Abstract
Cell therapy of myocardial infarction (MI) is under clinical investigation, yet little is known about its underlying mechanism of function. Our aims were to induce a sub-lethal myocardial infarction in a rabbit, to evaluate the abilities of labeled bone marrow mononuclear cells to migrate from the vessel bed into extracellular space of the myocardium, and to evaluate the short-term distribution of cells in the damaged left ventricle. Sub-lethal myocardial infarction was induced in rabbits by ligation of the left coronary vessel branch (in vivo). The Langendorff heart perfusion model (ex vivo) was used in the next phase. The hearts subjected to MI induction were divided into 3 groups (P1–P3), and hearts without MI formed a control group (C). Nanoparticles-labeled bone marrow mononuclear cells were injected into coronary arteries via the aorta. Perfusion after application lasted 2 minutes in the P1 group, 10 minutes in the P2 and C groups, and 25 minutes in the P3 group. The myocardium of the left ventricle was examined histologically, and the numbers of labeled cells in vessels, myocardium, and combined were determined. The numbers of detected cells in the P1 and C groups were significantly lower than in the P2 and P3 groups. In the P2 and P3 groups, the numbers of cells found distally from the ligation were significantly higher than proximally from the ligation site. Bone marrow mononuclear cells labeled with iron oxide nanoparticles proved the ability to migrate in the myocardium interstitium with significantly higher affinity for the tissue damaged by infarction.
Introduction
Bone marrow cell transplantation has been used in clinical hematology for almost 40 years, although the stem cell concept and role of stem cells in hematopoiesis were formulated by Maximov already in 1906. Sources other than bone marrow for hematopoietic stem cells, such as peripheral and umbilical cord blood, were later discovered. But discoveries of the last few years have renewed interest in the cells’ biology and use, primarily in non-hematological tissue repair processes and disease therapy. The impulse for such research in cardiology was the finding of Y chromosomes in patients after female-to-male heart transplantation (1). At the same time, questions were raised about the dogma that stem cells should be able to differentiate only into lineages of cells with the morphological markers and specialized tissue functions in which they appear.
Progress in the treatment of myocardium infarction (MI), in spite of urgent restoration of perfusion by angioplasty or by conventional drug therapy, still cannot replace dead myocardial cells because of their very limited regenerative ability. No other treatment process is able to repair the necrotic myocardium or the already formed scar.
Therapeutic administration of stem cells with the potential to repair damaged tissue has significant potential not only in cardiology but also in treating such degenerative diseases as Parkinson’s disease, Duchenne muscular dystrophy, traumatic damage of the spinal cord, and others.
Preclinical studies have tested the following cell types related to MI cell therapy in human medicine: skeletal myoblasts injected into the center and surrounding area (2), mononuclear cells of bone marrow (BM) applied into the coronary artery (3), blood and BM-derived progenitor cells (4), BM stem cells (5) applied directly into the infarction area or surrounding region, and BM stromal cells (6). In all these cases, some improvement in the efficiency of the affected heart was observed.
There are many experimental models for inducing myocardial infarction followed by the application of stem cells, but these are in heterogeneous groups (differing by choice of animal species, way of the cells’ application and their origin, and technique of labeling). In recent publications dealing with cell therapy in connection with reparation of affected myocardium, the intramyocardial administration of bone marrow cells prevails. From the clinical point of view, there is demand for intraluminal or, ideally, intra-coronary cell application. In considering the continuance of the experiment on an in vivo model, we chose a rabbit as an animal model because of its sufficiently large size for noninvasive examination (especially through echocardiography) and relatively wide range of antibodies that is necessary for separating the BM mononuclear cells. The work aimed to verify the possibility of inducing sub-lethal myocardium infarction in a rabbit by ligation of the coronary artery branch, verify the ability of the labeled BM mononuclear cells to leak from the vascular bed into the myocardium extracellular space, and evaluate the short-time distribution of the administrated mononuclear cells in the affected left ventricle. Long-time in vivo experiments, which could shed light on the underlying mechanism of BM cells activity in damaged myocardium, will follow this ex vivo kinetic study.
Materials and Methods
Experimental Animals.
The animals used were 30 male rabbits of the Hy-Plus breed, weighing between 2.5 and 3.5 kg, from a monitored production breeder (V. Pokorny, Malesovice, Czech Republic). There was a 7-day adaptation time before the experiments. The animals were kept in metal cages, 50 ×60 ×70 cm, during the experiment. The cages were situated in a room with a temperature of 19 ± 1°C and relative air humidity between 55% and 60%. The complete feed mixture (Biostan KV, Biosta Blucina, Czech Republic) and drinking water were administered ad libitum. A natural light regimen was maintained. During both the adaptation and experiment times, the health condition was checked daily. Before the experimental insertion, echocardiography examination was performed in all rabbits to reveal any sub-clinical abnormalities and to gain the physiological range of values for the in vivo experiments.
The experimental individuals were divided into 4 groups: Group C was the control (n =4). These animals were not subject to induction of myocardial infarction but only to application of the nanoparticle-labeled BM mononuclear cells, with 10 minutes of perfusion.
Twenty-six rabbits were used for the induction of myocardial infarction. Because the mortality after MI was quite high, however, just 14 animals were enrolled in the proper ex vivo model (groups P1, P2, P3), as follows: Group P1 (n =4)—induced MI, application of the nanoparticle-labeled BM cells, 2 minutes perfusion after the application. Group P2 (n =6)—induced MI, application of the nanoparticle-labeled BM cells, 10 minutes perfusion after the application. Group P3 (n =4)—induced MI, application of the nanoparticle-labeled BM cells, 25 minutes perfusion after the application.
The experiment was made in concordance with Act No. 246/1992 on the Protection of Animals from Maltreatment, as later amended.
Bone Marrow Collection.
Bone marrow was collected 21 ± 2 days before the MI induction in animals under general anesthesia (neuroleptanesthesia). First, pentazocine (3 mg/kg of live weight) was administered together with xylazine (5 mg/kg of live weight) in a single injection. This was followed by a standard preparation for collection (depilation, disinfection). After 5 minutes, half of a calculated dose of ketamine was administered quickly (35 mg/kg of live weight). Its next administration (a quarter dose) was indicated according to the clinical condition, with the aim to achieve zero protective reaction to the bone marrow aspiration. The extraction was made aseptically using an 18G syringe to heparin (1 ml of heparin to 10 ml of the aspirate) from the tuber ischiadicum. The aspirate was then processed by the gradient centrifugation method for the mononuclear cell separation (see below).
Bone Marrow Cells Processing, Culture and Labeling.
The aspirates of bone marrow cells were processed by gradient centrifugation method using a Histopaque 1077 (Sigma-Aldrich). Collected mononuclear BM cells were placed into 10 ml SFEM medium (Stem Cell Technologies, Canada) containing antibiotics (penicillin–streptomycin, Sigma-Aldrich) in cell culture flasks. Iron oxide nanoparticles (ferucarbotranum [Resovist inj.], Bayer Schering Pharma AG, Germany) were added to the culture to final concentration of 100 μg Fe per ml of media. BM mononuclear cells with the iron oxide nanoparticles were cultured at 37°C, 5% CO2 and 100% humidity for 21 days, on average. The viability of cells was checked weekly, and it exceeded 90% at the end of cultivation in all samples.
Induction of Myocardial Infarction.
Because of the planned surgical intervention, enrofloxacin (5 mg/kg of live weight) was administered intramuscularly 3 days before the induction of infarction. The rabbits were introduced into the general anesthesia by the modified protocol for rabbits (7) without the ether induction and atropine pre-medication. The protocol consists of intramuscular administration of diazepam (2 mg/kg of live weight) in combination with xylazine (5 mg/kg of live weight) and ketamine (35 mg/kg of live weight). Anesthesia was maintained according to the clinical condition by repeated application of ketamine in the original or a half dose.
In the zero minute, diazepam and xylazine and ketamine were administered. In the 10th minute of the protocol, intubation by the tracheal cannula Hi-Contour (Mallinckrodt Medical, IRL) followed, with I. D. 3 or 3.5 mm according to the size of the rabbit. In the 15th minute, connection was made to the registration devices (the pulse oximeter and ECG monitor) and the control ECG was recorded.
The left-sided approach was chosen for the operation. After cleaning the operation field, the preparation in layers and thoracotomy in the third intercostal space were made. The pericardium was fixed by Pean’s forceps in the area of the heart apex. We reached the necessary tension to enable cutting of the pericardium and approximation of the free left chamber into the thoracotomy cut by displacement of the Pean’s forceps.
Ramus proximalis ventriculi sinistri as the first ventral branch of ramus circumflexus a. coronariae sinistrae or ramus collateralis distalis as the second horizontal branch of ramus interventricularis paraconalis (in humans ramus interventricularis anterior) a. coronariae sinistrae was ligated, as illustrated in Figure 1a. The goal of surgery was to induce ischemia of the apex of the left ventricle by ligation of the dominant branch that supplies this area (Fig. 1b). Braided non-absorbable material (Ethibond 5/0, Ethicon, United Kingdom) was chosen. Ligature was performed by double pricking of the vessel in a Z pattern, so that the pricking did not penetrate the cavity of the left ventricle (Fig. 1c). The ligature was left in situ until the next day. The chest cavity was closed in the standard way by apposition of the muscle layers and then the pneumothorax was removed.
Extraction and Perfusion of the Isolated Heart by the Langendorff Method.
Eighteen to 24 hours after induction of myocardium infarction, total anesthesia was induced in the rabbits with respect to the protocol mentioned above. Anesthesia was followed by euthanasia after the tolerance stadium by an overdose of the anesthetics with a full initial dose administered intravenously. A sternotomy was performed immediately after breathing had ceased, and the heart was separated from the large vessels. The isolated organ was left in a 0.9% NaCl solution with ice for 5 minutes. During this time, the heart was connected with a cannula that was placed into the aorta. The cavity of the left ventricle was rinsed with physiological solution to remove the possible intracardial blood coagula. Afterward, the heart was connected to the apparatus and its tightness was controlled. The total time of cold ischemia was maximally 5 minutes. Perfusion of the isolated heart according by the Langendorff method followed. The volume of the column to which the heart was fixed was filled with the Krebs-Henseleit solution and albumin and saturated with 95% O2 and 5% CO2. The column temperature was kept between 37°C and 38°C, and constant pressure was maintained at 80 cm of water. The coronary artery ligature was removed, and 5 minutes of perfusion followed.
Application of Autologous Mononuclear Bone Marrow Cells.
A suspension of the nanoparticle-labeled mononuclear BM cells was applied by the catheter through the three-way valve. The average number of 1,480,000 cells in a volume of 1.8 ml was applied in fractions over approximately 4 minutes. The spontaneous perfusion was stopped after 2 minutes (group P1), 10 minutes (groups P2 and C), and 25 minutes (group P3). Then followed 5 minutes of intravascular heart fixation using 20 ml of 2.5% glutaraldehyde solution. The collected muscle samples were routinely processed by the method of histological evaluation (see below).
Histology Examinations.
The sample was taken from the cardiac muscle of the left ventricle, proximally (healthy myocardium) and distally to the ligature (MI site). The myocardium in the control group was taken out from the corresponding places. The cut was conducted approximately 5 mm above and under the ligature, respectively. Two samples from each rabbit were evaluated. The sample processing was made in a standard way using paraffin-embedded tissue. The paraffin-embedded sections were stained with hematoxylineosin and potassium ferricyanide for Turnbull’s blue.
The only preparations submitted to the semiquantitative evaluation of the number of labeled cells in the cut were those that stained positive for the presence of iron (by Turnbull’s blue). Every second viewing field (VF) was evaluated in the meandering examination when enlarged 400-fold. The total number of evaluated VFs per one sample was 15. Each nanoparticle-labeled cell (or stained spot) was evaluated as a positive finding if it properly corresponded in size, shape and color intensity. In every VF, cells located intravascularly and cells in the tissue were calculated separately.
In samples stained with hematoxylin-eosin, the presence of early post-infarction changes was evaluated qualitatively: protraction and undulation of the muscle fibers, loss of the coloring ability of the core, hypereosinophilia of the cytoplasm, interstitial edema, polymorphonuclear infiltrate and hemorrhage.
Transmission Electron Microscopy.
Dissected myocardial tissue was fixed in 400 mmol/l glutaraldehyde in cacodylate buffer pH 7.4. Then post-fixation was done using two baths of 40 mmol/l OsO4 in cacodylate buffer pH 7.4. Dehydration, immersion and embedding into Durcupan ACM were carried out using standard procedures. Ultrathin sections were made on an LKB Nova and Leica ultra-microtome, respectively, and stained with lead citrate or with uranyl acetate and lead citrate. The sections were viewed and photographed in a Morgagni 268D electron microscope.
Statistical Analysis.
The results are expressed as averages and standard deviations. For groups P1, P2 and P3, a simple linear regression model was used. It predicts the cell count dependent on the logarithm of the time of perfusion. The regression parameter for the regression curve was tested against the null hypothesis using standard parametric tests based on asymptotic normal distribution.
To test the dissimilarity in the cell count for different groups (the cell counts below and above ligature, in the heart muscle, and in blood vessels), the nonparametric Wilcoxon rank sum test was used.
Results
Myocardial infarction was induced on a group of 26 rabbits by ligation of the coronary artery branch. Intra-coronary application of the nanoparticle-labeled mononuclear bone marrow cells was performed in 14 animals that survived the surgery 18–24 hours after infarction was induced on the model of the perfused heart using the Langendorff method. Perfusion after the cell application in the P1 group continued for 2 minutes, in P2 for 10 minutes, and in P3 for 25 minutes. Just application of labeled cells ex vivo on the same model was performed on healthy hearts in the control group C with 10 minutes of perfusion.
The ability of the nanoparticle-labeled mononuclear cells to colonize the area of the infarction site was evaluated semiquantitatively. Each nanoparticle-labeled cell or focal area corresponding by its size, shape and color intensity was evaluated as a single positive event. In each viewing field, the labeled cells in vessels and cells colonizing the tissue were counted separately (Fig. 2a, b, c), above and below the place of ligation. Electron microphotographs of a BM cell attached to the capillary wall and a BM cell residing in the interstitial tissue are shown in Figure 3a, b, c, d.
The semiquantitative evaluation of the labeled cells proved only occasional presence of the labeled cells in group C in both evaluated regions of the heart. Similarly, few labeled cells above and below the place of ligation were found in group P1 (2 minutes of perfusion) and it did not differ significantly from the control group. In group P2 (10 minutes of perfusion), however, the maximum counts of the labeled cells in vessels (33/15 VF), as well as in the muscle (37/15 VF), were noticed below the place of ligation. Similar observation was found for group P3. The average cell counts observed at both locations in 15 VFs for each group are summarized in Table 1.
A graphic presentation of the perfusion-time kinetic profile for labeled cell counts found above and below the ligation, as well as for both locations together, is shown in Figure 4a, b, c for all groups. The null hypothesis, which stated that the groups P1, P2 and P3 would be the same, was rejected at level P =0.036. In this case, the required condition of normality of the measured random variable is acceptable because the cell counts were obtained as the means of cell counts of the samples’ parts and can be considered as asymptotic normal values.
The statistical description of the measured data for all groups, comparing the sites above and below the ligation, is given in a graphical form in Figure 5. Using the non-parametric Wilcoxon rank sum test, there were the significant differences, as described below and shown in Table 2. The cell counts below ligature (in muscle, blood vessels, and total, respectively) differed significantly between groups C and P2 (P =0.014, P =0.013, P =0.014), differed significantly between groups C and P3 (P =0.029, P =0.027, P =0.029), differed significantly between groups P1 and P2 (P =0.014, P =0.014, P =0.001), and differed significantly between groups P1 and P3 (P =0.029, P =0.042, P =0.028). The null hypothesis was not rejected for location below ligature between groups C and P1 in muscle, blood vessels and total cell count. Also, the data in groups P2 and P3 are similar.
Above ligature, there were significant differences only between groups C and P3 for the total cell count (muscle =blood vessels, P =0.028) and between groups C and P2 for the total cell count (P =0.027).
Evaluation of the histopathological changes in heart muscle under the ligation of the coronary artery branch proved the presence within 24 hours of changes that are characteristic of an early period of myocardial infarction. Modest evidence of ischemic myocardium (hypereosinophilia of cytoplasm, interstitial edema) appeared also in some samples that were taken above the place of ligation.
Of the 26 rabbits that underwent myocardium infarction, 12 died. Death resulted from ventricular fibrillation. Fibrillation appeared always approximately 5–10 minutes after ligation. The total mortality caused by the myocardium infarction induction was 46.2%.
Discussion
The possibility to induce sub-lethal myocardial infarction in a rabbit by ligation of the branch of a coronary artery was confirmed in the work presented. The experiment used the Langendorff perfused heart model. The ability of the autologous mononuclear bone marrow cells after their labeling with nanoparticles of iron oxide to migrate from the coronary bed and to colonize the damaged tissue was proven.
The model of the isolated heart has all the drawbacks of an ex vivo organ model. Its use does not reduce the number of experimental animals. Neurohumoral mechanisms, which regulate homeostasis on the level of different organ systems during the pathophysiological response, cannot be considered. On the other hand, it enables research on the myocardium, including its metabolic functions, in a way that would not be feasible in situ from the viewpoint of practical implementation. The influence of reperfusion on the ischemic myocardium is, generally, a solved problem (8). Measurement of biochemical changes as a result of hypoperfusion (myocardial glycogen, lactate concentration, and some preferred amino acids) is possible (9, 10). The optimal solution for measuring different hemodynamic parameters in an easy approach (the left ventricle–coronary flow, pressure ratio) is recommended (11).
The Langendorff model in our setting made intra-coronary application of the labeled cells significantly easier. The model allows easy and effective evaluation of the short-term migration ability of the labeled cells (after checking their viability), as it does the counting of implanted cells.
Induction of myocardial infarction can be performed in various ways. From the viewpoint of simulating acute myocardial infarction in humans, those methods based on ischemic damage of the myocardium can be regarded as most appropriate. For this reason, we chose the ligation of coronary artery branches to induce MI. We performed ligation of the r. collateralis distalis rami interventricularis paraconalis, which, with regard to the range of damage, is the best way to simulate the results of spontaneous acute myocardium infarction in humans (12). If the dominant blood supply vessel was ramus circumflexus a. coronariae sinistrae and its distal branch ramus proximalis ventriculi sinistri, the ligation was performed at the distal half of this branch.
We proved the acceptability of this approach with regard to the mortality rate and evaluation of histopathological changes. The exact comparison of this index with data in the literature is difficult since the experimental protocols are not the same. When we try to compare our approach in spite of these difficulties, we can only state that the mortality in our group did not differ significantly from that (38.9%) in the cryogenically induced MI performed by Thompson et al. (13). The MI induction made by Norol et al. (14) and Terrovitis et al. (15) recorded mortality rates of 50% and 47.4%, respectively.
Another possibility for inducing myocardial ischemia is through obturation of a coronary artery with a balloon. We did not use this method because of the technical difficulties posed by the rabbit’s anatomical system. This method can be used easily in a pig model (16). We did not consider the method of cryogenic damage of the heart muscle, because changes that are induced in this way do not correspond in their intensity and dynamics with the effect of occlusion at the myocardial infarction in humans. The irreversible character of the local vessel support damage does not create a favorable setting for the homing of stem cells (13). We cannot deny a certain advantage of this method, however, inasmuch as the scar formed after the cryogenic damage is homogenous and has constant size as well as localization. It enables proper sample collection by its strong demarcation. The necrotic part does not expand, and the results are not variable when compared with the ligation method (17). The model of inducing myocardial infarction over 24 hours creates a certain possibility to reduce the early mortality. Orlic et al. (18) state that mortality increased dramatically between the third and sixth day after the MI induction and that it was at 40% on the fourth day. We achieved minimization of mortality with the duration of ischemia (18–24 hours) followed by development of the histopathological changes. We did not use the model with reperfusion after 1 hour of ligation mentioned in the literature (19) because we feared interference with the results from possible reperfusion toxicity (negative effect of the oxygen reactive forms).
Antiarrhythmics and resuscitation were not used because of the necessity to standardize the protocol, i.e., it would be difficult to filter out the effects of the resuscitation duration, protective influence of the beta-blocker and other medications on the evaluated results.
The histopathological changes (extension and undulation of the muscle fibers, loss of staining ability, hypereosinophilia of cytoplasm, interstitial edema, hemorrhage, and polymorphonuclear cells infiltration) in a place below the ligation corresponded with those of an early MI period (20). The same changes—but without the presence of hemorrhages and polymorphonuclear cells infiltrate—in samples collected above the place of ligation are proof of an inflammation reaction development in the vicinity of the infarction, especially if we consider their absence in the control group.
Intramyocardial cell application directly into the focus of damage or into the periinfarction zone prevails in the literature (21–24). This does not, however, account for the potential ability of the cells to home in the infarction focus during their autologous transplantation in an active way. These cells survive in a focus passively, without their rejection. That is why we proceeded to the cell application in the coronary artery.
In spite of the experiment’s short duration, we used autologous transplantation of bone marrow cells because we hoped to eliminate any undesirable immune reactions to their administration and to reduce the influence of their allogeneic application. The easy gain of BM mononuclear cells prevailed over the time requirement for their in vitro expansion (15), although it did not play any role in our experiment.
Many techniques, especially immunofluorescence methods, can be used to identify the applied cells. We used the method of cell detection based on incorporating iron oxide nanoparticles because of the possibility to use the model during noninvasive monitoring of cell therapy in the course of the myocardial infarction, which is not possible with the immunofluorescence methods. No increased toxicity is indicated for iron oxide, the marker we used (25, 26), and, moreover, these compounds can be identified by light and electron microscopy. The questions of the correct concentration of the iron oxide nanoparticles in the suspension, duration of incubation, and possible combination of the iron oxide nanoparticles with transfection agents were tested separately.
It was necessary to define the positive finding through semiquantitative evaluation of the cell migration. We defined it as the presence of cells without damaged or characteristic morphology, as did Orlic et al. (18). We tried to eliminate the false positive reactions, which result from the histological sample processing and arise from washout of the nanoparticles from the cells. Jackson et al. (27) took every cell or cluster of cells as a positive finding. The technique of staining for Turnbull’s blue in reaction with the ferricyanide requires free iron (Fe2+) ions. For this reason, it was easy to eliminate the false positive reactions, which could appear in the presence of the erythrocytes aggregates. Based on examination of other organs after the heart extraction, we expected no distortion of the results. Distortions could appear if the hemosiderin concentration in the organism were to be increased.
The duration of incubating the mononuclear cells with iron oxide nanoparticles was determined from their incorporation into cells. We took the median value of 80% stained cells, based on the cytological examination, as satisfactory. Considering the results, we can say that the incubation time can be shortened to 7 days for future experiments and still maintain satisfactory incorporation.
The semiquantitative evaluation of migration of cells stained by nanoparticles within the groups proved only occasional presence of the labeled cells in the heart base, as well as in the apex. Considering these results, we can state that minimal homing of the applied cells in the undamaged myocardium occurs. Poor cell retention in the P1 group, after 2 minutes of perfusion, did not differ significantly from the retention in the control group. The statistically important differences from the two sides of the ligation (above and below) in both evaluated localizations (vessels, muscle) in the P2 group, with 10 minutes of perfusion, suggest a striking tendency for the labeled mononuclear cells to migrate into the damaged place. A significant difference in total number of cells in the muscle between the places above and below the ligation in the P3 group, with 25 minutes of perfusion, is proof that the cells are able to persist in a damaged focus.
When we compare the groups by number of the labeled cells above the ligation, we see no statistically significant differences in vessels and muscle. This finding is in accordance with our hypothesis that increased retention of the labeled cells does not occur where the myocardium is undamaged. It was difficult to distinguish the areas of full necrosis from those of reversible ischemia within 24 hours after MI. More cells were found in ischemic areas, although some cells were found in the areas of full necrosis, as well. The time period was too short, however, to observe the migration of cells into the area of central necrosis. This could be confirmed by in vivo experiments.
A significant increase in the total number of the labeled cells in vessels, and also in muscle, was proven by comparison of the C and P1 groups with the P2 group (10 minutes of perfusion). This indicates that the chosen time interval of 10 minutes from application is enough for the cells to attach to the vessel endothelium (Fig. 3b) and to migrate into the area of the tissue damage. This is in concordance with data presented by Saito et al. (28) but gathered from an in vivo rat model. A significant increase in the retention of the labeled cells in muscle above the ligation place was found when comparing the C and P1 groups with the P3 group (25 minutes of perfusion). Thus, the administered cells persist in the vessels after 25 minutes of perfusion, they are not washed out from the damaged place, and they further migrate into interstitial space. Although this work cannot shed light on the underlying mechanism of implanted cells on the improvement of heart function (various mechanisms are thoroughly discussed in the literature, including stem cell plasticity and trans-differentiation), it is likely that at least a part of this effect depends upon cytokines and proangiogenic factors released from BM mononuclear cells, as shown by Kupatt et al. in an example of mice embryonic endothelial progenitor cells (29).
Conclusion
We verified the possibility of inducing sub-lethal myocardial infarction by ligation of the coronary artery in a rabbit. We succeeded in inducing the myocardial infarction in all animals in the range, allowing us to evaluate migration of the applied mononuclear cells. If the animals survived the first 24 hours, this was considered satisfactory for our work.
The nanoparticles-labeled BM mononuclear cells proved the ability to attach to the endothelium of blood vessels and to migrate into the myocardium interstitial space while showing significantly higher affinity for the tissue that was damaged by the infarction.
Iron Oxide Nanoparticles-Labeled BM Cells in the Myocardium Above and Below the Ligation a
Wilcoxon Rank Sum Test P Values. Comparison of Labeled BM Cells Found Below the Ligation (MI Area) in 15 Random Visual Fields for Each Group*
Coronary arteries in a rabbit. a) The scheme of left coronary artery branching in a rabbit. Place of usual artery ligation is shown. 1, ramus interventricularis paraconalis a. coronariae sinistrae; 2, ramus collateralis proximalis; 3, ramus collateralis distalis; 4, ramus circumflexus a. coronariae sinistrae; 5, ramus proximalis ventriculi sinistri; 6, ramus marginis ventricularis; A, aorta; B, truncus brachiocephalicus; C, a. subclavia sinistra; D, truncus pulmonalis; E, v. cava caudalis. b) Eosin-stained ex vivo perfused heart. Blue arrow shows the ligation site, blue dot shows the center of ischemic region of left ventricle. c) Place of the ligation shown (blue dot) on the picture taken during surgery.
Iron nanoparticle-labeled BM cells detected in the myocardial tissue. Light microscopy, hematoxylin-eosin and Turnbull’s blue staining. a) Objective =20×. Longitudinal cut through myocardium, muscle striation is retained. Labeled cell (stained blue) are found inside the capillaries and in the interstitial space next to the myocytes (arrows), as well. ×200 enlarged. b) Objective=40×. Tangential cut through arteriola in ischemic region with erythrocytes and several labeled cells (stained blue) inside the vessel (arrow). Striations of myofibrils are untouched. ×400 enlarged. c) Objective =40×. Arteriola in transversal cut through ischemic myocardium. Its interior is almost occupied by the blue stained cell (arrow). ×400 enlarged.
Electron microphotographs of nanoparticles-labeled BM cell. a) Electron microphotograph of BM cell fully occluding the blood vessel. Two of the deposits of iron nanoparticles inside the cell are shown by arrows. b) Microphotograph of nanoparticles-labeled BM cell in tangential cut attached to the interior of the capillary wall at three points. The areas of cell contacts are shown by arrows. c) Electron microphotograph of BM cell (arrow) loaded with iron nanoparticles, partially occluding the blood vessel in a transversal cut (compare with Fig. 2c). d) Microphotograph of nanoparticles-labeled BM cell in the interstitial tissue (arrow).
Time kinetics of the labeled BM cells retention and homing, depending on the perfusion time. Comparison of total labeled cell count for each group C, P1, P2 and P3. a) Non-ischemic myocardium above the ligation site. b) Ischemic zone of myocardial infarction, below the ligation site. c) Total cell count from both locations.
Comparison of labeled BM cell counts found in the blood vessels, in the interstitial space, and total count according to perfusion time. The data are shown for each group separately for the region above and below the ligation. Box-and-whiskers plot.
Footnotes
2
These authors contributed equally to this work.