Background
The authors conducted an ultrasound interrogation of a mutant mouse model with a Dnah5 mutation to determine whether cardiac mechanics may be affected by reversal of cardiac situs. This mutant is a bona fide model of primary ciliary dyskinesia, with surviving homozygous mice showing either situs solitus (SS) or situs inversus totalis (SI).
Methods
High-frequency ultrasound interrogations of 27 neonatal and infant Dnah5 mutant mice, 16 with SS and 11 with SI, were conducted using an ultra-high-frequency biomicroscope. Electrocardiographic and respiratory gating were used to reconstruct high-resolution two-dimensional cines at 1,000 Hz, with speckle-tracking echocardiography used to further analyze midchamber and apical rotation.
Results
All SS mice exhibited the expected counterclockwise apical rotation as viewed caudocranially, and surprisingly, the same counterclockwise motion was also observed in SI mice. Speckle-tracking analysis confirmed counterclockwise systolic rotation in both SS and SI mice, and this increased in magnitude from the subepicardium to the endocardium and from the papillary muscles to the apex. The magnitude of apical endocardial rotation was not different for SS and SI mice (5.64 ± 0.75° and 5.76 ± 1.90°, respectively, P = .93). The anatomic segments responsible for the largest components of apical endocardial systolic rotation differed between the SS and SI hearts ( P = .004). In both, the two largest contributors to rotation were offset 180° from each other, but the anatomic regions differed between them. In SS hearts, maximal regional rotation occurred at the anterior mid-septum and posterolateral free wall, while in SI hearts, it was derived from the posterior septum and the anterolateral free wall. Analysis by episcopic fluorescence image capture histology of representative SI and SS mice showed normal intracardiac and segmental anatomy ({S,D,S} or {I,L,I}) without intracardiac defects.
Conclusions
These results show that mirror-image cardiac looping did not result in mirror-image rotation of the morphologic left ventricle. These findings suggest that further studies are warranted to evaluate whether fiber orientation and cardiac mechanics may be abnormal in individuals with reversal of cardiac situs. The results of this study indicate that cardiac looping and myofiber orientation may be independently regulated.
The concept of the helical ventricular myocardial band holds that the heart is the product of a continuous muscle band that assumes its shape by wrapping upon itself. Francisco Torrent-Guasp has been the most recent proponent of this theory, which he exhibited more than 30 years ago by demonstrating that a boiled cow’s heart can be unwrapped in a consistent, predetermined fashion dictated by the planes of the myocardial fibers. The Torrent-Guasp hypothesis has direct implications for myocardial fiber orientation and therefore ventricular mechanics. Furthermore, the determination of fiber orientation may be at least partially independent of ventricular orientation.
Abnormalities of ventricular looping may result in and be particularly sensitive to the impact of abnormal fiber orientation. Our goal, therefore, was to examine the ventricular mechanics of abnormally looped hearts and compare them with normal hearts. A mouse model of primary ciliary dyskinesia (PCD) has been described that has a recessive mutation in Dnah5, an ortholog of the human gene DNAH5, which has been identified as the gene most often mutated in patients with PCD. Analysis of the homozygous mutant fetuses showed a variety of abnormalities of ventricular looping, with 25% having normal organ situs, 40% having heterotaxy, and 35% having situs inversus totalis (SI). However, as the heterotaxy phenotype is lethal prenatally or soon after birth, surviving homozygote neonates have either normal situs solitus (SS) or SI. As a result, transthoracic echocardiography can be used to interrogate cardiac function in the surviving homozygous mutant neonatal mice with SI or SS phenotypes.
In normal hearts, the apex of the left ventricle has a net counterclockwise (CCW) rotation during systole and the base a net clockwise (CW) rotation; it is the difference between these two movements that constitutes twist. Myocardial fibers in the normal heart assume a spiral shape, which in part creates this twisting or wringing motion. Disruptions in this complex helical structure would be expected to cause disruptions in ventricular twisting. Thus, analysis of changes in ventricular twist within the context of a given cardiac malformation can give insight into the macroarchitecture of the heart and its impact on function.
Until recently, qualitative and quantitative observations about ventricular rotation were limited to studies using sonomicrometry, magnetic resonance imaging (MRI), or other novel techniques, which have significant limitations due to invasiveness, availability, or ease of use. In contrast, speckle-tracking echocardiography (STE) allows the rapid analysis of ventricular mechanics using conventional echocardiographic images. This newer modality tracks the dynamic motion of pixels comprising B-mode images of the heart through its cardiac cycle in two-dimensional (2D) space. This allows for qualitative and quantitative assessments of rotation (and other measures of ventricular function) in addition to conventional assessments of ventricular wall motion. Furthermore, STE has been validated against MRI, Doppler tissue imaging, and sonomicrometry in animal and human models including specific measurements of ventricular twist.
Few data exist on functional abnormalities in patients with SI and mirror-image dextrocardia. Historically, there have been observations of unexpected fiber orientation in SI on the basis of gross examination, specifically, that deep muscle layers had near mirror-image fiber orientation but that superficial layers did not have mirror-image symmetry. However, little attention was given to the functional implications. Recent studies using MRI have suggested that ventricular deformation does not conform to the expected mirror-image pattern, but assessments of left ventricular (LV) rotation in SI using conventional echocardiography or STE are lacking. Additionally, this evaluation has not been carried out in a pediatric model, and it has been established that there are normal age-related changes in ventricular twist. Given recent advances in technology, the determination of this aspect of function is relatively easy to acquire via ultrasound. The aim of this study was to assess whether mice with SI would have ventricular rotation that differed in magnitude and/or direction from that of normal mice. Additionally, our goals included the use of high-frequency ultrasound in very small neonatal mice with congenital heart disease and the application of speckle-tracking technology to high-frequency electrocardiographically gated echocardiographic images.
Methods
All animal studies were conducted under an animal study protocol approved by the National Heart, Lung, and Blood Institute’s Animal Care and Use Committee.
Mouse Breeding and Echocardiography
Heterozygous Dnah5 mutant mice were maintained in the C57BL6/J or C3H background and were intercrossed to generate homozygous offspring. Mice were observed at birth for cardiac and stomach situs, which can be evaluated easily through the translucent body wall. Visualization of the position of the heart and stomach (via the milk spot) allowed the identification of mice that had either SS and normal cardiac anatomy or SI and mirror-image dextrocardia. Sixteen mice with SS and 11 mice with SI were selected for echocardiography. Both phenotypes were frequently seen in the same litter and scanned on the same day. Imaging was performed between 1 and 20 days of life using the VisualSonics Vevo 770 (VisualSonics, Inc, Toronto, ON, Canada) with a 30-MHz or 40-MHz mechanical transducer. The mouse handling table of the VisualSonics Integrated Rail System was modified to enable electrocardiographically gated imaging of neonatal mice. Copper strips were used to extend the limb leads to a size appropriate for newborn mice. Anesthesia was induced using 4% isoflurane and maintained throughout the imaging using 1.5% to 2.5% isoflurane. Warmed ultrasound gel and a heating lamp maintained a body temperature of 35°C to 37°C. The transducer was positioned to optimize image quality and minimize stationary artifacts induced by lung or skeletal tissue. High resolution one-beat cine loops were recorded at presentation quality using electrocardiography-based kilohertz visualization (EKV) imaging. A single-beat high-resolution 2D cine image was reconstructed using 1,000 M-mode images gated to the electrocardiographic and respiratory channels. Data were saved and exported to uncompressed Digital Imaging and Communications in Medicine format files for offline analysis using STE. Images were reviewed by two pediatric cardiologists for the determination of systolic twist. Distributions of direction of twist in SS and SI mice were compared using Fisher’s exact test.
Anatomic Analysis
To confirm the ultrasound assessment of situs and to exclude complex congenital heart disease, euthanized mice were fixed in 10% phosphate-buffered formalin and necropsied to identify the cardiac and abdominal situs. Further anatomic analysis to exclude live-born mice with intracardiac abnormalities was carried out using episcopic fluorescence image capture (EFIC), a histologic imaging technique that provides high-resolution 2D image stacks. For EFIC analysis, the specimens were processed using a standard protocol involving dehydration in ethanol and xylene followed by infiltration and embedding in 70% paraffin wax containing 25% Vybar, 4.4% stearic acid, and 0.4% aniline dye Sudan IV. The paraffin blocks were sectioned using a sliding microtome (Leica SM2500; Leica Mikrosystem Vertrieb GmbH, Wetzlar, Germany), with the autofluorescence in the block face visualized using epifluorescent illumination with a mercury lamp and excitation and emission filters of s425 and 480 nm, respectively. Fluorescence images were captured serially using a low-light charge-coupled device camera (Hamamatsu ORCA-ER; Hamamatsu Photonics, Hamamatsu City, Japan) to generate registered 2D image stacks. For 3-dimensional rendering, the 2D image stacks were processed using Improvision Volocity software (PerkinElmer, Waltham, MA). The 2D image stacks were viewed in different imaging planes digital resectioning to evaluate the internal anatomy of the heart.
Rotational Analysis
Myocardial motion was analyzed using Velocity Vector Imaging (VVI), a speckle-tracking software package (Siemens Medical Solutions, Inc, Malvern, PA). Single-beat EKV short-axis views obtained at the level of the papillary muscles and apex were used for analysis. Initial traces were performed on all images at early systole of the cardiac cycle. Each image underwent analysis at the endocardium–ventricular chamber border and within the subepicardial wall. Three traces were performed at each of these points on the myocardium (endocardium and subepicardium) as well as at each level within the ventricle (papillary muscles and apex). Repeat measurement variability was calculated by anatomic level (apex or papillary muscles) and component of the myocardium (endocardium or subepicardium) for each mouse undergoing quantitative analysis and is reported as the mean percentage deviation from the average of the 3 individual measurements. Additionally, the intracluster correlation coefficient (ICC) was used to assess the reproducibility and correlation of the repeated measurements from the same mouse. The ICC was estimated as the proportion of the total variance explained by the between-subject variance on the basis of a one-way random-effects analysis of variance.
The cine images with superimposed velocity vector data were visually screened for the quality of motion tracking. Rotational displacement during systole was calculated in degrees for each of 6 evenly divided segments of the myocardium as well as for the average for the entire short-axis image. The 6 segments were defined as follows: septal, anteroseptal, anterolateral, lateral, posterolateral, and posteroseptal. These segments were defined and numbered automatically by the software algorithm regardless of cardiac situs. Wilcoxon’s rank-sum test was used to compare global twist between the SS and SI mice, and the two-sample Hotelling’s t 2 test statistic was used to compare mean vectors of percentage contributions from the 6 segments between the two groups of mice. A P value < .05 was considered statistically significant. The statistical analysis was performed using SAS version 9.1 (SAS Institute Inc, Cary, NC).
Results
Mice were interrogated using the Vevo 770 ultra-high-frequency biomicroscope. The mice were anesthetized with isoflurane titrated to maintain the heart rate between 350 and 500 beats/min. Electrocardiographic data were obtained using 4 limb leads for monitoring heart rate and gating of the echocardiographic images. Initial imaging consisted of a wide-angle short-axis view including the chest wall and ribs to identify cardiac situs ( Figure 1 ). Further imaging was conducting with the field of view zoomed in to maximize resolution of the heart. Images were obtained at the level of the LV papillary muscles and cardiac apex. For each level, multiple B-mode clips were obtained. Additionally, high-resolution one-beat cine loops at 1,000 Hz were reconstructed using EKV imaging.
A total of 27 neonatal and infant mice (mean age, 11.1 ± 5.0 days; mean weight, 5.9 ± 3.2 g) were interrogated using echocardiography, 16 with SS and 11 with SI. The results are summarized in Table 1 . Fourteen of the 16 SS mice and 9 of the 11 SI mice provided ultrasound data suitable for observation of apical rotation. Thirteen of 14 SS mice exhibited CCW apical rotation in systole as viewed caudocranially; 1 mouse did not display easily discernable ventricular rotation. Nine of 9 SI mice also exhibited systolic CCW rotation, contrary to the expected mirror symmetric CW rotation. Fisher’s exact test for the 23 mice with adequate images showed that the distribution of CCW rotation was similar in both populations ( P = 1). A conservative imputation analysis assuming that the SS mice with inadequate images rotated CCW and the SI mice with inadequate images rotated CW again showed no significant difference in their distributions of rotation (Fisher’s exact test, P = .16). EFIC analysis of 5 SI and 3 SS mice showed normal segmental anatomy ({S,D,S} or {I,L,I}), without evidence of intracardiac abnormalities.
| Animal | Age (d) | Weight (g) | Cardiac phenotype | Apical systolic rotation |
|---|---|---|---|---|
| 1 | 11 | 2.6 | SS | CCW |
| 2 | 14 | 10 | SS | Image inadequate |
| 3 | 14 | 5.1 | SS | Image inadequate |
| 4 | 17 | 10.1 | SS | CCW |
| 5 | 17 | 10 | SS | CCW |
| 6 | 13 | 9.4 | SS | No obvious twist |
| 7 | 13 | 8.7 | SS | CCW |
| 8 | 18 | 9.6 | SS | CCW |
| 9 | 10 | 8.1 | SS | CCW |
| 10 | 14 | 4.8 | SS | CCW |
| 11 | 14 | 7.2 | SS | CCW |
| 12 | 14 | 12.5 | SS | CCW |
| 13 | 20 | 10.2 | SS | CCW |
| 14 | 10 | 2.8 | SS | CCW |
| 15 | 14 | 4.6 | SS | CCW |
| 16 | 14 | 5 | SS | CCW |
| 17 | 14 | 6.2 | SI | Image inadequate |
| 18 | 13 | 2.7 | SI | CCW |
| 19 | 10 | 2.5 | SI | Image inadequate |
| 20 | 3 | 1.9 | SI | CCW |
| 21 | 5 | 3.5 | SI | CCW |
| 22 | 5 | 3.2 | SI | CCW |
| 23 | 5 | 4.7 | SI | CCW |
| 24 | 5 | 3.6 | SI | CCW |
| 25 | 6 | 4.7 | SI | CCW |
| 26 | 6 | 5.4 | SI | CCW |
| 27 | 1 | 1.4 | SI | CCW |
Speckle-tracking analysis was performed using VVI on the high-resolution reconstructed cine loops obtained from 5 SS and 6 SI mice. This confirmed CCW rotation in both groups, with the magnitude of rotation greater toward the apex than near the papillary muscles. The magnitude of CCW rotation was also greater at the endocardium than at the subepicardium. Comparison of the magnitude of rotation between SS and SI hearts showed no significant difference at the apical endocardium, apical subepicardium, papillary muscle endocardium, or papillary muscle subepicardium ( P = ns for all groups; Table 2 ). Vector diagrams help illustrate the magnitude and direction of endocardial movement ( Figure 2 ), with the direction of each arrow indicating the direction of instantaneous tangential motion and the length of the arrow indicating the velocity of the myocardium. This analysis showed that CCW rotation was present throughout systole, while CW rotation was observed in diastole. Repeat measurement variability as measured by mean percentage deviation for individual measurements ranged from 7.5% to 13.1% at different points and levels, except those at the papillary muscles and subepicardium (22.1%) ( Table 2 ). The ICC for measurements at two points (the apex and papillary muscles) of endocardium ranged from 0.715 to 0.886, suggesting a strong correlation and good reproducibility. At the papillary muscles and subepicardium, the ICC was 0.625, suggesting a moderately strong correlation and fair reproducibility ( Table 2 ).
| Apex | Papillary muscles | |||
|---|---|---|---|---|
| Variable | Endocardium | Subepicardium | Endocardium | Subepicardium |
| Rotation ∗ | ||||
| SS (n = 5) | +5.64 ± 0.75° | +2.32 ± 1.27° | +2.92 ± 1.14° | +1.22 ± 0.31° |
| SI (n = 6) | +5.76 ± 1.90° | +2.67 ± 1.16° | +2.94 ± 0.89° | +0.96 ± 0.50° |
| Wilcoxon’s test exact P value | .93 | .66 | .93 | .33 |
| Repeat measurement variability † | 9.6 ± 5.6% | 13.1 ± 13.1% | 7.5 ± 4.6% | 22.1 ± 11.2% |
| ICC ‡ | 0.715 | 0.841 | 0.886 | 0.625 |
∗ Mean ± SD as viewed in the short axis caudocranially; positive values denote CCW rotation and negative values denote CW rotation.
† Mean percentage deviation of individual measurements from the mean rotation for each mouse.
‡ Estimated on the basis of a one-way random-effects analysis of variance.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
