Abstract
The amniotic membrane has long been applied for wound healing and treatment of ophthalmological disorders, even though the mechanisms underlying its actions remain to be clarified. Recently, cells derived from fetal membranes of human term placenta have raised strong interest in regenerative medicine for their stem cell potential and immunomodulatory features. Our study aimed to investigate the possible utility of amniotic membrane to limit postischemic cardiac injury. A fragment of human amniotic membrane was applied onto the left ventricle of rats that had undergone ischemia through left anterior descending coronary artery ligation. Echocardiographic assessment of morphological and functional cardiac parameters was then performed over a 3-month period. We demonstrated that application of an amniotic membrane fragment onto ischemic rat hearts could significantly reduce postischemic cardiac dysfunction. The amniotic membrane-treated rats showed higher preservation of cardiac dimensions and improved cardiac contractile function in terms of higher left ventricle ejection fraction, fractional shortening, and wall thickening. These improvements were apparent by day 7 after application of the amniotic membrane, persisted for at least 2 months, and occurred independently of cardiac injury severity. No engraftment of amniotic cells was detected into host cardiac tissues. Our results suggest that use of amniotic membrane may constitute a convenient vehicle for supplying cells that produce cardioprotective soluble factors, and reinforce the notion that this tissue constitutes a cell source with clinical potential that has yet to be completely revealed.
Introduction
Myocardial infarction causes an irreversible loss of cardiomyocytes, which, together with the inability of remaining tissue cells to adequately compensate for the injury, results in impaired cardiac function and possible heart failure.
In the last decade, a number of studies have been performed both in animal models and in clinical trials to assess if and to what extent cell therapy may be beneficial in repairing infarcted hearts (31,36). The obvious primary intent of cell therapy would be to regenerate the injured myocardium, and several studies have been undertaken in order to investigate the ability of adult stem cells to differentiate into cells with cardiomyocyte-like features. Although the results obtained are still subject to debate, some authors have demonstrated cardiomyogenic differentiation potential for mesenchymal stromal cells derived from the bone marrow (11,23,47,52) as well as for adipose tissue-derived cells (13,29,54).
In parallel, a variety of cell types, including fetal cardiomyocytes (37), skeletal myoblasts (10,17), resident myocardial progenitors (6), bone marrow cells (34, 48), adipose tissue-derived cells (29), and umbilical cord blood cells (19,30), have been applied in preclinical studies aimed at investigating the efficacy of these cells in cardiac repair.
Interestingly, most groups have reported that cell therapy can improve postischemic cardiac function, although such effects do not appear to be associated with cardiac differentiation of the transplanted cells. Therefore, the current predominant hypothesis regarding the mechanism underlying these beneficial effects is that, rather than regeneration of the damaged myocardial tissue, progenitor and stem cells promote repair of heart injury through paracrine secretion of growth and survival factors that induce neovascularization (18), reduce scar formation, and improve myocardial contractility (14,36).
Among the different types of cells with potential clinical applicability for cell therapy, fetal stem cells have recently gained particular interest in the field of cell transplantation and regenerative medicine for several important reasons, including their low immunogenicity, which has been demonstrated through both in in vitro (21,22) and in vivo studies (2,26,55), as well as for their ability to engraft in allogeneic and xenogeneic hosts (2,5,25). Preclinical studies have shown that transplantation of xenogeneic and allogeneic fetal membrane-derived cells results in a reduction in bleomycin-induced lung fibrosis in a mouse model (4) and improvement of spinal cord injury in monkeys (39), while potential applications for these cells have also been suggested for ischemic stroke (53) and treatment of liver diseases (27,38,46). Furthermore, placental fetal membrane- and amniotic fluid-derived cells have progenitor/stem cell potential (7,35) and, even though their ability to differentiate toward cardiomyocytes is still debated (40,50,55), a promising study by Ventura et al. (50) has recently reported improved myocardial function after transplantation of fetal membrane-derived cells into infarcted rat hearts.
On the other hand, amniotic membrane patches have long been applied in ophthalmology and wound healing due to their anti-inflammatory and antiscarring properties (9).
Therefore, considering that amniotic membrane patches could act both as a source of fetal stem cells and as a vehicle for local cell delivery to ischemic areas, thus avoiding the risks of intramyocardial cell inoculation (12), we have hypothesized that their use after myocardial infarction could represent a potential new strategy for treatment of ischemic injuries.
In this study, we demonstrate that application of amniotic membrane fragments to infarcted rat hearts significantly reduces postischemic cardiac dimensional alterations and improves myocardial function for up to at least 60 days after ischemia induction.
Materials and Methods
Amniotic Membrane
Human term placentas were obtained with maternal consent according to the guidelines of the Ethical Committee of the Catholic Hospital (CEIOC). The amnion was immediately peeled from the chorion and washed extensively in phosphate-buffered saline containing 100 U/ml penicillin and 100 μg/ml streptomycin. The amnion was then cut into 5 × 5-cm fragments, which were stored separately at room temperature, as reported by Hennerbichler et al. (15), in 50-ml vials filled with serum- and phenol red-free DMEM in sterile conditions. The amniotic fragments were used within 24 h. We had tested the number and viability of cells in amniotic membranes which had been stored for 24 h as described, observing that these parameters were not significantly reduced compared to those that were seen using fresh membrane (unpublished data).
Myocardial Infarction Model
All animal experiments were carried out in accordance with current Italian and European regulations and laws on the Use and Care of Animals for research (DL. 116/27 January 1992).
Seventy-seven male Sprague-Dawley rats were anaesthetized with isoflurane and, after lateral thoracotomy and heart exteriorization, left ventricle infarction was induced by permanent ligation of the left anterior descending (LAD) artery at 1 mm from the atrioventricular groove. Ischemia induction was confirmed by a paling in epicardial color and dyskinesis. The heart was then repositioned into the thorax.
For animals in the amnion treatment group, immediately after LAD ligation, an amniotic membrane fragment was gently placed onto the left ventricle with its mesenchymal side in contact with the epicardial surface. In order to fasten the amniotic membrane fragment to the heart, the excess suture that was left uncut after LAD artery ligation was passed around an edge of the amniotic membrane fragment and tightened (Fig. 1A). Finally, the amniotic membrane fragment was laid flat upon the ventricle surface (Fig. 1B). The amniotic membrane fragment was therefore applied to the ventricle surface without any direct suture on the heart. The modality that was used to keep the membrane fragment lying in place on the ventricle was to tie one of the edges of the membrane at the point of coronary ligation using the suture thread that had been used for the ligation itself, as represented in Figure 1. Therefore, the membrane fragment was only made to lie on the left ventricle without any artificial reinforcing effect. After this procedure was performed, the sternum and the skin incisions were sutured. Each rat was given perioperatory analgesia and antibiotics after surgery.
Application of an amniotic membrane fragment onto the cardiac ischemic surface. After being placed gently onto a plastic spatula, the amniotic fragment was fastened to the heart using the suture that had been left uncut after LAD artery ligation. The suture was passed around an edge of the amniotic fragment and tightened (A). Next, the amniotic fragment was laid flat upon the ventricle surface with its mesenchymal side in contact with the epicardial surface (B).
Experimental Groups
After surgery, the ischemic rats were randomly assigned to two experimental groups: Ischemia group (infarcted rats that received no treatment) and Ischemia + amnion group (infarcted rats in which amniotic membrane was applied to the ventricle surface). A group of untreated noninfarcted rats (n = 8) served as normal controls. Another group of noninfarcted rats (n = 4) was treated with amniotic membrane fragments to allow assessment of any effects of the membrane on nonischemic hearts.
Echocardiography
Seven days after surgery, a complete echocardiography study (B Mode, M Mode, Doppler, Color Doppler) was performed in the 52 surviving rats using MyLab 30 CV apparatus (Esaote, Genova, Italy) equipped with a phased array 10-5 MHz probe. Infarct size was estimated as the percentage of the myocardial akinetic area corrected for left ventricle endocardial circumference, monitored on a midventricular short axis in diastole, as reported by Sakakibara et al. (37). Rats showing detectable left ventricle akinetic areas (n = 30) continued to be monitored by echocardiography at 30, 60, and 90 days after LAD artery ligation.
Echocardiographic measurements were performed under mild anesthesia (induced with an intramuscular injection of 10 mg/kg xylazine and 50 mg/kg ketamine) and executed under continuous ECG monitoring by a blinded operator in accordance with the guidelines of the American Society of Echocardiography (41) and following previously described procedures (51).
For each rat, the following measurements were performed: complete left ventricle M Mode, EPSS (E Point Septal Separation), and left ventricle B Mode evaluation (area-long axis method), aortic annulus and root, left atrium diameter, aortic and mitral flow determination. On the basis of these measurements, the following parameters were calculated: fractional shortening, ejection fraction, septal and parietal thickening, systolic and diastolic sphericity index (the ratio between left ventricle length and the respective diameter), and stroke volume. The mitral diastolic filling pattern was also analyzed. Heart rate was measured by a single channel electrocardiogram and/or by calculating the interval between two subsequent cardiac cycles at pulsed Doppler.
Scar Size Assessment
Scar sizes were assessed in ischemic untreated and amnion-treated rats at 30 and 90 days after ischemia induction. After euthanasia, the hearts were excised and each scar size determined as a percentage of the entire left ventricle area, as reported by Nagaya et al. (32).
Assessment of Amniotic Cell Engraftment in Rat Cardiac Tissues
Immunohistochemistry
Immunohistochemical studies were performed on formalin-fixed and paraffin-embedded tissue sections using the Vector® M.O.M. Immunodetection kit (Vector, Burlingame, CA). To detect the presence of human amniotic cells, samples were immunostained with monoclonal antibodies specific for human cytokeratin 19 (CK 19, Ab-4) (clone BA17) diluted 1:50 and vimentin (clone 3B4) diluted 1:100. The primary antibody was applied for 1 h at room temperature. Novared (Vector) was used as the chromogen, and hematoxylin for counterstaining.
PCR Analysis
Total DNA (50–100 ng) was extracted from heart tissue using a Bio robot EZ1 (Qiagen, Hilden, Germany) and the EZ1 Tissue Kit (Qiagen) according to the manufacturer's instructions, and was then amplified in 50-μl reactions containing dNTPs (200 μmol) and GoTaq DNA polymerase reagents (Promega, Madison, WI, USA) and the following primers (25 pmol) specific for the human sequence of the cytochrome B mitochondrial gene: 5′-CCCATACATTGG GACAGACC-3′ (forward); 5′-GACGGATCGGAGAA TTGTGT-3′ (reverse) (43). All PCR reactions included an initial step at 95°C for 10 min. This was followed by 40 cycles (94°C for 30 s, 58°C for 30 s, 72°C for 1 min) for DNA amplification of human mitochondrial cytochrome B. PCR products were analyzed by agarose gel electrophoresis and ethidium bromide staining, followed by Southern blotting. For Southern blot analysis, PCR products were transferred to nylon membranes (Hybond N+, Amersham Biosciences, Little Chalfont, UK) and hybridized to specific probes labeled with horseradish peroxidase enzyme (ECL Gold, Amersham Biosciences). Chemiluminescent signals were detected with a Gel Doc 2000 system (Bio-Rad, Hercules, CA).
Statistical Analysis
The echocardiographic and infarct size measurements are expressed as means ± SE. The effect of treatment and time on the echocardiographic outcomes was assessed by a repeated measures analysis of variance via linear mixed effects models. Treatment and time were included as fixed effects while subjects were treated as random effect to take into account correlation among measures on the same rats. All models were adjusted for the severity of the induced ischemic damage to control for a possible confounding effect. A treatment group by time interaction term was tested to investigate possible time effect on treatment. A treatment group by severity interaction term was tested in order to evaluate if treatment effect could change across severity of differing grade. The differences among groups with regard to scar sizes were assessed by Student's t-test. Values of p < 0.05 were considered to indicate statistical significance. A correlation coefficient with respective 95% confidence interval (CI) was estimated to assess the relationship between akinetic areas measured at day 7 and the scar sizes measured 90 days after LAD ligation. All statistical analyses were performed using the R software (version 2.7.0).
Results
Myocardial Injury Induced by Left Anterior Descending Coronary (LAD) Artery Ligation
LAD artery ligation was performed in 77 rats, 25 of which died within 24 h following surgery. In order to ensure the presence of a myocardial defect in the surviving animals, we selected those rats that showed a detectable cardiac akinetic area at day 7 after surgery. Thirty rats fit this criterium, with akinetic areas ranging from 11.4% to 36.1%. Of these, 14 belonged to the Ischemia group and 16 to the Ischemia + amnion group. Rats from both groups showed cardiac akinetic areas that were not statistically different (19.83 ± 2.41% vs. 19.88 ± 2.27%, respectively). Two late deaths occurred: one in the Ischemia group at day 60 and one in Ischemia + amnion group at day 89 after LAD artery ligation.
In order to demonstrate the reliability of the echocardiographically determined cardiac akinetic area as an index of the severity of the induced ischemia, the correlation between the akinetic areas measured at day 7 and the scar sizes measured post mortem 90 days after LAD artery ligation was assessed in rats from the Ischemia group. The calculated correlation coefficient of 0.75 (95% CI: 0.33–0.92) confirmed the validity of the approach, in line with previously reported data (28).
According to Sakakibara et al. (37), the echocardiographically evaluated cardiac akinetic areas were used both to select rats to be included in the study, as well as to classify the severity of the ischemic injury induced in those rats.
Rats' Distribution With Respect to Ischemia Severity Categories
Effects of Amniotic Membrane on Ischemia-Induced Cardiac Dimensional Changes
Cardiac dimensions were evaluated in all rats, ischemic or not, amnion-treated or untreated, at four time points: 7, 30, 60, and 90 days. As expected, in Ischemia group rats, LAD artery ligation produced alterations in cardiac geometry with high systolic and diastolic left ventricle diameters and lengths (Fig. 2A, B, Table 2); low systolic and diastolic sphericity indexes, indicating that ischemic ventricles displayed more spherical shapes than those of control hearts (Fig. 2C, D); thin left ventricle walls were associated with ischemic areas (Fig. 2E, F), indicative of loss of ventricle muscle in these areas; high values of E point septal separation (Fig. 2G), indicative of increased left ventricle cavities and reduced mitral inflow; and high left atrium diameters (Fig. 2H). No changes were observed in the interventricular septum (IVS) thickness (Table 2).
Echocardiographic cardiac dimensional assessment in amniotic membrane-untreated and treated ischemic rats. Values of left ventricle diameters, in systole (A) and in diastole (B); left ventricle sphericity indexes, in systole (C) and in diastole (D); left ventricle wall thickness, in systole (E) and in diastole (F); E point septal separation (EPSS) (G) and left atrium diameter (H), which were recorded in the Ischemia and Ischemia + amnion groups at days 7, 30, 60, and 90 after LAD artery ligation, are reported. Control group is regarded as the reference healthy control. n = 14 rats are included in the Ischemia group (n = 13 at days 60 and 90 because one death occurred); n = 16 in the Ischemia + amnion group (n = 15 at day 90 because one death occurred), and n = 8 in the Control group (n = 5 at day 7). Statistical analysis was performed by a repeated measures ANOVA adjusted for injury severity. The p-values shown in the lower left corner of individual panels indicate either significant (p < 0.05) or nonsignificant (p > 0.05) differences in parameter values between Ischemia + amnion and Ischemia groups at all reported time points. The p-values reported within the diagram indicate instances where the effects of treatment with amniotic membrane were significant only at specific time points. Echocardiographic Parameters Monitored in Nonischemic (Control) and in Ischemic Rats Treated (Ischemia + Amnion Group) or not (Ischemia Group) With Amnion LV SD, LV systolic diameter; LV DD, LV diastolic diameter; IVS ST, interventricular septum systolic thickness; IVS DT, interventricular septum diastolic thickness; IVS thick., IVS thickening; SLVWT, systolic LV wall thickness; DLVWT, diastolic LV wall thickness; LVAW thick., LV wall thickening; FS, fractional shortening; EF, ejection fraction; EPSS, E point septal separation; LV SL, LV systolic length; LV DL, LV diastolic length; LA diameter, left atrium diameter; HR, heart rate. p < 0.01 versus Ischemia Group at same time point.
Application of a fragment of amniotic membrane onto the cardiac ischemic injury resulted in better preservation of cardiac dimensions with respect to untreated ischemic rats. Compared to the Ischemia group, the Ischemia + amnion group showed significantly lower systolic left ventricle diameters (p < 0.05) (Fig. 2A); higher systolic left ventricle sphericity indexes (p < 0.01 at day 60) (Fig. 2C); and markedly higher left ventricle wall thickness values, either measured in systole (p < 0.05) (Fig. 2E) or in diastole (p < 0.05) (Fig. 2F). The Ischemia + amnion group also showed lower EPSS (p < 0.05) (Fig. 2G) and lower left atrium diameters (p < 0.05) (Fig. 2H). Treatment with amniotic membrane did not result in any statistically significant effects on left ventricle diameters and sphericity indexes measured in diastole (Fig. 2B, D).
Effects of amniotic membrane on cardiac ischemia-induced dimensional changes were already evident at the first echocardiographic monitoring performed 7 days after ischemia induction, and were maintained up to day 60. However, at day 90, no appreciable differences were recorded among Ischemia and Ischemia + amnion groups. Interestingly, treatment with amniotic membrane resulted in beneficial effects on cardiac ischemia-induced dimensional changes, and these effects were not influenced by injury severity of different degrees.
In order to explore whether amniotic membrane could exert effects independently of ischemia induction, amniotic membrane fragments were also applied to the left ventricle surface of non ischemic rats (n = 4). Compared to healthy controls, these rats showed no appreciable changes at any time points in the echocardiographically evaluated dimensional parameters (data not shown).
Effects of Amniotic Membrane on Ischemia-Induced Left Ventricle Dysfunctions
The infarct produced a rapid myocardial functional failure that was characterized by both a systolic deficit, as shown by a decline in left ventricle fractional shortening (from 35.5 ± 1.6% to 23.8 ± 2.4% in control and in Ischemia group rats, respectively, at day 7 after ligation) (Fig. 3A) and ejection fraction (from 69.8 ± 2.0% to 51.4 ± 4.1%, at day 7) (Fig. 3B); and a diastolic deficit in terms of alterations in the pattern of mitral inflow, detected as increments in the early filling velocity (E velocity) (Table 2) and in the ratio of early to late filling velocities (E/A) (Fig. 3D). In particular, except for some rats at day 90, E/A measurements in control rats never exceeded the value of 2.0, which has been proposed by some authors as a threshold value for restrictive pattern occurrence (28). On the contrary, Ischemia group rats showed E/A values over this threshold at days 7, 30, and 60, with some rats also showing higher values at day 90.
Echocardiographic cardiac functional assessment in amniotic membrane-treated and untreated ischemic rats. Values of left ventricle fractional shortening (A), left ventricle ejection fraction (B), left ventricle stroke volume (C) and E-wave velocity, A-wave velocity ratio (E/A) (D), which were measured in the Ischemia and Ischemia + amnion groups at days 7, 30, 60, and 90 after LAD artery ligation, are reported. Control group is regarded as the reference healthy control. n = 14 rats are included in the Ischemia group (n = 13 at days 60 and 90 because one death occurred); n = 16 in the Ischemia + amnion group (n = 15 at day 90 because one death occurred), and n = 8 in the Control group (n = 5 at day 7). Statistical analysis was performed by a repeated measures ANOVA adjusted for injury severity. The p-values shown in the lower left corner of individual panels indicate nonsignificant (p > 0.05) differences in parameter values between Ischemia + amnion and Ischemia groups at all reported time points. The p-values reported within the diagram indicate instances where the effects of treatment with amniotic membrane were significant only at specific time points.
In our study, and in agreement with findings reported by others (20), the stroke volume in Ischemia group rats was similar to that calculated for control rats (Fig. 3C).
Application of amniotic membrane was able to preserve myocardial function and, indeed, left ventricular fractional shortening (FS%) and ejection fraction (EF%) in Ischemia + amnion group rats were higher than those recorded in Ischemia group rats at days 7 and 60 (p < 0.01) (Fig. 3A, B). At day 30, the differences observed between Ischemia + amnion and Ischemia groups for FS% and EF% were marked, although they did not reach statistical significance (p = 0.06 and p = 0.09, respectively).
In agreement with the better systolic function detected in Ischemia + amnion group rats, higher thickening values of left ventricle wall were also observed in this group (Table 2). Moreover, even though no worsening in stroke volume was observed in the Ischemia group, the values of stroke volume calculated in the treated rats were higher than those observed in the Ischemia group except for 90 days after LAD artery ligation (Fig. 3C).
Interestingly, at days 7, 30 and 60, Ischemia + amnion group rats showed lower values for the ratio of early to late mitral filling velocities (E/A ratio) compared to Ischemia group rats, although these differences were not statistically significant. At days 30 and 60, E/A values exceeding the threshold value of 2.0 were more frequently detected in Ischemia group rats (in 6/14 rats, or 42.9%) with respect to in Ischemia + amnion group rats (in 3/16 rats, or 18.8%) (Fig. 3D, Table 2). As also reported by other authors, we detected the highest E/A values in rats with severe ischemia, indicating the occurrence of severe diastolic dysfunction (20,28).
The higher degree of preservation of hemodynamic conditions in amniotic membrane-treated rats was also confirmed indirectly by the lower heart rate values recorded at days 7 and 30, in these rats, compared to Ischemia group rats (p < 0.05) (Table 2), which, during the first month, developed a transient compensatory tachycardia.
As observed for cardiac dimensional changes, treatment with amniotic membrane also provided early beneficial effects on ischemia-induced cardiac functional impairment irrespective of the different degrees of the ischemic injury induced. The better fractional shortening and ejection fraction values were measured just 7 days after amniotic membrane application. However, by day 90, myocardial systolic function was similar in Ischemia + amnion and Ischemia groups.
No cardiac dysfunctions were observed at any time points when amniotic membrane fragments were applied onto ventricle surfaces of non ischemic rats (data not shown).
Effects on Ischemia-Induced Myocardial Scarring
Scar sizes in untreated and amniotic membrane-treated ischemic rats were assessed at two different time points (30 and 90 days) after injury induction. Rats euthanized 30 days after ischemia induction (n = 14; 7 in Ischemia group and 7 in Ischemia + amnion group) belonged to a separate experimental set to that described above in which all animals were assessed for echocardiographic parameters and euthanized at day 90.
At 30 days from ischemia induction, amniotic membrane-treated rats showed scar sizes significantly smaller than those of ischemic untreated rats (p < 0.05) (Fig. 4A). At 90 days, as expected, based on the decline in observed effects of amniotic membrane on the echocardiographic parameters, scar extensions in the Ischemia + amnion group rats were not statistically different from those measured in Ischemia group rats (Fig. 4B).
Myocardial scar size assessment in amniotic membrane-treated and untreated ischemic rats. Scar sizes were evaluated in untreated and amniotic membrane-treated ischemic rats at days 30 (A) and 90 (B) after LAD artery ligation. The rats that were euthanized at day 30 after ischemia induction belonged to an experimental set from a separate study. The number of rats in each experimental group is indicated in the bars. Statistical significance was assessed by Student's t-test between Ischemia and Ischemia + amnion groups (each analysis was performed between groups of the same experimental set).
Amniotic Cell Engraftment
Cell engraftment was evaluated in host cardiac tissues by immunohistochemistry and PCR analysis after 30 and 90 days from amniotic membrane application. No cell engraftment was observed at either time point (data not shown).
Discussion
This study demonstrates that, when applied to a myocardial ischemic surface, amniotic membrane has protective effects against deleterious ischemia-induced changes in cardiac contractile function and dimensions. Our study combines the advantages of a well-established and commonly applied cardiac ischemia model (LAD artery ligation performed in rats) with echocardiographic analysis, which allows reliable measurement of ischemia-induced morphological and functional cardiac alterations (1,20,28,51). Furthermore, given that this procedure is noninvasive, it allows repeated measurements to be taken for each animal over several time points.
Previous studies performed in both preclinical cardiac ischemia models and in patients suffering from ischemic diseases (31,36) have so far been aimed at investigating the efficacy of cell therapy on cardiac repair by using different cell types including bone marrow-derived cells (16,32,34,42,48,49), skeletal myoblasts (10), fetal cardiomyocytes (37), and endothelial progenitor cells (56). Interestingly, it has recently been shown that transplantation of placental fetal membrane-derived cells into ischemic rat hearts results in a reduction in cardiac scar size as well as an improvement in cardiac contractile function with less reduction of left ventricle wall thickness, when assessed 28 days after cell treatment (50).
In our study, for the first time to our knowledge, we have instead tested the effects of applying amniotic membrane patches onto rat myocardial ischemic surfaces in order to explore the potential efficacy of this tissue as a possible treatment for ischemic diseases. We have studied the effects of amniotic membrane in this model for a prolonged period and have classified the different levels of severity of cardiac injury induced by LAD artery ligation, to allow evaluation of the effects of amniotic membrane on dimensional and functional cardiac alterations of different degrees (mild, moderate, and severe).
Our results show that treatment of ischemic rats with amniotic membrane plays a major role in preserving myocardial contractile function. In fact, the echochardiographic left ventricle functional parameters and morphological measures, assessed during systole, were preserved to a significantly higher degree compared to untreated ischemic rats. In addition, amniotic patching resulted in absence of compensatory tachycardia.
Treatment with amniotic membrane not only improved the systolic function of ischemic rats, but also resulted in some positive effects on the cardiac diastolic function, as demonstrated by the better diastolic filling parameters (E/A ratio and EPSS) and the lower dilation of the left atrium recorded in these treated animals.
Beneficial cardiac effects resulting from application of amniotic membrane were already apparent by day 7 when the first echocardiography examination was performed. This is in agreement with previous studies where cell therapy was applied in cardiac ischemia models using other cell types such as bone marrow-derived cells (12,33) and skeletal muscle cells (10). The majority of these previous studies report assessment of cardiac parameters for up to 30 days after cell therapy treatment; instead, we have performed cardiac evaluations until day 90 and observed that amniotic membrane patching exerts benefits on ischemia-induced cardiac functional and dimensional alterations until day 60. The echocardiography evaluations performed at day 90 revealed no differences between amniotic membrane-treated and untreated ischemic rats, except for left ventricle wall thickness, which remained significantly higher in amniotic membrane-treated rats up until the last echo-analysis. These data could suggest that further application of amniotic membrane may be required to prolong the cardiac benefits.
Interestingly, our data have also demonstrated that cardiac protection resulting from application of amniotic membrane is independent of the severity of the injury induced by LAD artery ligation, and it is therefore tempting to speculate that amniotic membrane could have beneficial effects on a wide spectrum of ischemic pathology severities.
In this study, we also explored the possible mechanisms through which the amniotic membrane may exert beneficial effects on ischemic rats. Firstly, we considered the possibility that these effects could have been due to the engraftment of cells derived from the human amniotic membrane into the host tissues. The fact that we were unable to detect human cells in rat cardiac tissues at 1 and 3 months after application of the amniotic membrane (data not shown) suggests that engraftment of amniotic cells was either absent or very low. Therefore, it is very likely that the effects observed were instead associated with the release of soluble factors by cells of the amniotic membrane, which we have recently hypothesized as the mechanism by which placenta-derived cells may also work to reduce bleomycin-induced lung fibrosis (4). It has been reported that amniotic membrane, as well as cells derived from this tissue, are able to release cytokines that have potent immunomodulatory and anti-inflammatory effects, such as interleukin-10 and interleukin-6 (8). Furthermore, amniotic cells have also been reported to release growth factors associated with wound healing, including angiogenetic factors [vascular endothelial growth factor, angiogenin and platelet-derived growth factor (44)], as well as cell proliferation [i.e., epidermal, keratinocyte, hepatocyte, and basic fibroblast growth factors (45)] and differentiation [i.e., transforming growth factors β1, β2, and β3 (44)]. Therefore, amniotic cells may act via a paracrine mechanism, as reported for other cells that have been applied in cell therapy approaches and that have been shown to counteract cardiac injury by releasing mediators that influence the survival, differentiation, and proliferation of host tissue cells (10,14).
It is also interesting to note that, besides the factors mentioned above, the amniotic membrane also expresses high levels of thymosin-β4 (data not shown), a short peptide that has been demonstrated to accelerate wound healing (24) and, more recently, has also been shown to protect the murine heart against ischemic injury when administered exogenously (3).
It is very likely that more than one of the factors mentioned above may play a role in amniotic membrane-mediated preservation of ischemic myocardial function, and it is therefore tempting to hypothesize that amniotic membrane patches may be useful as sources for release of soluble factors. The amniotic membrane could represent a very interesting alternative to cell sheets and cardiac cell patches, which have recently been proposed as an innovative local cell delivery system for assuring homogeneous distribution of transplanted cells onto the myocardial ischemic area (29,42). In addition, the use of the amniotic membrane patching may also prolong the survival of the transplanted cells, given that the cells are embedded in a collagen-rich stromal matrix, which may preserve their integrity by preventing direct contact with ischemic and necrotic cardiac tissues that may otherwise accelerate their death.
Finally, but not of secondary importance, the use of amniotic membrane as a treatment for myocardial ischemic diseases offers the evident advantages of being in plentiful supply and being applicable immediately without the need for any cell isolation, selection, or culture steps, which makes it a low-cost approach; in addition, amniotic membrane has good preservability (15) and is immunologically tolerated in both allogeneic and xenogeneic conditions, both in vitro (21) and in vivo (2).
These considerations, together with our experimental findings, suggest that application of amniotic membranes for wound healing could be extended to cardiac ischemic injuries.
In conclusion, the application of amniotic membrane onto the ischemic region of the left ventricle is able to limit myocardial dimensional alterations and contractile dysfunction in a rat model of myocardial infarction. These effects are already evident 7 days after application of the amniotic membrane, persist for at least 2 months, and are independent of the severity of the induced cardiac injury. The benefits conferred by treatment with amniotic membrane do not seem to be related to the engraftment of human amniotic cells into the ischemic rat hearts, but they are likely due to release of soluble factors that may modulate the ischemic inflammatory process, and translate to prolongation of the survival of host tissue cells.
Footnotes
Acknowledgments
The authors thank the physicians and midwives of the Department of Obstetrics and Gynaecology of Fondazione Poliambulanza Istituto Ospedaliero, Brescia, Italy. The authors are indebted to Dr. Fabio Candotti for critically reviewing the manuscript and to Dr. Marco Evangelista for help in editing the manuscript. The authors are grateful to Paolo Locatelli and Esaote-Italia for making the echo-cardiograph available for this study. This study was in part supported by grants from Fondazione Cariplo (Bando 2005 and 2006). A European patent application has been filed with the application No. PCT2008-004845.