Myocardial stretch determines cardiac performance. Upon stretch, there is an immediate increase in contractility (the well-known Frank-Starling mechanism) and a secondary slow increase in contractility (the Anrep effect). The Anrep effect is due to an increase in Ca²⁺ supply to the myofilaments. This phenomenon is observed in the intact heart, in isolated muscle and in the single myocyte, yet the signalling pathway(s) and ultimate mechanism(s) that underlie this slow increase in contractility are a subject of continuing debate.

Several kinases have been identified as playing a role in the slow response to stretch. The protein kinase A (PKA) pathway was the first to be highlighted. We have observed a functionally significant increase in myocardial cAMP during the slow response in ferret papillary muscle (Calaghan et al. 1999). Endothelin 1 (ET1) and angiotensin II (Ang II)-stimulation of protein kinase C (PKC) has also been implicated. In ferret papillary muscle, ET_A and ET_B receptor antagonism attenuated the slow response, although blockade of Ang II receptors was without effect (Calaghan & White, 2001). Other studies using cardiac muscle from the cat have implicated Ang II upstream from ET1 (Pérez et al. 2001). Most recently, Vila Petroff et al. (2001) proposed that activation of the PtdIns-3-OH kinase-Akt-endothelial NO synthase axis mediates the slow increase in [Ca²⁺]_i seen following stretch of rat cardiac myocytes.

In terms of effectors, the target protein phosphorylated by PKA has yet to be identified although, given that the slow response in muscle is not attenuated by pharmacological blockade of the L-type Ca²⁺ channel or the sarcoplasmic reticulum (SR) (e.g. Chuck & Parmley, 1980; Kentish & Wrzosek, 1998), we speculate that increased cAMP and consequent stimulation of PKA occurs in a compartment inaccessible to these sites. Evidence suggests that ET1-dependent activation of PKC during the slow response stimulates Na⁺-H⁺ exchange and thereby Ca²⁺ influx, via reverse mode Na⁺-Ca²⁺ exchange (Pérez et al. 2001). The work by Vila-Petroff et al. (2001) is consistent with S-nitrosylation of the ryanodine receptor enhancing the capacity of the SR to release Ca²⁺.

There is evidence that the mechanism that underlies the slow response in the single cell is different in some respects to that seen in muscle. For example, ET1 antagonism does not attenuate the slow response in the single rat myocyte (Calaghan & White, 2002); neither is an increase in intracellular [Na⁺] seen upon stretch in the single cell, as it is in muscle (Hongo et al. 1996). Furthermore, when endocardial endothelium is removed from papillary muscle using Triton X-100, the slow response is absent (Calaghan et al. 2001). One interpretation of these data is that in intact muscle the slow response is predominantly a paracrine phenomenon, and there is no autocrine release of ET1 from the myocyte.

Although these pathways and effectors have been identified to date there is still a great deal that is not understood. One anomaly is the finding of enhanced capacity of the SR to release Ca²⁺ in the single myocyte (Vila-Petroff et al. 2001), coupled with evidence from muscle that a functional SR is not required for the slow response (e.g. Kentish & Wrzosek, 1998). The difference between the mechanisms that underlie the response of intact muscle and single cells to stretch could be ascribed to paracrine function of endothelial cells (or fibroblasts) present in muscle; however, it is also possible that mechanotransduction pathways are different in cells that are physically isolated from each other.

Finally, what of the components of the cell that act as mechanotransducers? The stretch-activated channel (SAC) has been identified as one such mechanotransducer. We have found recently that the SAC blocker streptomycin (40 mM) reduces the magnitude of the slow response by 80 % in the rat cardiac myocyte. Cationic SACs could allow entry of Ca²⁺, and Ca²⁺-dependent changes in action potential may also be important in this scheme. The role of other potential mechanotransducers (such as the cytoskeleton) in the slow response has yet to be established.

This work was supported by the BHF.