction. The initi-Consequently, we examined the presence of macrophages (CD45 + CD11b + F4/80 + cells) in the myocardium during the fi rst week post-AMI. Our analysis has shown a progressive accumulation of macrophages into the heart, from day 1 to day 7 ( Figure 5C, Figure 6D). By day 7 post-AMI, the vast majority of macrophages acquired a reparatory phenotype, characterized by high surface expression of the efferocytosis receptor MerTK (Figure 5 C-d and Figure 6E), while the percentage of infl ammatory Ly6C hi MerTK low macrophages was much lower ( Figure 6F).

DISCUSSION
Despite remarkable progress in reducing mortality from AMI in recent decades by streamlining myocardi- Figure 6. Immune cell dynamics post-AMI in blood and myocardium during the fi rst week post-AMI, determined by fl ow cytometry Percentages of: A, neutrophils in the infarcted myocardium out of total myeloid cells. B, circulating monocytes out of total myeloid cells in blood. C, infl ammatory Ly6Chi monocytes out of total blood monocytes. D, macrophages in the infarcted myocardium out of total myeloid cells. E, reparatory Ly6C lo MerTK hi macrophages in the infarcted myocardium out of total macrophages. F, Ly6C hi MerTK lo macrophages in the infarcted myocardium out of total macrophages. ventricular wall, and smaller infarctions usually have a central location, located in the middle of the wall thickness. Monitoring the ST segment elevation on ECG after the procedure allows to confi dently exclude non-infarcted mice from the study groups. We also observed that short-term death within the fi rst hour post-AMI is usually due to surgical errors or very large infarct size and is the main factor that impairs overall survival rate. Conversely, even in the case of induction of a large infarction, the long-term survival rate (from 1 hour after surgery until harvest) is excellent if the surgery is performed properly and fast (LAD ligation time within 30 seconds). Besides, care should be taken to minimize disturbing animals during the period when myocardial rupture may be triggered by stress, particularly at days 1-7 post-AMI. Ultimately, especially when practicing the model, we strongly recommend the researchers to perform autopsies for all prematurely dead mice in order to evaluate the potential cause of deaths. The early deaths usually occur due to technical issues such as excessive blood loss into the thorax, potentially caused by injuries to large vessels of the thorax wall or mediastinum, or lung injuries. Later deaths are usually due to AMI complications such as cardiac wall rupture or heart failure.
Integrating this AMI animal model in our platform allowed us to further investigate the histological changes induced by myocardial ischemia and to perform an in-depth analysis of tissue and blood cellular populations by fl ow cytometry. The local infl ammatory infi ltrate following an AMI has a crucial infl uence on both the size of the injury and on the effi ciency of the reparatory mechanisms that determine myocardial remodeling and post-AMI prognosis 24 . The myocardial ischemia followed by necrosis triggers a potent immune reaction involving two mechanistically distinct phases: the infl ammatory phase and the repa ratory phase 25 .
Histologically, we showed that in the initial infl ammatory phase (the fi rst 3 days after AMI) the infarcted myocardium was massively infi ltrated by infl ammatory cells, mainly consisting of neutrophils 11 . The fl ow cytometry analysis revealed a robust initial infl ux of Ly6G+ neutrophils into the infarcted myocardium, followed by a marked decrease by day 7 after AMI. Importantly, the infi ltration of S100A8/A9-positive cells correlated with neutrophil dynamics. Of note, S100A8/A9 (calprotectin) is a proinfl ammatory alarmin that is readily produced and stored in large amounts in neutrophils. S100A8/A9 has been intensively studied by our group in recent years, with promising results as a potential ally-employed procedure to ligate LAD in mice involves a large intercostal thoracotomy, requiring tracheal intubation and mechanical ventilation to keep the animal alive 22,23 . Access and ligature of the coronary artery is done through a large thoracic opening and the entire procedure takes up to 1 hour. During this entire period, the mouse has to be sedated and mechanically ventilated. After surgery, the mouse has to be weaned off the ventilator under close supervision until it regains spontaneous breathing, a process that can take between 30 minutes to a few hours. The main downside of this traditional model is the requirement for long-term parenteral general anesthesia, intubation and use of an external ventilator. The procedure causes extensive tissue damage, is associated with high mortality due to intubation and ventilation, and is very time consuming.
Here, we described the implementation in our laboratory of a minimally invasive approach initially proposed by Gao et al. 9 , and we propose a complex model of histological, immunohistochemical, and fl ow cytometry analyses to evaluate the immune response post-MI. This surgical method is more effi cient to induce AMI and causes less injuries compared to the classic procedure, as it is minimally invasive and does not require invasive mechanical ventilation and long-term parenteral anesthesia. The method leads to a major decrease of intra-operative and post-operative recovery times along with an increase in short-and long-term survival rates 9 . When performing the procedure, there are several important details that in our experiments have been proven to be crucial in obtaining favorable results. It is very important to ligate the LAD at the same anatomical region in all mice, 1 to 3 mm distal of the left atrium. This ensures that large infarctions of similar sizes are induced in all animals. However, differences in the infarction size will still occur due to natural variations in the coronary tree anatomy even in same-sex siblings. Failed surgery is characterized by missing the coronary artery, which will not generate an infarction and will only lead to myocardial damage from the suture. These failed cases can be identifi ed during the histological examination as small infarctionlike areas exclusively located at the same level as the suture. The success rate depends on the operator and increases with the level of experience. The successful infarctions are rather large, usually extend to the apex, and decrease in size when they approach the suture level, as they follow the anatomy of the irrigated territory. Large infarctions extend throughout the stage is also characterized by stimulation of angiogenesis ( Figure 4B). Importantly, the reparatory macrophages are critical stimulators of the fi brogenic and angiogenic responses 32 . In experimental models of AMI, depletion of macrophages causes severe reparative defects associated with impaired fi broblast activation and attenuated angiogenesis 33 . Here, we demonstrate that robust neoangiogenesis occurs, especially in the border zone between the infarcted and the remote myocardium.
The resolution of infl ammation and initiation of fibrosis are fi nely tuned processes aiming to limit the infl ammatory myocardial damage and lead to optimal healing. However, as the reparatory macrophage accumulation is dependent on the initial infl ammatory activation, all interventions that inhibit the initial infl ammatory response invariably affect the reparatory phase as well. Affecting the timing and magnitude of the infl ammatory response can help reduce the infl ammatory damage but can also impair the adequate wound healing process. For example, we have shown that pharmacological blockade of the S100A8/A9 alarmin has important therapeutic benefi ts when administered during the infl ammatory phase of the AMI but led to impaired repair and cardiac dysfunction when given long-term 10,11 . Therefore, care should be exercised when developing immunomodulatory therapies for AMI and the effects on both phases of the immune response should be carefully considered.

CONCLUSION
The innate immune system plays a critical role in the repair and remodeling of the infarcted myocardium. An initial infl ammatory phase is required to recruit the cells that repair the myocardium, but excessive infl ammation leads to further damage. In contrast, a weakened infl ammatory phase leads to ineffi cient repair and also affects function. These are very important aspects that have to be taken into consideration by all attempts to develop anti-infl ammatory therapies for AMI.
The AMI model described here, alongside the subsequent tissue and cellular analyses, is an important tool that supports the understanding of the role of the immune system in AMI and the testing of novel therapeutics approaches. Any potentially successful immunomodulatory therapy would have to limit the infl ammatory damage to the myocardium, while improving or keeping intact the reparatory mechanisms. therapeutic target after AMI 11,12 .
Following neutrophils, infl ammatory monocytes form the second wave of cellular infi ltration into the myocardium. We demonstrated by fl ow cytometry that the ischemic injury is followed by a sharp increase in the number of infl ammatory Ly6C hi monocytes in the blood stream, starting from day 1 after the coronary occlusion and reaching peak numbers by day 3. These monocytes infi ltrate the myocardium and subsequently shift away from their pro-infl ammatory phenotype to give rise to macrophages with a predominantly reparative phenotype (Ly6C low macrophages) 26,27 .
The early infl ammatory response initiates uptake and removal of dead cardi ac tissue, and paves the way for the reparative phase, which is also mediated by immune cells. The reparatory phase involves resolution of infl ammation, myofi broblast proliferation, scar formation and neovascularization 28,29 . After AMI, the relatively small population of resident macrophages expands by abundant recruitment of monocytes from the blood stream. The spleen has been identifi ed as the main reservoir supplying cells for the initial monocyte infi ltration 30 . The fl ow cytometry gating strategy ( Figure 5C) demonstrates a progressive accumulation of reparatory Ly6C low macrophages starting from day 3 after AMI and increasing to day 7 ( Figure 6D). Novel insights into the post-AMI healing process indicate a critical role for the efferocytosis receptor MerTK, expressed by reparatory macrophages ( Figure 5C-d, lower right quadrant). In previous experimental studies, it has been shown that the absence of MerTK led to impaired efferocytosis of dying cardiomyocytes and impaired cardiac healing, leading to progressive loss of function 31 . Here, we demonstrate a progressive accumulation of CD11b + F4/80 + macrophages in the myocardium, and a phenotype switch from a predominant Ly6C hi MerTK low population on day 1, to a population of Ly6C low MerTK hi cells on day 7, which clearly dominate the macrophage infi ltrate (59 % out of total macrophages, Figure 5C-d, 5C-e, 6D-F).
The extent of the myocardial damage, as revealed by the fi brous scar, can be visualized by tissue staining with Masson's trichrome starting 5-7 days after injury ( Figure 4A). This proliferative phase of the infarct healing is characterized by expansion and activation of the cardiac fi broblast population, which acquire a myofi broblast phenotype characterized by expression of alfa-smooth muscle actin (-SMA) and produce the collagen fi bres that form the scar ( Figure 4B-4C). This