Annals of Clinical & Laboratory Science 36:333-340 (2006)
© 2006 Association of Clinical Scientists
Experimental Ischemic Cardiomyopathy: Insights into Remodeling, Physiological Adaptation, and Humoral Response
Gabriella Agnoletti1,
Anna Cargnoni2,
Laura Agnoletti2,
Marco Di Marcello3,
Piera Balzarini4,
Evasio Pasini2,
Gianluca Gitti2,
Patrizia Martina2,
Roberta Ardesi2 and
Roberto Ferrari5
1 Necker Enfants Malades, Paris, France; 2 Cardiovascular Pathophysiology Center, Salvatore Maugeri Foundation, IRCCS, Gussago, Italy; 3 Veterinary Hospital, Cellatica, Brescia, Italy; 4 II Pathology Institute, University of Brescia, Brescia, Italy; and 5 Department of Cardiology, University of Ferrara, Ferrara, Italy
Address correspondence to Gabriella Agnoletti, M.D., Pediatric Cardiology Service, Groupe Hospitalier Necker Enfants Malades, 149 rue de Sèvres, 75743, Paris, France; tel 33 144 494356; fax 22 144 495724; e-mail gabriella. agnoletti{at}nck.ap-hop-paris.fr.
 |
Abstract
|
|---|
Ligation of the left anterior descending coronary artery (LAD) is used to induce experimental myocardial infarction (MI). Most previous studies have focused on the early postoperative period, while data on mid-term follow-up are scanty. This study examined the mid-term effects of LAD ligation in 95 MI rats and 28 controls. The following parameters were evaluated: systemic blood pressure (SBP), heart rate (HR), and plasma brain natriuretic peptide (BNP) level. In addition, M-mode and B-mode echocardiography, histologic examinations, and cardiac hydroxyproline assays were performed. Forty-seven perioperative and 5 late deaths were recorded. Left ventricular dilation, observed 1 mo after MI, did not progress with time. Septal thickening was similar in the 2 groups, while wall thickening was lower in the MI rats at 1 mo only. Stroke volume was diminished in MI rats, while cardiac output was depressed only at 1 and 2 mo, due to increased heart rate. SBP was unchanged and plasma BNP level was similar in the 2 groups. The infarcted area (mean ± SD) was 35 ± 10%. The ventricles in MI rats were heavier and had increased hydroxyproline content. In conclusion, these data show that LAD ligation is not only a model of acute MI, but at mid-term it provides a model of chronic ischemic dilated cardiomyopathy.
Keywords: myocardial infarction, chronic ischemia, dilated cardiomyopathy, echocardiography
|
Introduction
|
|---|
Ligation of the left anterior descending coronary artery (LAD) in experimental animals has become the method of choice to study acute myocardial infarction [1–4]. This model, although burdened by elevated mortality during the first 24 hr after coronary ligation [5], results in a low mortality rate after the acute phase and offers the possibility to study the mid-term effects of myocardial infarction (MI) on cardiac morphology, functional parameters, and humoral responses.
Most studies that have used this model have focused on the first 30–50 days after ligation [1–5]. However, morphologic and functional adaptation to MI is a process that develops over several months [6,7]. One study focused on echocardiographic aspects of coronary ligation after 3-mo follow-up, paying particular attention to measurements of left ventricular (LV) end diastolic pressure [8]. Another study examined the effects of coronary ligation in mice that were followed for 1 yr, when the presence of strong compensatory mechanisms became evident [9].
The aim of our study was to examine the mid-term effects of LAD ligation in rats, with particular attention to functional adaptation and modifications at 1 to 3 mo post-MI. We examined several echocardiographic parameters, measured plasma BNP levels to monitor the progression to heart failure (HF) [10], and determined the hydroxyproline content of cardiac ventricles as an index of fibrous substitution.
|
Materials and Methods
|
|---|
Experimental design.
All procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institute of Health (NIH Publication 85-23, Revised 1996). Ninety-five male Sprague-Dawley rats (Charles River Co., Calco, Italy) weighing 230–270 g at time of surgery and 28 age-matched controls were used. LV infarction was induced by ligation of the LAD artery under general anesthesia, as described by Fishbein et al [11]. In detail, a lateral thoracotomy was performed between the 4th and 5th intercostal space. The heart was exteriorized and a 5-0 silk suture was tightened around the proximal LAD artery at 1 mm from the atrioventricular groove. Control rats were sham-operated without coronary artery ligation. Echocardiography was performed before and 1, 2, and 3 mo after ligation. Systemic blood pressure (SBP) and heart rate (HR) were measured at the same times. Blood samples for plasma BNP assay were obtained during the echocardiographic examination. Twenty-five infarcted rats were killed at 1 mo after LAD ligation and the cardiac chambers were weighed. The remaining infarcted rats and the controls were killed at 3 mo after surgery; the cardiac chambers were weighed, histological examinations were performed, and ventricular hydroxyproline content was measured.
Echocardiography.
Echocardiographic measurements were made under mild anesthesia (pentothal, 36 mg/kg, ip) with spontaneous ventilation. We used a commercially available echocardiograph (300S Pandion Vet, Esaote, Genova, Italy) equipped with a 10 MHz phased array transducer. Measurements were made by 2 operators blinded to the rat’s hemodynamic status, with a mean of 5 consecutive measurements. The chest was shaved and the rat placed in supine position on a heating pad. A single channel electrocardiogram was obtained on the imaging system. Right and left decubitus were used to obtain the following views: right long axial, right short axial at the level of papillary muscles, mitral cordae, and mitral leaflets, right short axial at the level of left atrium, aortic root, and pulmonary valve, subcostal, and left long axial view [12,13]. LV internal diameters were measured as outlined by the American Society of Echocardiography [14].
The following parameters were measured: end diastolic/ systolic interventricular septum thickness, end diastolic/ systolic LV diameter and length, end diastolic/systolic posterior wall thickness, and end diastolic distance between the anterior mitral leaflet and the interventricular septum (E-point septal separation). The following parameters were calculated: interventricular septal thickening (%), posterior wall thickening (%), fractional shortening (FS, %), sphericity index, and ejection fraction (EF, %). The sphericity index was calculated by the following formula: index = (LV diastolic length – LV diastolic diameter) / LV diastolic length. EF was calculated by a version of Simpson’s biplane analysis [14].
Pulsed Doppler was performed to register the aortic, mitral, and tricuspid flows. Cardiac output and cardiac index were calculated by previously published methods [15]. Color flow mapping allowed evaluation of any incompetence of the atrioventricular valves. HR was measured by a single channel electrocardiogram and/or by calculating the interval between 2 subsequent cardiac cycles at pulsed Doppler. Doppler and oscillometric measurement of SBP was performed during the echocardiographic examination, utilizing an echograph (Parks Medical Electronics Inc., Aloha, OR, USA) equipped with an 8 mHz probe applied on the femoral artery, as described by Haberman et al [16].
BNP determination.
One ml of blood was collected from the caudal vein under sedation, during echocardiographic examination. Plasma brain natriuretic peptide (BNP) concentration was determined by an immunoenzymatic method (EIA Kit, Phoenix Pharmaceuticals Inc., Belmont, CA, USA). The minimal detectable concentration of BNP was 0.43 ng/ml; the upper limit of normal was 2.35 ng/ml, expressed as the mean + 2 SD in our control rats.
Hydroxyproline determination.
Hydroxyproline was assayed in cardiac tissue from surviving infarcted rats and controls at the end of the study. The ventricles were excised, weighed, and stored at –80°C. Hydroxyproline was determined by Berg’s method [17], as modified by Morais-Lopes et al [18]. The normal hydroxyproline content in the LV was <4µg/mg of dry weight (dw) and in the right ventricle <8 µg/mg dw.
Histological examination.
Rats were deeply anaesthetized and sacrificed. Any pleural and peritoneal effusion was collected by needle aspiration and measured. The heart was excised and weighed. The left atrium, right atrium, left ventricle plus septum, and right ventricle were separated and weighed. For histological examination, tissues were fixed in 10% neutral buffered formaldehyde. Each cardiac chamber was embedded in paraffin and 4 µm transverse sections were stained with H&E, as well as Gomori’s trichrome stain [19]. Sections corresponding to the short-axis echographic view were projected and the infarct size was estimated by measuring the percentage of the total endocardial circumference that was replaced by scar tissue [3]. Slides were examined by light microscopy (Eclipse E600 microscope; Nikon Italia, Florence, Italy); images were captured by a Nikon digital camera and quantified using Nikon ACT 1 2.11 image analysis software. Ventricles were processed to measure their hydroxyproline content. When possible, hearts of rats that died during the study were excised and analyzed in the same manner.
Statistics.
Results were expressed as mean ± SE. Two-way ANOVA was used for repeated measures, followed by Bonferroni’s t-test for group comparisons. Evaluations of cardiac output and stroke volume used repeated-measures ANOVA. Linear regression curves and correlation coefficients were obtained by the least-squares method. Statistical analyses were performed with GraphPad Prism software (version 3.02). A p value <0.05 was deemed significant.
|
Results
|
|---|
Forty-seven animals died during the initial 24 hr after LAD ligation (50% perioperative mortality). Among 48 survivors, 5 late deaths (at 8, 13, 20, 31, and 59 days post-surgery) were likely due to congestive heart failure (HF) (10% late mortality). It was feasible to perform histological examinations of the heart in only 2 of these cases. Among 28 sham-operated controls, no early or late deaths occurred. Heart rate (HR) and systolic blood pressure (SBP) data are reported in Table 1
. The HR was higher in the infarcted rats than the controls only at 3 mo after LAD, whilst SBP was similar in both groups during the entire study.
Longitudinal evaluation of echocardiographic parameters.
Diameters of aortic annulus, pulmonary annulus, and aortic root increased similarly during the study in both groups. Systolic and diastolic LV lengths increased significantly in infarcted rats compared to controls. Systolic and diastolic LV diameters were significantly higher in infarcted animals during the entire study, but the increment did not progress after 1 mo. LV fractional shortening was less in infarcted rats than controls during the whole study and did not change over time (Table 2). Sphericity index was significantly higher in infarcted rats only at 1 mo after ligation (Fig. 1), confirming that ventricular dilation did not progress. Accordingly, E point septal separation was higher in infarcted rats during the entire study, but did not increase with time. Diastolic interventricular septal diameter of infarcted rats, systolic diameter, and percent interventricular septal thickening were similar to those of controls and did not change during the study. The LV posterior wall was thinner in diastole and systole in infarcted rats compared to controls only at 1 mo after ligation. Similarly, the percent posterior wall thickening was lower in infarcted rats at 1 mo after ligation and increased thereafter, albeit not significantly.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 1. Sphericity index in infarcted and control rats at baseline and at 1, 2, and 3 mo after surgery.
|
|
Doppler and color Doppler evaluation.
In all rats, it was possible to register the aortic flow and calculate cardiac output. Aortic incompetence was never present. Five rats developed mild to moderate mitral incompetence at 1 mo after ligation and this remained stable throughout the study. Two rats showed associated tricuspid regurgitation. Stroke volume and cardiac output data are shown in Fig. 2. Stroke volume significantly increased over time. In infarcted rats, the stroke volume was lower during the whole study. Cardiac output, although lower in infarcted rats during the entire study, was significantly reduced only at 1 and 2 mo after ligation, due to increase of HR.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2, Stroke volume and cardiac output in infarcted and control rats at baseline and at 1, 2, and 3 mo after surgery.
|
|
Longitudinal evaluation of body weight and cardiac chamber weight.
Body weight increased similarly in both groups. The weight of cardiac chambers increased during the study in both groups; heart weight was significantly higher in infarcted rats than controls at 3 mo after ligation. LV weight was significantly higher in infarcted rats only at 3 mo after surgery. Right ventricular and left atrial weights were higher in infarcted rats at 3 mo. Right atrial weight was similar in the 2 groups during the entire study (Table 3). Significant negative correlations were found between shortening fraction and heart wt (r = –0.343; p = 0.049) and between EF and heart wt (r = –0.37; p = 0.047).
View this table:
[in this window]
[in a new window]
|
Table 3. Longitudinal evaluation of body weight and cardiac chambers weight in infarcted and control rats (mean ± SE).
|
|
Longitudinal evaluation of plasma BNP levels.
Plasma BNP concentrations remained stable in infarcted rats and were similar to controls. BNP levels in infarcted rats at 0, 1, 2, and 3 mo averaged 1.34 ± 0.56, 1.50 ± 0.55, 2.18 ± 1.71, and 1.35 ± 0.85 ng/ml, respectively. The corresponding plasma BNP levels in controls averaged 1.34 ± 0.56, 1.47 ± 0.55, 1.06 ± 0.31 and 1.88 ± 0.32 ng/ml. Only 2 rats with thoracic effusion showed plasma BNP levels that exceeded the upper normal limit.
Macroscopic examinations.
Two of the 18 rats that survived 3 mo after surgery had measurable pleural effusion (400 µl); none had peritoneal effusion. The heart occupied most part of the thoracic cavity and had a round shape. Remodeling was evident at macroscopic examination in most rats. Transverse sections of the heart showed thinning of the infarcted area and marked hypertrophy of the remaining portions of the LV (Fig. 3), leading almost to LV cavity obliteration at the apical level. Scar tissue and hypertrophy involved the right ventricle in 50% of the cases.
View larger version (125K):
[in this window]
[in a new window]
|
Fig. 3. Panel A: The heart occupies most of the thoracic cavity and has a round shape. Panel B: Macroscopic examination shows cardiac remodelling.
|
|
Histological examinations.
Collagen was visualized by its green color on histological sections stained by the Gomori trichrome technique. Fibrosis was diffusely present in transvere sections of the heart and was seen in the anterior wall, with extension to the posterior and right ventricular walls (Fig. 4). The infarcted area, expressed as fibrotic area/normal myocardium, averaged 35 ± 10% (range from 15 to 70%). Negative correlation was found between the infarcted area and the EF at 3 mo after surgery (r = –0.9, p <0.0005).
View larger version (145K):
[in this window]
[in a new window]
|
Fig. 4. Fibrosis and collagen deposition are shown by Gomori’s trichrome staining (green color) in transverse sections of the hearts from control and infarcted rats. Panel A: control right ventricle (x100); panel B: infarcted right ventricle (x100); panel C: Control left ventricle (x40); panel D: infarcted left ventricle (x40).
|
|
Hydroxyproline content.
Cardiac hydroxyproline content was significantly increased in the LV and septum of infarcted rats compared to controls (9.3 ± 1.9 vs 2.5 ± 0.1 µg/mg, dw, p <0.02) and in the right ventricle of infarcted rats compared to controls (8.0 ± 0.6 vs 6.2 ± 0.3 µg/mg, dw, p <0.05).
|
Discussion
|
|---|
Previous studies have examined the physiological changes in rats after LAD ligation [1–9]. We focused our attention on the changes at a mid-term period after MI, in order to understand better the mechanisms of adaptation to impaired LV function. It is known that, after the acute phase of MI, fibroblast proliferation and collagen deposition occur in the long-term. Depending on the species, the whole process is completed within weeks to months [11]. Before and during the period of reabsorption of necrotic tissue, there is extensive deposition of collagen, with thinning and elongation of the infarcted region [20]. At 2 mo after coronary ligation, we observed that (a) ventricular remodeling prevented the development of overt HF, (b) ventricular dilation did not progress, (c) SBP was maintained, and (d) the rats had normal survival and growth. In accordance with previous data, early mortality in our study was 50% and no deaths occurred after the second mo after surgery. That only a small minority of our rats developed pleural effusion reflects the fact that scar area rarely involved more than 30% of the LV circumference.
The increment of heart weight at 3 mo after coronary ligation was due to increases in weights of both ventricles and left atrium. At 1 mo after LAD, the weights of cardiac chambers in the infarcted animals were not higher than controls, likely because thinning and elongation prevailed over hypertrophy, in accordance with other studies [3]. It must be emphasized that significant negative correlation was found between heart weight and systolic function, suggesting that ventricular remodeling was more pronounced when ventricular function was severely depressed.
Echocardiographic data showed that normal growth of the heart occurred in infarcted animals. Although previous studies demonstrated progressive enlargement of cardiac chambers and inadequate hypertrophy during the first 6 weeks after MI [1–3], we found that, at mid-term follow-up, ventricular dilation did not progress. We observed that LV diameter was maximally dilated at 1 mo after ligation and that the dilation did not increase over time, leading to normalization of the sphericity index at 2 mo after coronary ligation. Systolic and diastolic septal thickness and percent septal thickening were maintained during the entire study, most likely due to remodeling and inotropic activation. Similarly, LV posterior wall thickening was depressed only at 1 mo after ligation.
Taken together, our data indicate that dramatic changes in ventricular geometry occurred at 1 mo after coronary ligation; during the following 2 mo, progressive development of compensatory hypertrophy of non-infarcted areas contributed to the maintenance of LV systolic function. After the second mo, the heart disease had a chronic course; no mortality occurred and no further echocardiographic changes were evident, at least until the third mo of disease. We found that infarcted rats had diminished stroke volume. This does not agree with previous studies that demonstrated, in the early post-MI period, normal stroke volume but decreased EF [3,5]. Stroke volume could be maintained through acute distension of the viable myocardium as well as augmentation of inotropic activity through adrenergic receptor stimulation [5,8]. However, these mechanisms are inadequate when the non-contractile region involves more than 20% of the LV circumference [21].
Our results can be explained either by a relatively higher infarction area or, alternatively, by the timing of stroke volume measurements, which were made after the acute phase of the disease. This is consistent with the observation that, 1 mo after MI, the HR was not significantly higher and the SBP was not significantly lower than in controls. On the other hand, in the later phase of the disease, cardiac output was maintained through an increment of chronotropic activity [3,8]. Thus, we conclude that in the early phase after MI, vasoconstrictor response prevailed over chronotropic activation, preventing the development of overt congestive HF, at least at 3 mo follow-up. Doppler data, in accordance with previous reports [3], detected mitral incompetence in only a minority of infarcted rats and it was never severe. The absence of severe mitral incompetence together with the presence of moderately dilated left atrium could explain our data concerning plasma BNP.
To our knowledge, no data have been previously reported for plasma BNP levels in rats after acute MI induced by coronary ligation. In rat models of high volume HF, circulating BNP is elevated either in the presence or absence of overt HF, whereas ventricular BNP is activated only when overt HF develops [22]. In men, plasma BNP levels are correlated with end diastolic LV pressure and can be measured to follow the clinical progression [23]. The observation that most of our rats with MI had normal plasma BNP values could confirm that, in this model, the only survivors are those animals in which functional and humoral activations occur that prevent the development of overt HF. Alternatively, we might conceive that in this model, vasoconstrictor mechanisms aimed at maintaining SBP prevailed over the natriuretic and vasodilator response to acute ischemic damage.
In rats following LAD, the hydroxyproline content increased in both ventricles. Moreover, fibrous substitution associated with collagen deposition was always present, confirming the chronic stage of the disease and explaining why the rats could not maintain stroke volume at a mid-term follow-up [20].
In conclusion, our data demonstrate that coronary ligation in rats is not only a model of acute myocardial ischemia, but, after 3 mo of disease, can serve as a model of chronic ischemic dilated cardiomyopathy. Thus, this model can be used to evaluate therapeutic options aimed at preventing or reducing the post-infarction remodeling process, such as trials of drugs, cell therapy, or percutaneous surgical palliation.
|
Acknowledgments
|
|---|
The authors thank Professor Piergiovanni Grigolato for valuable suggestions and Dr. Alessandro Bettini for editorial assistance.
|
References
|
|---|
- Baily RG, Lehman JC, Gubin SS, Musch Ti. Non-invasive assessment of ventricular damage in rats with myocardial infarction. Cardiovasc Res 1993;27:851–855.[Abstract/Free Full Text]
- Sjaastad I, Sejersted OM, Ilebekk A, Bjornerheim R. Echocardiographic criteria for detection of postinfarction congestive heart failure in rats. J Appl Physiol 2000;89:1445–1454.[Abstract/Free Full Text]
- Litwin SE, Katz SE, Morgan JP, Douglas PS. Serial echocardiographic assessment of left ventricular geometry and function after large myocardial infarction in the rat. Circulation 1994;89:345–354.[Abstract/Free Full Text]
- Solomon SD, Greaves SC, Rayan M, Finn P, Pfeffer MA, Pfeffer JM. Temporal dissociation of left ventricular function and remodeling following experimental myocardial infarction in rats. J Card Fail 1999;5:213–223.[Medline]
- Pfeffer MA, Pfeffer JM, Fishbein MC, Fletcher PJ, Spadaro J, Kloner RA, Braunwald E. Myocardial infarct size and ventricular function in rats. Circ Res 1979; 44:503–512.[Abstract/Free Full Text]
- Pfeffer JM, Pfeffer MA, Fletcher PJ, Braunwald E. Ventricular performance in rats with myocardial infarction and failure. Am J Med 1984;76:99–103.[Medline]
- Sun Y, Zhang JQ, Zhang J, Lamparter S. Cardiac remodeling by fibrous tissue after infarction in rats. J Lab Clin Med 2000;135:316–323.[Medline]
- Prunier F, Gaertner R, Louedec L, Michel JB, Mercadier JJ, Escoubet B. Doppler echocardiographic estimation of left ventricular end-diastolic pressure after MI in rats. Am J Physiol Heart Circ Physiol 2002;283:H346–352.[Abstract/Free Full Text]
- Pons S, Fornes P, Hagege AA, Eudes D, Giudicelli JF, Richer C. Survival, haemodynamics and cardiac remodeling follow up in mice after myocardial infarction. Clin Exp Pharmacol Physiol 2003;30:25–31.[Medline]
- Jiang W, Cai DY, Pan CS, Qi YF, Jiang HF, Geng B, Tang CS. Changes in production and metabolism of brain natriuretic peptide in rats with myocardial necrosis. Eur J Pharmacol 2005;507:153–162.[Medline]
- Fishbein MC, Maclean D, Maroko PR. Experimental myocardial infarction in the rat: qualitative and quantitative changes during pathologic evolution. Am J Pathol 1978;90:57–70.[Abstract]
- Watson LE, Sheth M, Denyer RF, Dostal DE. Baseline echocardiographic values for adult male rats. J Am Soc Echocardiogr 2004;17:161–167.[Medline]
- Bollano E, Waagstein F, Omerovic E. Stress echocardiography using trans esophageal atrial pacing in rats. J Am Soc Echocardiogr 2003;16:326–332.[Medline]
- Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH, Tajik AJ. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989;2:358–367.[Medline]
- Oh JK. The ECH Manual, 2nd ed. Lippincott-Raven, Philadelphia, 1999.
- Haberman CE, Morgan JD, Kang CW, Brown SA. Evaluation of Doppler ultrasonic and oscillometric methods of indirect blood pressure measurement in cats. Int J Appl Res Vet Med 2004;2:279–289.
- Berg RA. Determination of 3- and 4-hydroxyproline. Meth Enzymol 1982;82:372–398.[Medline]
- Duarte de Morais Lopes J, da Silva Gomes RA, Hial V, Rezende Lopes IC, dos Reis MA, de Paula Antunes Teixeira V. Correlations between the collagen content of the human left ventricular myocardium, measured by biochemical and morphometric methods. Arq Bras Cardiol 2002;79:15–19.
- Gomori G. A rapid one-step trichrome stain. Am J Clin Pathol 1950;20:661–664.[Medline]
- Hutchins GM, Bulkley BH. Infarct expansion versus extension: two different complications of acute myocardial infarction. Am J Cardiol 1978;41:1127–1132.[Medline]
- Klein MD, Herman MV, Gorlin R. A hemodynamic study of left ventricular aneurysms. Circulation 1967;35:614–630.[Abstract/Free Full Text]
- Langenickel T, Pagel I, Hohnel K, Dietz R, Willenbrock R. Differential regulation of cardiac ANP and BNP mRNA in different stages of experimental heart failure. Am J Physiol Heart Circ Physiol 2000;278:H1 500–506.[Abstract/Free Full Text]
- Dokainish H, Zoghbi WA, Lakkis NM, Al-Bakshy F, Dhir M, Quinones MA, Nagueh SF. Optimal noninvasive assessment of left ventricular filling pressures: a comparison of tissue Doppler echocardiography and B-type natriuretic peptide in patients with pulmonary artery catheters. Circulation 2004;109:2432–2439.[Abstract/Free Full Text]
Top
Abstract
Introduction
