In the heart, the process by which mechanical forces on the myocardium can induce electrical disturbances (mechano-electric feedback or MEF) has recently excited much interest, because it seems likely that it is responsible for many clinically important arrhythmias. The cellular mechanisms underlying mechanoelectric feedback are uncertain, but it seems likely that they involve mechanosensitive ion channels in the myocardial cell membrane (Sakin, 1995; Hu & Sachs, 1997). The electrical response of the myocardium to stretch can be either depolarising or hyperpolarising (Zabel et al. 1996), suggesting that two different types of channels may be involved. Indeed, stretch-sensitive non-specific cation channels (SACs) and potassium channels have been described (Kim, 1992).

We have used patch-clamp techniques to investigate the properties of a mechanosensitive potassium channel recorded in isolated ventricular myocytes from adult rats. The channel is highly selective for potassium ions, has a conductance of ca 100 pS, and responds strongly to membrane stretch applied as hydrostatic pressure across the membrane patch. Until recently the molecular identity of this channel was unclear. However, in 1998, a new family of potassium channels was cloned (the 4T2P channels: Lesage & Lazdunski, 2000). Members of this gene family are widely expressed in a variety of tissues such as brain, kidney and heart. The channels fall into three phenotypic groups: (1) the acid-sensitive channels TASK-1, -2 and -3, (2) the mechanosensitive channels TREK-1 and -2 and TRAAK and (3) the weakly inwardly rectifying channels TWIK-1 and -2. In the mechanosensitive subfamily only TREK-1 has been found to be expressed in cardiac tissues. TREK-1, when expressed in heterologous systems, has been shown to have the combination of properties that matches our functional measurements in cardiac cells (Maingret et al. 1999), and RT-PCR has shown a gene similar to TREK-1 to be expressed in the heart (Aimond et al. 2000).

The function of TREK in the heart is still a matter of speculation. However, it has been known for some time that the cardiac action potential has a very different morphology in different parts of the heart, a crucial determinant of this action potential heterogeneity being the variation in the expression level of the voltage-dependent potassium channel Kv4.3. If TREK showed a similar heterogeneous expression pattern, it would provide clues as to its function. We have now shown that TREK is indeed differentially expressed in epicardial and endocardial cells from rat hearts, and that the level of expression correlates with the size of an anaesthetic-induced current in these cells.

This differential expression of a mechanosensitive channel may be related to the distribution of stress across the ventricular wall. In this case, the factors that govern the gene expression of TREK, and similar channels, and the way that these change in pathological states such as heart failure and hypertrophy may provide fruitful avenues for new investigations.

This work was supported by the NHMRC.