Transthoracic Echocardiography in Models of Cardiac Disease in the Mouse
Background Transthoracic echocardiography (M-mode and Doppler) offers a noninvasive approach for in vivo evaluation of the mouse heart. The present study examines its usefulness for assessing the morphological/functional phenotype of the left ventricle (LV) in several transgenic and surgical murine models of cardiac disease.
Methods and Results Observations were made in 83 intact, anesthetized mice. In mice with a surgical arteriovenous fistula, volume overload and LV dilation were detected. In normal mice, echocardiographic indexes of increased contractility (dobutamine) were confirmed by LV dP/dtmax. In transgenic mice with overexpression of the β2-adrenergic receptor, heart rate and mean velocity of circumferential fiber shortening were increased, indicating enhanced contractility. In colony screening of transgenic mice overexpressing the H-ras gene, 45% had increased LV wall thickness (>0.9 mm), and those showing a striking increase were selected for breeding. In mice with LV hypertrophy (aortic constriction) and normal mice, the actual LV mass determined by echocardiography correlated well (r=.93), and 95% confidence limits were determined. The maximum intraobserver and interobserver coefficients of variation for M-mode data were 0.03±0.29 mm (±2 SD), <10% for LV internal dimensions but 27% to 30% for wall thickness.
Conclusions These studies provide the first application of transthoracic echocardiography for morphological/functional characterization of the cardiac phenotype in transgenic and surgical murine models, including (1) high reliability for detecting LV chamber dilation and function; (2) reliability (and its limits) for determining abnormal LV wall thickness and LV mass; (3) identification of marked, sometimes asymmetrical, hypertrophy in a transgenic model of hypertrophic cardiomyopathy; and (4) usefulness for transgenic colony screening to identify markedly abnormal phenotypes.
The mouse has become a primary target for genetic manipulation because of extensive knowledge of the mouse genome and the availability of mouse embryonic stem cell technology. Transgenic and gene targeting methods have now been successfully applied to develop murine models of altered cardiac size and function1 2 3 4 5 and, together with the development of microsurgical techniques for inducing cardiac overload in the mouse,6 7 8 have led to the need for approaches to reliably assess in vivo the functional and anatomic cardiac phenotype. The small size of the mouse heart has limited the use of traditional approaches, but hemodynamic studies of the LV have been performed in a miniaturized Langendorff preparation of the isolated mouse heart,9 aortic pressure has been measured in intact anesthetized mice,6 and high-fidelity LV pressure and LV dP/dtmax have been measured in situ in normal and transgenic mice under anesthetized open-chest conditions.2 4 Also, to assess both function and morphology of the mouse heart, we applied a miniaturized contrast angiographic technique to study ventricular size and function.8 However, the use of angiography requires surgical anesthesia with tracheal intubation and cannulation of the vena cava for injections of x-ray contrast medium,8 and therefore we sought a more physiological, noninvasive approach.
The use of echocardiography for estimating LV mass in mice has been described,10 11 and Hoit et al12 applied transthoracic echocardiography to define differences in LV function between normal mice and transgenic mice having enhanced myocardial contractility. Pollick et al13 also reported averaged echocardiographic data in relatively large normal mice (BW, 48 g; range, 38 to 65 g), and the present work has been described in part in preliminary form.14 The ultimate usefulness of this approach will be dependent on its ability to assess quantitative differences between morphological and functional cardiac phenotypes in vivo. Accordingly, the present study was designed to further examine the reliability of transthoracic echocardiography in mice and, in particular, to assess its utility for detecting abnormalities of LV chamber size, LV WTh, and LV mass and function in transgenic and surgical murine models of cardiac disease and its application for transgenic colony screening.
Animals were handled according to the animal welfare regulations of the University of California, San Diego, which are in accordance with American Heart Association guidelines, and the experimental protocol was approved by the Animal Subjects Committee of this institution.
Eighty-three mice were studied, including 10 C57BL/6 mice (average BW, 27.2±5.4 g [mean±SD]) in which an AVF between the infrarenal descending aorta and the inferior vena cava was produced surgically 1 week before study; these mice were compared with 12 weight-matched C57BL/6 controls (BW, 27.0±4.5 g); 6 normal mice studied before and after dobutamine administration; 15 TGβ targeted with the α-myosin heavy chain promoter (BW, 36.0±8.0 g),2 with 14 weight-matched wild-type controls (BW, 35.5±7.8 g); 12 mice with transverse aortic constriction6 7 maintained for 1 week to produce concentric LV hypertrophy, which were used together with 15 normal mice from the control and dobutamine groups for the calculation of LV mass; and 8 TG-ras targeted to the ventricles with the myosin light chain-2v promoter4 and 6 controls (5 wild-type and 1 heterozygote).
In most studies, mice were anesthetized with a mixture of ketamine (100 mg/kg IP) and xylazine (5 mg/kg IP). The anterior chest was shaved, and small-needle ECG leads were placed through the skin on the right and left upper extremities and the left leg. In 12 mice, anesthesia with chloral hydrate (450 mg/kg IP) was used, which had an onset of sedation within 20 to 25 minutes and was maintained for about 30 minutes.
The echocardiograph system used (Apogee CX, ATL Interspec) employs dynamically focused symmetrical annular array technology15 16 for 2D and M-mode imaging. It also has color flow mapping and spectral Doppler capabilities. The 7.5-MHz transducer has a wide bandwidth and can be shifted to carrier frequencies of 9.0, 7.5, or 5.0 MHz. To optimize resolution and penetration with 2D and M-mode imaging, the 9.0-MHz frequency was used at a minimum depth setting of 3 cm and sector width of 40° to 50°. The 2D images were obtained in white-on-black format with M-mode and spectral Doppler tracings in black-on-white format. For pulsed-wave Doppler recordings, the minimum sample size (0.6 mm) was used and the pulse repetition frequency was 5 kHz.
In later studies, a specially constructed standoff for the ultrasonic transducer was used that consisted of a short cylindrical polyethylene tube (1.1-mm WTh, 2.54 cm in diameter, 1.59 cm high) attached to the convex head of the transducer. The standoff was filled with ultrasound transmission gel so that in use the distance between the top of the transducer and the chest wall was ≈0.66 cm. The standoff made it possible to substantially improve near-field image quality, so that the IVS could be particularly well visualized in small mice (17 to 21 g).
With the mouse in the left lateral decubitus position, the transducer was placed on the left hemithorax. Care was taken to avoid excessive pressure, which can induce bradycardia. Using the 2D parasternal short-axis imaging plane as a guide, an LV M-mode tracing was obtained close to the papillary muscle level with a sweep speed of 50 or 100 mm/s. Pulsed Doppler tracings of the estimated peak LV outflow tract velocity and the mitral inflow velocities were obtained in a modified parasternal long-axis view at a sweep speed of 100 mm/s. M-mode tracings of the LV were recorded on videotape and a Sony digital color printer (UP-520MD or UP-1800MD). Doppler tracings were recorded on videotape and a black-and-white thermal printer (Sony UP-870MD). Echocardiographic data were obtained by two examiners (N.D. and N.T.).
In four of the six normal mice studied, within 10 days before the echocardiographic examination using dobutamine (1.0 to 1.5 μg/g BW IP) a significant effect on myocardial contractility was confirmed by measuring the LV pressure and estimating changes in LV dP/dtmax by use of a stretched PE-50 catheter attached to a Statham fluid-filled transducer (P-50) inserted via the right carotid artery into the LV. The appropriate timing for the echocardiographic study using the same dose of dobutamine also was determined.
M-mode tracings were digitized with a scanner (ScanJet IIcx, Hewlett Packard) with an application program (Desk Scan II) and visualized in 8×10-in format to facilitate accurate measurements. The images were measured by two individuals using NIH Image (version 1.52) on a 15-in computer screen by the American Society of Echocardiography leading-edge technique.17 Heart rate was determined from at least three consecutive RR intervals on the LV M-mode tracings and also from spectral Doppler tracings of the LV outflow tract and mitral inflow region. LVDed and LVDes, LV PWTh, and IVSWTh were determined, end diastole being defined as the maximal LV diastolic dimension and end systole as the most anterior systolic excursion of the LV posterior wall. All LV dimension data are presented as the average of measurements by two observers of two or three selected beats. The FS of the LV is expressed as a percentage as %FS=(LVDed−LVDes)/LVDed×100.
LV mass calculation
In normal mice and transverse aortic constriction mice with concentric LV hypertrophy, after echocardiography the experiment was terminated with an overdose of intraperitoneal sodium pentobarbital, the heart was excised, the atria and vessels were dissected free, and the LV including the IVS was blotted and weighed on an analytical balance (Sauter model RE1614).
Estimated echocardiographic LV mass was calculated (in mg) as LV mass (echo)=[(LVDed+IVSWTh+PWTh)3−LVDed3]×1.055, where 1.055 (mg/mm3) is the density of myocardium.11
Data from Doppler studies were also scanned and measured on the 15-in computer monitor. Vp and the peaks of the E and A waves of the mitral inflow velocity tracing were recorded as the average of three beats. Using the aortic ET obtained from the Doppler tracing of the LV outflow tract, we also calculated the mean Vcf as mean Vcf=(πLVDed−πLVDes)/(ET×πLVDed)=[(LVDed−LVDes)/LVDed]/ET.
Although it is sensitive to acute changes in loading conditions, under basal conditions in the absence of acute changes in arterial pressure, the mean Vcf provides an approach for assessing myocardial contractility.18
To determine intraobserver variability of M-mode measurements, one observer remeasured 23 sets of LVDed, LVDes, PWTh, and IVSWTh on a second occasion in 12 animals.
Dimension data from the LV M-mode tracings were determined by two individuals using the same two or three beats. Interobserver variability was tested with 23 sets of LVDed, LVDes, PWTh, and IVSWTh measurements in 12 animals.
The same 12 mice also were studied under ketamine and xylazine anesthesia on different days (4 mice 5 days, 1 mouse 4 days, and 6 mice 9 days after the first studies). One mouse died just after ketamine and xylazine anesthesia before the second examination. Thus, 22 pairs of LVDed, LVDes, PWTh, and IVSWTh measurements were analyzed from M-mode tracings in 11 animals. In addition, in 5 TGβ and 5 weight-matched controls, we studied reproducibility at 3 or 4 days using chloral hydrate sedation on both occasions.
All data are shown as mean±SD unless otherwise noted. The significance of group differences was determined with unpaired t tests with a significance level of P<.05. The effect of dobutamine was evaluated with a paired t test with a significance level of P<.05. Agreement between two measurements in the validation studies was determined according to the method of Bland and Altman,19 coefficients of variability and repeatability being calculated as the mean±2 SD of the differences between two measurements. The correlation coefficient and 95% confidence limits were determined from the linear correlation between directly measured and echocardiographically determined LV mass.
The BWs of the mice studied ranged from 17.0 to 51.5 g. In our experience, the initial success rate for obtaining a satisfactory echocardiographic study was very high (95%), and studies are generally completed within 20 minutes. In later phases of this study, a standoff was attached to the transducer (see “Methods''), which markedly improved the near-field visualization and resulted in equally high success rates (even in mice as small as 17 g). The normal right ventricular chamber was sometimes seen, particularly when the standoff was used. The ascending aorta could be readily identified, and the mitral valve leaflets were frequently seen.
LV Dimensions and Function in the Normal Mouse Heart
Representative 2D and M-mode echocardiograms in a normal mouse are shown in Fig 1⇓. Normal mice weight-matched to AVF mice (control group 1) and those weight-matched to TGβ (control group 2) differed significantly from each other in BW because of the age difference, and there was a nonsignificant trend for cardiac dimensions to be larger in control group 2. There was no difference in %FS between the two control groups (Table 1⇓).
Normalizing cardiac dimension data by BW did not prove satisfactory, since BW differences were proportionately much larger than cardiac size differences. Therefore, different control groups with matching BWs were needed to test for abnormalities in the AVF and TGβ. This matter is further illustrated by the much younger mice in Table 2⇓ (average age, 7 weeks); in these animals, BW, LVDed, IVSWTh, and PWTh are all significantly smaller (P<.05) than control group 2 mice (Table 1⇑).
The estimated Vp (Fig 2⇓) was 0.65±0.10 m/s in control mice (control group 1) and 0.70±0.13 m/s in control group 2 (P=NS).
In control group 1, early mitral inflow velocities (E waves) averaged 0.46 m/s and velocities during atrial systole (A waves) were lower (0.23 m/s) (Table 1⇑, Fig 2⇑). In control group 2, values were similar and not significantly different.
Effects of Dobutamine
The effects of the intraperitoneal dobutamine lasted more than 20 minutes during LV pressure and dP/dt monitoring. Mean LV pressure increased from 89±10 to 115±22 mm Hg, and LV dP/dtmax increased from 5746±1239 to 11 842±5213 mm Hg/s at an average of 13 minutes after intraperitoneal dobutamine injection.
On the basis of the above findings, echocardiographic recordings when dobutamine was given on a separate occasion were performed in the same animals at an average of 13 minutes after dobutamine administration. Fig 3⇓ shows representative M-mode tracings before and after intraperitoneal dobutamine administration. Dobutamine significantly increased heart rate and shortened the LV ET; the LVDed did not change, but LVDes became significantly smaller and %FS increased. The mean Vcf was greatly increased, from 5.67 to 9.00 circumferences/s (Table 2⇑).
LV Dimensions and Function in the Abnormal Mouse Heart
During the initial 3 to 5 minutes of each echocardiographic study, color flow mapping was performed in the abdomen, and in 9 of 10 mice found to have a patent AVF at postmortem study (≈1 mm in diameter), a shunt flow signal was detected. Also, a continuous flow signal was identified in the abdominal aorta at the fistula site by pulsed Doppler echocardiography. In mice with a significant AVF, the mitral inflow velocity signal was disturbed by signals from increased pulmonary venous flow in some animals, so that transmitral data could not be reliably evaluated in all studies (Table 1⇑).
LV dimensions and %FS
The mean heart rates were not significantly different (Table 1⇑). Dilation of the LV was readily detected, the LVDed being significantly larger than in controls without a difference in LVDes, and the LV %FS was significantly higher in the AVF compared with control group 1. The LV ET was prolonged in AVF animals, and there was no significant difference in mean Vcf between AVF and controls. The LV PWTh and IVSWTh were mildly increased (Table 1⇑). Together, these findings are consistent with eccentric LV hypertrophy.
In AVF mice, Vp was markedly increased compared with controls (Table 1⇑). The E-wave velocities were significantly higher in AVF mice than in controls, but the difference in A waves was not significant (probably because of the small number of satisfactory tracings in the AVF group) (Table 1⇑). These findings indicate substantial volume overload on the LV.
The heart rate was significantly higher in TGβ than in control group 2 mice (P<.01) (Table 1⇑). There were no significant differences in LV dimensions, %FS, and WTh between TGβ and control group 2 mice, and Vp also was not different. In TGβ, the ET was shortened, and the mean Vcf was 6.20±1.52 circ/s, significantly higher than in control group 2 (Table 1⇑).
Doppler studies showed no difference in Vp and mitral flow velocities between TGβ and controls (Table 1⇑).
When a subset of TGβ (n=5) and control group 2 (n=5) mice was restudied under chloral hydrate sedation, heart rates were significantly higher in both controls and TGβ (474±98 and 525±73 bpm, respectively) and no longer significantly different. Also, no significant differences could be detected between controls and TGβ in cardiac dimensions and %FS. The mean Vcf was increased in both groups, but there was no significant difference between controls and TGβ (8.78±3.02 versus 8.65±2.55 circ/s, respectively, P=NS).
LV WTh and Mass by Echocardiography
The linear correlation and 95% confidence limits for the relation between LV mass estimated by echocardiography and that directly measured at postmortem study are shown in Fig 4⇓. There is an excellent correlation (r=.93), with some overestimation of LV mass by echocardiography. The values >150 mg were obtained in mice with concentric LV hypertrophy. Those values lying outside the 95% confidence limits are consistent with the coefficients of variation for LV WTh (see “Reproducibility,'' below) and constitute a significant percentage error, which is related to the limits of resolution of the instrument, as discussed subsequently.
Echocardiography in TG-ras and Transgenic Colony Screening
Among the 49 homozygous TG-ras mice screened for the presence of hypertrophy, 22 (45%) were found on echocardiography to have either increased IVSWTh or PWTh (>0.9 mm). Among these 22 mice, 8 were selected to encompass a range of increases in estimated WTh.
In the selected group of H-ras homozygous mice (n=8) and controls (n=6), both WThs were then measured by echocardiography by two observers as described above. Examples of echocardiograms in homozygous mice exhibiting moderately and markedly thickened LV walls are shown in Fig 5⇓.
PWTh averaged 1.43±0.55 mm in homozygous mice compared with 0.73±0.12 mm in controls (Fig 6⇓). The IVSWTh averaged 1.05±0.50 versus 0.71±0.12 mm (P=NS), respectively, in the same two groups. The average of the PWTh and IVSWTh was 1.24±0.50 versus 0.73±0.12 mm (P<.05). This figure provides an example of how echocardiography can be used in colony screening to select mice with a marked phenotype for further breeding.
Vp by Doppler was elevated in two of the transgenic mice, but the mean values averaged 0.90±0.28 (n=8) in H-ras mice and 0.81±0.17 m/s in controls and were not significantly different statistically. In the homozygous H-ras mice, %FS also was normal, 41.5±9.1%. The number of studies performed at the mitral valve was insufficient for analysis.
The agreement was generally good between two measurements by the same observer for all M-mode data (Fig 7A⇓), with mean differences of 0.03 mm and coefficients of variation of 0.22 to 0.24 mm (representing ±2 SD of the differences) (Table 3⇓). The maximum difference expressed as a percentage of LV chamber diameter was 8%. The absolute differences for WTh measurements were small (mean difference, 0.03 mm) between measurements and the coefficient of variation good (Table 3⇓), but when expressed as a percentage of the normal WTh (average, 0.72 mm), the maximum difference was 30% (Table 3⇓).
The agreement between measurements by two individuals for M-mode dimension data was good and similar to the intraobserver data (Fig 7B⇑). Absolute differences and coefficients were small, and percentage differences in LV chamber diameter were <10%. Similarly, for WTh the average absolute difference was very small, but the maximum difference expressed as a percentage of WTh was 27% (Table 3⇑).
Repeatability on different days with ketamine and xylazine anesthesia showed an average percent difference of 5.2% to 5.6% but with a large SD because of marked heart rate variability between the two studies. When chloral hydrate was used for sedation (Fig 7C⇑), repeatability was better than with ketamine and xylazine anesthesia (Table 3⇑). Chloral hydrate significantly increased the heart rates in both TGβ and control mice and reduced the SD of repeatability by 40%.
This study identifies an important role for noninvasive echocardiographic studies in assessing changes in LV morphology and function in transgenic and surgical murine models of cardiac disease. As reported previously by others,12 satisfactory intraobserver and interobserver variabilities were achieved in this study, and the success rate of echocardiography was high (>95%) in mice weighing as little as 17 g when a standoff was used with the transducer.
The annular array technology used in this study and in previous studies12 13 has certain favorable features for such studies in small animals.15 16 The transducer consists of several concentric ring-shaped elements driven mechanically, resulting in a cylindrical ultrasound beam that can be electronically focused in two planes; such focusing results in improved penetration and resolution because of its ability to produce a narrow beam width and thin tomographic slice, which, in turn, permit use of higher-frequency transducers.15 Thus, the 9-MHz mode of the transducer used in this study allowed satisfactory images to be obtained in almost all mice.
Assessment of LV Chamber Dimensions and WTh
LV chamber diameters
Two SD of the mean difference in measuring LV chamber diameter was 0.24 mm for intraobserver and 0.29 mm for interobserver variability (<10% error). Therefore, chamber enlargement could be detected with 95% confidence in a mouse population having LV chamber enlargement by use of these coefficients of variability, and they also could be used to detect an increase in chamber diameter >0.29 mm in an individual animal. In AVF mice, the LV chamber was dilated, the average LVDed being increased 0.6 mm above controls, and the difference was statistically significant. Thus, M-mode echocardiography will be very reliable in identifying even modest chamber enlargement in mouse models of heart failure as well as impaired systolic diameter shortening.
LV WTh and LV mass
Two SD of the mean difference for measuring WTh was 0.22 mm for intraobserver and 0.20 mm for interobserver variability, indicating that a 0.22-mm thickness increment could be used to identify a mouse population having increased LV WTh with 95% confidence, depending on group size, by these strict criteria. Echocardiography could also be used to detect LV wall thickening in an individual mouse, keeping in mind that 0.20 to 0.22 mm represents 28% to 30% of the normal WTh (0.7 mm), and an LV WTh ≥0.9 mm was used in initial colony screening of TG-ras. In AVF mice, the difference in WTh from control group 1 was 0.17 mm for PWTh, and in the sizable group of AVF mice, this relatively mild increase was statistically significant.
We obtained an excellent correlation between echocardiographically estimated and actual LV mass in normal and pressure-overloaded hearts, with some overestimation of LV mass by echocardiography and, as expected, some points falling outside the 95% confidence limits. These findings appear to be consistent with the range of the percentage coefficients of variation for WTh of 27% to 30% (±2 SD) (Table 3⇑). Manning et al10 also overestimated LV mass by echocardiography, with a good correlation in 15 normal mice (young and senescent) with scatter (SEE, 18 mg), whereas Gardin et al11 showed significant correlations but markedly underestimated LV mass by echocardiography in normal and hypertrophied hearts (scatter was not analyzed statistically). For calculating the LV mass, an assumption of cubic LV geometry has been made because the LV long axis cannot be measured accurately in the mouse, and errors obviously may result from cubing WTh values. Nevertheless, our echocardiographic methods, using the averaged data from two observers and taking into account the strict coefficients of reproducibility (±2 SD), appear to provide useful estimations of LV mass in groups of animals. The reliability of measurements of LV WTh and calculating LV mass from echocardiographic data in the mouse is shown to be limited primarily by the resolution of current echocardiographic instrumentation, as discussed further under “Study Limitations,'' below.
Assessment of LV Function
Invasive methods have been useful as a terminal procedure to assess LV function in the mouse,2 4 8 9 but the noninvasive technique of echocardiography provides a more physiological approach, because the chest is not opened and the animal is not intubated. Moreover, valuable animals such as transgenic mice or surgically produced models of overload can be preserved and followed over time by serial echocardiographic studies.
In the AVF mice, a patent shunt could be detected, and LV chamber enlargement with increased WTh was identified. Moreover, the LV %FS and peak aortic velocity were increased, with increased ET and normal mean Vcf. These findings indicate the presence of chronic volume overload on the LV with eccentric LV hypertrophy, enhanced LV function, and normal contractility (assessed by mean Vcf), as observed in large-animal models of compensated volume overload.20
In TGβ, LV dP/dtmax has been reported to be significantly higher than in control mice under open-chest conditions.2 The average heart rates were slower in these transgenic animals studied by echocardiography under ketamine-xylazine anesthesia (331 bpm) compared with the open-chest state (500 bpm).2 We were not able to clearly identify increased contractility by echocardiography related to β2-adrenergic effects on the LV myocardium, since the observed increase in mean Vcf could have reflected an effect of the increased heart rate in TGβ compared with that in controls (331 versus 264 bpm), although no information is yet available on force-frequency effects in the intact mouse. The higher mean Vcf in TGβ compared with control mice was absent at the higher heart rates associated with chloral hydrate sedation, further suggesting that the difference observed with ketamine-xylazine anesthesia may have been largely a result of the heart rate difference. With chloral hydrate, the TGβ as well as controls may have reached the maximum level of positive inotropic state achievable by β-adrenergic stimulation, as in open-chest mice when TGβ and controls were treated with the highest dose of isoproterenol.2 Conversely, it is possible that under closed-chest basal conditions with ketamine-xylazine anesthesia, neurohumoral feedback mechanisms were operative and precluded detection of clearly enhanced myocardial contractility. Finally, it is possible that the echocardiographic approach is less sensitive than measurement of LV dP/dtmax, since the latter has been shown to be useful for detecting very small changes in myocardial contractility.21
Changes in LV contractility induced by dobutamine administration were readily identified by echocardiography in our study (Table 2⇑), with changes in Vcf and LV %FS comparable in magnitude to the increase in LV dP/dtmax observed at the earlier hemodynamic study in the same animals. The mean Vcf and %FS are known to be affected by both contractility and loading conditions,18 but LVDed by echocardiography was unchanged after dobutamine, and these measures increased despite a moderate increase in LV systolic pressure. Heart rate also was significantly increased by dobutamine (267 to 327 bpm) and undoubtedly contributed to the contractility increase. The observed augmentation of mean Vcf produced by dobutamine was large and probably reflected the combined effect of direct β-adrenergic stimulation of the myocardium, together with amplification of the force-frequency effect by β-adrenergic stimulation.22
The study by Hoit et al12 did not examine differences in morphology between groups of animals, but evidence for enhanced contractility with increased heart rate was found in mice with phospholamban deficiency. Our findings on LV dimensions and velocities in normal mice of comparable BWs are similar to those described in that report. As might be expected, we observed a much lower heart rate response to dobutamine than to the isoproterenol used in that study10 ; the increase in LV FS was similar, whereas the increase in Vp with dobutamine was less marked.
Variability of Measurements
Hoit et al12 presented variability data on normal LV dimensions that generally resemble those reported here. We did not perform intraobserver and interobserver variability studies on Doppler data, but such studies from the article by Hoit et al12 show narrower confidence limits for aortic and mitral velocities than for the M-mode data. This finding might be expected, since peak velocities by Doppler can be determined more consistently than M-mode measurements of endocardial motion.
Effects of Anesthesia
With ketamine-xylazine anesthesia, heart rates were generally relatively slow (258 and 264 bpm in the two control groups) compared with rates of 450 to 500 bpm reported in conscious mice.23 The heart rates often tended to slow somewhat during the study, and repeatability of echocardiographic dimensions several days apart by the same observer was inadequate in mice anesthetized with ketamine and xylazine, largely because of the very wide variability in heart rates observed in individual animals on two separate occasions.
Studies under chloral hydrate yielded heart rates closer to those in conscious mice (464 to 474 bpm in normal mice), and the reproducibility of studies in the same mice on different days was somewhat improved. However, this agent does not appear suitable, since the LVDed and LVDes and ET were considerably lower and mean Vcf was greatly increased with chloral hydrate, suggesting that endogenous catecholamine levels were higher and probably associated with a nonbasal state.
Transgenic Colony Screening
Echocardiography was applied in screening a transgenic mouse colony to identify mice with a markedly abnormal hypertrophic cardiac phenotype within a population of homozygous, heterozygous, and wild-type mice. In TG-ras, very mild LV hypertrophy occurs in heterozygous animals, with variable hypertrophy among homozygous offspring.4 By echocardiographic screening for WTh, we were readily able to identify increased thickness as well as homozygous mice with extreme LV hypertrophy for further breeding.
To compare LV chamber size, WTh, and function in echocardiographic examinations, it would be desirable to match heart rates in comparing individuals or groups of mice. In practice, it may be difficult to match heart rates (particularly in models exhibiting increased or decreased heart rate) without cardiac pacing or use of bradycardic drugs, which have not yet been tested in the mouse under closed-chest conditions. We have not found a satisfactory method of mathematically correcting cardiac dimensions for heart rate differences, and the problem with normalization of cardiac dimensions for BW was mentioned previously.
Axial resolution depends on the frequency of the ultrasound and the velocity of ultrasound transmission. With the 9-MHz frequency and under the assumption that velocity is 1540 m/s, one wavelength corresponds to 0.171 mm; thus, the ideal axial resolution of the system used can be estimated at 0.18 to 0.2 mm.24 In a normal mouse, assuming an LV WTh of 0.70 mm, this limit of resolution could result in a ≥28% to 30% maximum error in measuring WTh, in agreement with the intraobserver and interobserver variability data reported in this study. For more precise assessment of WTh in the individual mouse and for clearly delineating small structures in the mouse heart, frequencies in the 10- to 15-MHz range will probably be needed, although depth of imaging would be limited. Good near-field resolution may be commercially useful for many clinical and experimental applications, and although the current spherical probe shape works well in the mouse, improved transducer materials, a smaller footprint, a built-in standoff, and better focusing would be advantageous. Finally, the availability of multiline acquisition would increase the frame rate, although accompanying improvements in display technologies also would be necessary.
Such improvements should also enhance the quality and resolution of 2D images, which are useful for guiding M-mode and Doppler studies but are currently of limited value for direct quantitative measurements.
Spectral Doppler tracings were obtained from the LV outflow tract. However, the absolute value of the peak velocity is uncertain, since the angle between the Doppler beam and the targeted blood stream was not zero in the parasternal view because of the limited echocardiographic window in the mouse and the size of the transducer footprint; thus, we were not able to image the mouse heart from the apex, and the probe is too large for use in the suprasternal window. Nevertheless, the variability of the values for Vp in normal mice was small, and the approach was sufficient for detecting a relative change, as observed in the AVF mice. Studies on mitral flow velocity may be more reliable, since a relatively small angle between the ultrasonic beam and mitral inflow can be achieved.
In the pulsed Doppler method, the sampling volume is relatively larger than that used in human studies, which can lead to contamination with other information; for example, signals from the right ventricular outflow tract and pulmonary artery lie very close to the aortic signal in the mouse, and the sampling volume position should be confirmed frequently by 2D imaging or color Doppler flow mapping.
In summary, assessment of LV chamber dimensions by transthoracic M-mode echocardiography allows highly reliable assessment of abnormal LV chamber enlargement and detection of increased WTh in a variety of surgical and transgenic mouse models of disease. When combined with Doppler echocardiography, the anatomic and functional cardiac phenotype can be defined noninvasively and can also be used for colony screening.
Selected Abbreviations and Acronyms
|IVSWTh||=||interventricular septal wall thickness|
|LV||=||left ventricular, left ventricle|
|LVDed||=||LV end-diastolic internal diameter|
|LVDes||=||LV end-systolic internal diameter|
|PWTh||=||posterior wall thickness|
|TGβ||=||transgenic mice with targeted cardiac overexpression of the β2-adrenergic receptor|
|TG-ras||=||transgenic mice homozygous for targeted ventricular overexpression of the oncogenic H-ras gene|
|Vcf||=||velocity of circumferential fiber shortening|
|Vp||=||peak flow velocity in LV outflow tract|
This study was supported in part by SCOR grant in Heart Failure HL-53733 and Program Project grant HL-46345 awarded by the National Heart, Lung, and Blood Institute. The authors express appreciation to Anthony N. DeMaria, MD, for his assistance and advice during the initial phase of this work. We are also grateful to Robert J. Lefkowitz, MD, PhD, for supplying breeder TGβ. The work of Farid Abdel-Wahhab and Cheryl Bugsch is gratefully acknowledged.
- Received January 10, 1996.
- Revision received February 19, 1996.
- Accepted March 4, 1996.
- Copyright © 1996 by American Heart Association
Jackson T, Allard MF, Sreenan CM, Doss LK, Bishop SP, Swain JL. The c-myc proto-oncogene regulates cardiac development in transgenic mice. Mol Cell Biol. 1990;10:3709-3716.
Milano CA, Allen LF, Rockman HA, Dolber PC, McMinn TR, Chien KR, Johnson TD, Bond RA, Lefkowitz RJ. Enhanced myocardial function in transgenic mice overexpressing the β2-adrenergic receptor. Science. 1994;264:5582-5586.
Luo W, Grupp IL, Harrer J, Ponniah S, Grupp G, Duffy JJ, Doetschman T, Kranias EG. Targeted ablation of the phospholamban gene is associated with markedly enhanced myocardial contractility and loss of β-agonist stimulation. Circ Res. 1994;75:401-409.
Hunter JJ, Tanaka N, Rockman HA, Ross J Jr, Chien KR. Ventricular expression of a MLC-2-v-Ras fusion gene induces cardiac hypertrophy and selective diastolic dysfunction in transgenic mice. J Biol Chem. 1995;270:23173-23178.
Dyson E, Sucov HM, Kubalak SW, Schmid-Schoenbein GW, DeLano FA, Evans RM, Ross J Jr, Chien KR. Atrial-like phenotype is associated with embryonic ventricular failure in RXRα −/− mice. Proc Natl Acad Sci U S A. 1995;92:7386-7390.
Rockman HA, Ross RS, Harris AN, Knowlton KU, Steinhelper ME, Field L, Ross J, Chien KR. Segregation of atrial specific and inducible expression of an ANF transgene in an in vivo murine model of cardiac hypertrophy. Proc Natl Acad Sci U S A. 1991;88:8277-8281.
Rockman HA, Knowlton KU, Ross J Jr, Chien KR. In vivo murine cardiac hypertrophy: a novel model to identify genetic signaling mechanisms that activate an adaptive physiologic response. Circulation. 1993;87(suppl VII):VII-14-VII-21.
Rockman HA, Ono S, Ross RS, Jones LR, Karimi M, Bhargava V, Ross J Jr, Chien KR. Molecular and physiological alternations in murine ventricular dysfunction. Proc Natl Acad Sci U S A. 1994;91:2694-2698.
Ganim JR, Luo W, Ponniah S, Grupp I, Kim HW, Ferguson DG, Kadambi V, Neumann JC, Doetschmann T, Kranias EG. Mouse phospholamban gene expression during development in vivo and in vitro. Circ Res. 1992;71:1021-1030.
Manning WJ, Wei JY, Katz SE, Litwin SE, Douglas PS. In vivo assessment of LV mass in mice using high-frequency cardiac ultrasound: necropsy validation. Am J Physiol. 1994;266:H1672-H1675.
Gardin JM, Siri FM, Kitsis RN, Edwards JG, Leinwand LA. Echocardiographic assessment of left ventricular mass and systolic function in mice. Circ Res. 1995;76:907-914.
Hoit BD, Khoury SF, Kranias EG, Ball N, Walsh RA. In vivo echocardiographic detection of enhanced left ventricular function in gene-targeted mice with phospholamban deficiency. Circ Res. 1995;77:632-637.
Tanaka N, Dalton N, Mao L, Rockman HA, DeMaria AN, Peterson KL, Ross J Jr. Transthoracic echocardiography in mice with surgical arteriovenous fistula and targeted beta receptor overexpression. J Invest Med. 1995;43(suppl 2):385A. Abstract.
Ryan T, Armstrong WF, Feigenbaum H. Annular array technology: application to cardiac imaging. Echocardiography. 1987;4:203-214.
Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation. 1978;58:1072-1083.
Ross J Jr, McCullaugh WH. Nature of enhanced performance of the dilated left ventricle in the dog during chronic volume overloading. Circ Res. 1972;30:549-556.
Kambayashi M, Miura T, Oh BH, Rockman HA, Murata K, Ross J Jr. Enhancement of force-frequency effect on myocardial contractility by adrenergic stimulation. Circulation. 1992;86:563-571.
Bernstein D, Desai K, Rohrer D, Link R, Barsh G, Kobilka B. Resting and stressed cardiorespiratory parameters in the mouse: new tools for the assessment of adrenergic receptor biology. The Mouse as a Model for Cardiovascular and Pulmonary Research. Bethesda, Md: NIH/NHLBI; April 1995. Abstract.