Histo-anatomical structure of the living isolated rat heart in two contraction states assessed by diffusion tensor MRI

Abstract

Deformation and wall-thickening of ventricular myocardium are essential for cardiac pump function. However, insight into the histo-anatomical basis for cardiac tissue re-arrangement during contraction is limited. In this report, we describe dynamic changes in regionally prevailing cardiomyocyte (fibre) and myolaminar (sheet) orientations, using Diffusion Tensor Imaging (DTI) of ventricles in the same living heart in two different mechanical states. Hearts, isolated from Sprague-Dawley rats, were Langendorff-perfused and imaged, initially in their slack state during cardioplegic arrest, then during lithium-induced contracture. Regional fibre- and sheet-orientations were derived from DTI-data on a voxel-wise basis. Contraction was accompanied with a decrease in left-handed helical fibres (handedness relative to the baso-apical direction) in basal, equatorial, and apical sub-epicardium (by 14.0%, 17.3%, 15.8% respectively; p < 0.001), and an increase in right-handed helical fibres of the sub-endocardium (by 11.0%, 12.1% and 16.1%, respectively; p < 0.001). Two predominant sheet-populations were observed, with sheet-angles of either positive (β+) or negative (β-) polarity relative to a 'chamber-horizontal plane' (defined as normal to the left ventricular long-axis). In contracture, mean 'intersection'-angle (geometrically quantifiable intersection of sheet-angle projections) between β+ and β- sheet-populations increased from 86.2 ± 5.5° (slack) to 108.3 ± 5.4° (p < 0.001). Subsequent high-resolution DTI of fixed myocardium, and histological sectioning, reconfirmed the existence of alternating sheet-plane populations. Our results suggest that myocardial tissue layers in alternating sheet-populations align into a more chamber-horizontal orientation during contraction. This re-arrangement occurs via an accordion-like mechanism that, combined with inter-sheet slippage, can significantly contribute to ventricular deformation, including wall-thickening in a predominantly centripetal direction and baso-apical shortening.

Fig. 1

Schematic illustration to introduce fibre- and sheet-angle definitions. Top: conceptualization of the orientation of a ‘transmural tissue block’, removed from the left-ventricular (LV) free-wall, with reference to the heart's chamber-horizontal plane (blue dashed line). Bottom-Left: schematic representation of transmural cut, with indication of long-section tissue appearance (see Fig. 6 for original transmural long-section histology). Bottom-Right: excised transmural tissue block with an indication of epicardial fibre-orientation (black arrow) and transmural sheet-orientations (green double-lines). The helix-angle (α) describes the deviation (viewed from the epicardium) of fibre-orientation from the horizontal plane; it is negative if fibres point from top-left to bottom-right (and positive if they go from bottom-left to top-right). The angle β is used to describe the deviation, from the heart's chamber-horizontal plane, of the apparent sheet-orientation in transmural long-cuts of the ventricles, and is positive if sheet-projections are directed in an apico-basal (‘upward’) orientation, as one follows them from endo- to epicardium (and vice versa). Also shown is the sheet intersection-angle δ, discussed in more detail elsewhere in the text.

Fig. 2

MRI of a human heart in end-diastole (left) and end-systole (right). Note stable position of apex and base (A, B), and minimal lateral translocation of epicardial boundaries (arrow-head). Intra-ventricular volume reduction is dominated by (i) pronounced centripetal tissue thickening, and (ii) the baso-apical shift of the atrio-ventricular valve-plane (see white outline). This is different from the pronounced apex “skipping” seen in isolated coronary-perfused hearts, where the removal of pericardial constraints (which provide a pleura-like viscous coupling that allows for ventricular twisting and sliding, but prevents tissue disconnection) changes cardiac pump function from the normal mode of combined ventricular pressure and atrial suction generation to a ventricular pressure only mode. Modified from (Petersen, 2006); with permission.

Fig. 3

Photographs of the custom-made Langendorff perfusion rig. Top: view of a heart inside the imaging chamber, mounted to the rigid plastic support tube prior to insertion into the MRI scanner. The perfusate is carried through a water-jacketed line from the reservoir (not shown), through the support tube, and to the Langendorff-mounted heart in the chamber. Effluent solution leaves the system, via gravity, through the support tube. Bottom: magnified view of the heart chamber. Hearts were attached to an aortic cannula inside the chamber and placed on a flexible cradle made from Parafilm suspended between two support rods.

Fig. 4

Summary of changes in helix- (left) and sheet- (right) angles (n = 8 hearts), in basal (top-row), equatorial (middle-row), and apical (bottom-row) regions of the ventricles. The two states are shown in all panels as blue / solid bars (slack) and red / hatched bars (contracture). A, C and E: Binned fractions of left-handed helical fibres (LHF: α < −30°), circumferential fibres (CF: −30° ≤ α ≤ 30°), and right-handed helical fibres (RHF: α > 30°; all data shown as mean ± SD, *source of significant change in fibre population distribution, after decomposition of G statistic by fibre orientation (see Section 3.3)). B, D, and F: Normalized histograms of the distribution of sheet-angles. Solid / dashed lines show the mean values for all slack / contracture hearts studied respectively, while the shaded bands represent the 95% confidence interval for each mechanical state.

Fig. 5

Plots of helix-angles (left) and sheet-orientation (right), observed in an equatorial layer of one and the same perfused rat heart, in slack state (top) and contracture (bottom). A, C: helix-angles. B, D: colour-coded rectangles, representing locally-resolved spatial distribution of sheet-orientation (calculated as planes perpendicular to the tertiary eigenvector of the diffusion tensor). Colour-coding applies to all four panels: red = +90°, blue = −90°, see colour bar); white bars represent the location of the sheet-angle profiles and their spatial extent (local LV wall-thickness) relative to the ventricular cross-sections shown on the left.

Fig. 6

Long-axis histology sections (apical direction pointing down) of rat myocardium, following tissue fixation in contracture (10 μm section thickness, trichrome stained). Cleft spaces with alternating orientation, compatible with multiple transmural sheet populations, can be observed in the right ventricular wall (A), septum (B), and LV wall (C).

Fig. 7

High resolution DTI. A, B: Plot of helix angles derived from the high resolution DTI datasets in two hearts, chemically fixed in either slack (A) or contracture (B) states (red = +90°, blue = −90°). C: Normalized histograms of the distribution of sheet angles throughout the myocardium in the above slack (blue/solid line) and contracture (red/dashed line) hearts, illustrating that in contracture the fraction of sheet angles with low absolute values is higher than in the slack state.

Fig. 8

Schematic illustration of the proposed interaction between fibre- and sheet-orientations, and how their interplay may contribute to wall-thickening and apico-basal shortening. Clockwise, starting top-left: illustration of a transmural tissue-block, removed from LV free-wall (as in Fig. 1), with introduction of the x/y/z coordinate system used in this illustration. Top-Right (x/y): epicardial view onto different layers of the tissue block, illustrating decomposition of forces in x and y directions. Bottom-Right (x/z): top–down view onto the same tissue block (transmural/horizontal plane) with indication of how curvature of ventricular wall-segments affects decomposition of tangential (here taken to equal the x-component) into centripetal forces (z-direction). Bottom-Left (y/z): side-on view of the tissue-block (transmural/baso-apical plane), with indication of sheet-angle re-arrangement from rest (green) to contraction (orange) by applying y- and z-components established above. Please see more detailed explanation in the Discussion section. Illustration is a highly-simplified scheme and not to scale. Epi: sub-epicardial; mid: mid-myocardial; endo: sub-endocardial layers.