| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Cardiovascular Sciences, Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, Maryland 21224
Submitted 8 September 2003 ; accepted in final form 6 November 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
heart; anatomy; method; cardiovascular physiology
The objective of the present study was to improve the technique of ligation of nonvisible coronary artery branches in mice to permit the reliable creation of infarcts of specific size and location. Our hypothesis was that 1) there are specific, stereotyped anatomic variants of mouse coronary branching patterns and 2) a knowledge of specific distances from the origin of the LMDA at which branches are most likely to occur would improve the likelihood of creating an infarct of the desired size resulting from an otherwise "blind ligation" of these branches. We used latex casting to characterize left coronary artery (LCA) branching patterns in 53 mice. From these castings, we determined the probability for the branch points of three major LV branches to occur at given distances from the origin of the LMDA. In another subset of mice, we employed the information about the probability of branch point locations from the casting study to guide the placement of a ligature to create an infarction by ligating specific branches that could not be directly observed in the beating myocardium. The placement of ligatures, guided by the probability information gleaned from the prior anatomic casting experiment, was successful in creating specific small to moderately sized infarcts in the territory of two specific nonvisible branches of the LMDA in 57% of attempts. Ligation of the LMDA, by comparison, resulted in large MIs in 100% of attempts. A hemodynamic profile performed 5–8 wk after ligation showed that LV remodeling and LV function predictably varied with MI size.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Coronary artery casting. Fifty-three mice were used for coronary artery casting to analyze the branching pattern of the LCA. Under pentobarbital sodium anesthesia (43 mg/kg iv), the thoracic and abdominal cavities were opened, the descending aorta was retrogradely cannulated, the vena cava was cut open, and red latex (
100 µl) was infused into the heart after blood was flushed out with 3–6 ml of PBS (pH 7.4) containing heparin (20 U/ml). The hearts were excised and placed in neutral buffered formalin solution for 
3 days.
The fixed hearts were placed for examination, in surgical view, in a heart-shaped shallow groove of a scale plate (Fig. 1A). Examination was performed under a stereoscopic dissecting microscope (Karl Zeiss). The LMDA was examined, and its length from its origin to the apex was measured. The diameter of the heart and distances from the origin of the LMDA to the points of origin of its different branches were also measured. All distances were eventually normalized for the average heart size. For each heart, the LMDA branching pattern was drawn on scaled paper. Microdissection was often required for the measurements and the drawing in detail, because the smaller branches were deeply submerged within the myocardium and could not be traced at the surface. After the LMDA branching patterns were drawn, the soft tissue of hearts was dissolved in concentrated hydrochloric acid for 48 h, and the actual branching pattern of coronary arteries in latex was compared with its schematic drawing (Fig. 1, B and C).
|
Coronary artery ligation. Coronary artery ligation was performed in 58 mice. Under pentobarbital sodium anesthesia (43 mg/kg iv into the tail vein via a 30-gauge needle), a 24-gauge catheter (Abbocath) was inserted into the trachea through the oral cavity and connected to a rodent ventilator (model 683, Harvard). Artificial ventilation (room air, 120 strokes/min, 20 ml·kg–1·stroke–1) was provided throughout the procedure, which was conducted under aseptic conditions. The thoracic cavity was opened through the left fourth intercostal space, and the heart was exposed in its lateral oblique view. Identification of the LMDA was assisted by a controlling illuminator. To induce a small MI, a 7-0 polypropylene ligature was placed 1 mm to the left from the origin of putative L2 or L3 branches of the LMDA. Because LV branches could not be observed in the beating heart, even under microscopic amplification, the ligature was placed at the distance of highest probability from the origin of the LMDA as established by analysis of results of casting (see below), i.e., 4 and 6 mm distal from the origin of the LMDA for L2 and L3, respectively. The origin of the LMDA can be detected after the auricle of the left atrium is elevated using fine, smooth-tipped forceps. The ligature was placed and provisionally tied for observation of the change in color of an appropriate area of the myocardium. If attempts to ligate one of the branches failed, placement of the ligature was attempted at the other branch. Attempts were repeated twice before the failure of proper ligation was conceded. To induce a large MI, the ligature was placed and tied around the LMDA at the level of the tip of the auricle, similar to previous reports (6, 8, 9, 15). Proper placement of the ligature was confirmed by observing the change in color on a large area of LV surface. In sham-operated animals, a loosely tied ligature was placed in a similar location. The thoracic wall was closed by layered stitches. The animals were removed from artificial ventilation and placed in a warm cage until recovery.
Hemodynamic measurements. To determine the relation of MI size to LV hemodynamic function, the latter was assessed 5–8 wk after surgery using a conductance catheter and pressure-volume (P-V) loop analysis, as described elsewhere (14). Briefly, under pentobarbital sodium anesthesia (43 mg/kg ip) and artificial ventilation, the chest was opened and a Millar 1.4-Fr pressure-conductance catheter was inserted into the LV through the apex. P-V loops were recorded in real time (ARIA, Millar Instruments). The catheter was precalibrated in heparinized mouse blood, and the signal offset was calibrated in situ after data collection using a 10-µl bolus of 15% hypertonic saline injected into the jugular vein, as recommended by the manufacturer and described elsewhere (14). Variations in venous return were achieved by transient occlusions of the caudal vena cava, accessed through a central subxyphoid incision just below the level of the diaphragm. Analysis algorithms have been described in detail previously (14). Three indexes of LV performance are reported: ejection fraction (EF), the most clinically relevant index; maximum dP/dt (dP/dtmax), another frequently reported, but load-dependent, index of LV systolic function; and preload-recruitable stroke work (PRSW), an index of LV systolic function, regarded as load independent and available exclusively via P-V loop analysis (2). EF and dP/dtmax were measured from steady-state P-V loops; PRSW was estimated from the series of loops produced by preload variation.
Estimation of MI size. After completion of hemodynamic measurements, mice were euthanized by an injection of 0.5 ml of 1 M potassium chloride and heparin (20 U/ml) in PBS (pH 7.4) via the right jugular vein. The harvested hearts were fixed in 10% neutral buffered formalin solution and embedded in paraffin blocks. Sections (6 µm) of the heart were made at regular intervals of 600 µm (a total of 7–12 sections) and stained with hematoxylin and eosin. The MI size for each section was calculated as a ratio of the infarcted segment to the total LV perimeter averaged between outside and inside measurements using the Image Lab program, and the average MI of all sections was expressed as a percentage of total LV perimeter.
Statistical evaluation. Values are means ± SD. The infarct size resulting from ligatures placed at different locations and hemodynamic indexes were compared using ANOVA with post hoc comparisons.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
LMDA branching patterns. The average diameter of the hearts in surgical view after fixation was 5.8 ± 0.4 mm. The average length of the LMDA from its origin to the apex was 7.7 ± 0.5 mm. The LMDA runs obliquely and does not follow the paraconal interventricular groove (Fig. 1A). In 17% of mice (9 of 53), the LMDA bifurcates 2 mm proximally from the apex. The LV branches emanate from the LMDA and run obliquely but parallel to each other into atrial and marginal surfaces.
We classified the pattern of the left branching of the LMDA into three main types on the basis of the distribution of the branches. In type I (22 mice, 41.5%; Fig. 2A), L1, L2, and L3 originate independently from the LMDA and distribute separately on the surface of the LV. In type II (19 mice, 35.8%; Fig. 2B), L1 and L2 originate as a single common branch (L1&2), which eventually bifurcates into two subbranches, distributing in the upper and middle areas of the LV. L3 originates separately from the LMDA. In type III (8 mice, 15.1%; Fig. 2C), L1 is separate, while L2 and L3 originate from the LMDA as one common branch (L2&3), which eventually bifurcates, distributing in the middle and lower areas of the LV. Several mice (7.6%) did not fit into the above classification: two mice had two left common branches approximately corresponding to L1 and L3, but each one of them bifurcated; two other mice had a single LV branch that eventually divided into three branches.
Likelihood of LV branch point location within specific branching patterns. In type I (Fig. 2A), L1, L2, and L3 were each represented by a single artery in 12 of the 22 mice; in the other mice one of these three branches was represented by two parallel arteries, while the other two were single. The frequency distribution of the distances from the origin of the LMDA to the points at which L1, L2, and L3 originated from the LMDA is presented in Fig. 2A (right). For L1 the most frequently encountered distance (statistical mode) was 2 mm, for L2 it was 4 mm, and for L3 it was 6 mm.
In type II (Fig. 2B), the common L1&2 branch was always represented by a single artery. The origination of the common L1&2 branch from the LMDA most frequently occurred 2 mm below its own origin. This common branch bifurcated into two and distributed to the upper and middle areas of the LV. The most common distance of the origin of L3 from the origin of the LMDA was 5–6 mm. L3 also was mostly represented by a single artery, but in 3 of 19 mice, this branch was duplicated.
In type III (Fig. 2C), the single L1 most often originated from the LMDA 2 mm distally from the origin of the LMDA (statistical mode) and most often occurred as a single artery (7 of 8 mice). The common (L2&3) branch was always represented by a single artery and originated from the LMDA at 4 mm from its origin.
Right ventricular branches of the LMDA. We also observed one (25 mice, 47.2%) or two (20 mice, 37.7%) or more than three (8 mice, 15.1%) right ventricular branches of the LMDA. Their origin from the LMDA and their sizes were more variable than those of the LV branches. In 15 of 25 mice (28%), a single large right ventricular branch originates from the LMDA at such a high level that it resembles the second LMDA.
Induction of MI. On the basis of analyses of branching of the main descending artery, we can conclude that left branches appear to distribute more uniformly than right branches and, thus, represent more reliable targets for ligation. To induce a small MI located in the middle of the left anterior wall, L2 must be targeted for ligation, and the most probable place for attempted ligation of L2, on the basis of our findings in Fig. 2, is to the left of the LMDA, 4 mm distal to its origin. Attempts to ligate L3 by placing a ligature 6 mm distal to the origin of the LMDA should induce a small MI in the lower part of the left anterior wall. (L1 is very seldom accessible, because it is covered by the left atrium.)
Table 1 represents the results of coronary ligation of the LMDA and different left coronary branches with the intent to produce different-sized MIs. Small MIs were successfully induced in 57% of 28 mice, in which ligation of L2 and L3 were attempted. There was no mortality among these animals or among six mice subjected to the sham operation.
|
To contrast the functional deficit of small vs. large MIs, 30 mice were subjected to ligation of the LMDA at the level of the left atrium. All ligations were successful. There was no perioperative mortality; however, 20% of the animals died 2–6 days after surgery as a result of cardiac rupture, cardiac arrest, or cannibalism.
Histological examination 5–8 wk later showed that all MIs were transmural. The MI size after attempted L2 ligation (Fig. 3C) was 4.7 ± 1.5% of the LV, with scar tissue located in the middle third of the LV at the lateral-to-posterolateral free wall. The MI size after attempted ligation of L3 or the L2&3 common branch (Fig. 3B) was 16 ± 3.64% of the LV in the lower third of the LV near the apex at the lateral part of free wall. Ligation of the LMDA (Fig. 3A) resulted in large MIs, 44.5 ± 6.29% of the LV, spreading through the anterior-to-posterior wall and always including the apex.
|
Therefore, as a result of our "trial" to induce the MIs of a predetermined size and location, we produced three groups of relatively uniform transmural infarcts: small (lower third of the anterolateral-to-posterior wall near the apex), very small (middle third of the anterolateral-to-posterior wall), and large (most of the anterior wall, including the apex).
Hemodynamic measurements. Figure 4 illustrates the hemodynamic and LV functional changes 5–8 wk after ligation of the different branches of the LMDA or the LMDA itself. Small MIs (16 ± 3.64% of the LV) associated with ligation of L3 of the LMDA (or most probably with ligation of L2&3 common branches) resulted in a statistically significant decline of LV function: EF fell by 36%, dP/dtmax by 33%, and PRSW by 37%. LV function in mice with very small MIs (4.7 ± 1.5% of the LV), induced by ligation of L2, did not differ from that in sham-operated mice. In contrast, large MIs (44.5 ± 6.29% of the LV) associated with ligation of the LMDA resulted in a pronounced decline of LV function compared with sham-operated animals: EF fell >70%, dP/dtmax declined by 48%, and PRSW, the load-independent index of LV systolic function, was reduced by 57%.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Since the mouse model of experimental MI by coronary ligation was first introduced (16), nearly all subsequent research had concentrated on ligation of the left descending branch (LMDA in our study), with the intent to induce a large MI. This resulted in a vast heterogeneity of infarct sizes: 10–70% of the LV (1a, 4, 10, 11, 13). On the basis of our classification, we can infer that, in prior studies, when the ligature was properly placed at the main artery (LMDA) at the level of the left atrium (similar to our study), the large (on average 44.5% of the LV) MI was a predictable outcome in 100% of attempts. Some variability in size among MIs resulting from ligation of the LMDA in prior studies likely depended on the type of distal branching. For instance, the outcome of a large MI (>35% of the LV) indicates a type I (3 independently originating LV branches) or type III (L2&3 originate as a common branch) pattern. Creation of a moderately sized MI (30–35% of the LV) after ligation of the LMDA in prior studies suggests a type II branching pattern, in which L1&2 arises as a common branch from the LMDA above the level of ligation, i.e., a smaller portion of the LV was perfused by the ligated branch. Accordingly, the probability of a moderately sized MI (30–35% of the LV) after ligation of the LMDA at the level of left atrium might be expected to occur in 36% of cases.
A smaller MI (<30% of the LV) after ligation of the LMDA reported in some previous studies (1a, 7, 10, 11) suggests that the ligature was placed much more distally than intended or that ligation was incomplete, and the MI resulted probably from a partial interruption of the flow by scar tissue induced by the ligature. In the latter case, the MI is usually nontransmural. Animals with a smaller MI, unintentionally induced by ligation of the LMDA, have usually been removed from published study data analyses (7, 11), or their results have been presented separately (1a, 10), because those hearts did not show significant remodeling or functional decline. For instance, Bayat et al. (1a) reported that morphometric and functional indexes derived from echocardiography of mice with MI <30% of the LV (16 ± 3% of the LV) were not different from those of sham-operated animals 6, 12, or 18 wk after coronary ligation. In distinct contrast, in our study, small transmural MIs (16 ± 3.6% of the LV), purposely induced by ligation of L3 or the L2&3 common branch, showed statistically significant reduction of EF, dP/dtmax, and PRSW compared with sham-operated mice 5–8 wk after ligation. These results support our assertion that so-called "small MIs" resulting from ligation of the LMDA of the coronary artery tree, which are not accompanied by a significant remodeling and functional decline, were not clearly delineated transmural infarcts but, rather, a result of nontransmural myocardial damage related to a partial or temporary blockage of the artery.
In summary, our study of the branching patterns of the LMDA in C57BL/6 mice delineated a branch pattern probability-assisted technique for ligation of the left main descending artery itself and its small LV branches that permits reliable induction of small MIs of predictable location and uniform size. In vivo hemodynamic assessment of LV function 5–8 wk after induction of MI showed functional decline proportional to the infarct size. Application of the present method to reproducibly create infarcts of a uniform size and specific location will enhance the precision of experimental studies aiming to test novel pharmacological or gene therapy for ischemic heart disease. The ability to create small transmural infarcts will be particularly useful in long-term studies investigating interaction of aging and ischemic heart disease with hypertension or other experimental genetic pathology.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Lygate Letter to the editor: Infarct size measurements are critically important when comparing interventions affecting ventricular remodeling Am J Physiol Heart Circ Physiol, November 1, 2007; 293(5): H3221 - H3221. [Full Text] [PDF]
|
|
|
|
 |
|
 J. Takagawa, Y. Zhang, M. L. Wong, R. E. Sievers, N. K. Kapasi, Y. Wang, Y. Yeghiazarians, R. J. Lee, W. Grossman, and M. L. Springer Myocardial infarct size measurement in the mouse chronic infarction model: comparison of area- and length-based approaches J Appl Physiol, June 1, 2007; 102(6): 2104 - 2111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Klocke, W. Tian, M. T. Kuhlmann, and S. Nikol Surgical animal models of heart failure related to coronary heart disease Cardiovasc Res, April 1, 2007; 74(1): 29 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hazarika, M. R. Van Scott, and R. M. Lust Severity of myocardial injury following ischemia-reperfusion is increased in a mouse model of allergic asthma Am J Physiol Heart Circ Physiol, January 1, 2007; 292(1): H572 - H579. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Cozzi, S. Hazarika, H. W. Stallings III, W. E. Cascio, R. B. Devlin, R. M. Lust, C. J. Wingard, and M. R. Van Scott Ultrafine particulate matter exposure augments ischemia-reperfusion injury in mice Am J Physiol Heart Circ Physiol, August 1, 2006; 291(2): H894 - H903. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Cai, J. M. Holm, I. J. Duignan, J. Zheng, M. Xaymardan, A. Chin, V. L. T. Ballard, J. N. Bella, and J. M. Edelberg BDNF-mediated enhancement of inflammation and injury in the aging heart Physiol Genomics, February 23, 2006; 24(3): 191 - 197. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
