The physiological significance of the mechanical heterogeneity of myocardium in the regulation of its contractile function is not yet clear. We developed a new experimental approach to investigate this question, an In-Series Hybrid Duplex, which joined the family of earlier implemented duplexes of different configurations (Markhasin et al. 2003).

The hybrid duplex combines a biological cardiac muscle preparation with a ‘virtual muscle’ (a mathematical model), which interact in real-time, mimicking mechanical connection in-series. In-series elements actually influence each other during the isometric mode of duplex contractions, where the special control algorithm supports the equality of forces of natural and virtual muscles, at the constant sum of their lengths. Parameters of virtual muscles were selected to make uncoupled duplex elements different in their mechanical characteristics of contractions (Fig. 1A, ‘fast’ and ‘slow’ muscles), such as time to peak force.

For experiments all animals were anaesthetized with a lethal dose of pentobarbital (150 mg kg^-1). Results obtained in hybrid duplexes were compared with the activity of pure virtual combinations. In addition to the duplex settings, 1D models of heterogeneous cardiac tissue composed of serial chains of virtual elements have also been studied. Mechanical parameters of the chain elements varied between extreme values, which were used for the fast and the slow element of corresponding duplexes. Two types of heterogeneity were considered – (i) a regular pattern (gradual change in mechanical properties from the fast boundary element to the slow one), and (ii) a random array (uniform distribution of properties within the chain).

Mechanical effects of a time lag between stimulation of the elements imitating excitation propagation in heart tissue were investigated. Hybrid duplexes typically showed a stable (even increasing) contractile response on the stimulation delay of a faster element. On the contrary, the stimulation delay of a slower partner caused a steep decrease in contractility. The results confirmed predictions, which were made in virtual duplexes merging fast and slow elements (Fig. 1B). Mechano-dependent changes in Ca²⁺ handling (kinetics of Ca²⁺-troponin C complexes, Ca²⁺ transients) occurring in virtual elements of hybrid duplexes due to the stimulation sequences were very similar to those obtained in the corresponding elements of virtual duplexes, thus validating model assumptions (Fig. 1C). 1D models of cardiac tissue with ‘regular’ mechanical heterogeneity revealed stable contractility with an increasing delay (entrance-phase) in the activation of boundary elements if excitation had propagated from the slowest element to the fastest one. Inverted excitation sequence of the same models led to a decrease in force generation (Fig. 1D). Isometric force produced by the random distribution models fell between the forces generated by homogeneous samples, which consisted of either only fast or only slow elements (Fig. 1D). In these models, an increase in the time lag in stimulation of boundary elements was followed by decreasing peak force.

The results obtained show that contractile elements in heterogeneous myocardium do not act as ‘independent’ generators of tension/shortening because their inotropic characteristics change dynamically due to mechanical interaction. We conclude that special spatio-temporal heterogeneity is a physiological necessity to optimize cardiac performance. Alterations of the physiological heterogeneity (including inverted sequence of activation) will have pathological consequences.