Endogenous direct-current electric fields (dcEFs) occur in vivo in the form of epithelial transcellular potentials or neuronal field potentials, and a variety of cells respond to dcEFs in vitro by directional movement. This is termed galvanotaxis. The passive influx of Ca²⁺ on the anodal side should increase the local intracellular Ca²⁺ concentration, whereas passive efflux and/or intracellular redistribution decrease the local intracellular Ca²⁺ concentration on the cathodal side. These changes could give rise to `push-pull9 effects, causing net movement of cells towards the cathode. However, such effects would be complicated in cells that possess voltage-gated Ca²⁺ channels and/or intracellular Ca²⁺ stores. Moreover, voltage-gated Na⁺ channels, protein kinases, growth factors, surface charge and electrophoresis of proteins have been found to be involved in galvanotaxis. Galvanotactic mechanisms might operate in both the short term (seconds to minutes) and the long term (minutes to hours), and recent work has shown that they might be involved in metastatic disease. The galvanotactic responses of strongly metastatic prostate and breast cancer cells are much more prominent, and the cells move in the opposite direction compared with corresponding weakly metastatic cells. This could have important implications for the metastatic process and has clinical implications. Galvanotaxis could thus play a significant role in both cellular physiology and pathophysiology.

A two-part hypothesis has been tested, which proposes that (1) prostate cancer cells are galvanotactic (i.e. respond to an electric field by moving directionally) and (2) voltagegated Na⁺ channel activity, which was shown previously to be expressed specifically by strongly metastatic cells, controls galvanotaxis. Two well-defined rat (`Dunning') cell lines, originally derived from the same prostate tumour but differing markedly in their metastatic ability, were used. Cells were exposed to exogenous direct-current electric fields of physiological strength (0.1-4.0 V cm^-1), their reactions were recorded by light microscopy and analysed by a quantitative tracking method. Voltage-gated Na⁺ channel activity was modulated pharmacologically using a range of concentrations of a specific channel blocker (tetrodotoxin) or an opener (veratridine). The results showed that the highly metastatic MAT-LyLu cells responded to the application of the electric field strongly by migrating towards the cathode. By contrast, the weakly metastatic At-2 cells gave no such response. Tetrodotoxin suppressed the galvanotactic response of the MAT-LyLu cells whereas veratridine enhanced it. Both compounds had little effect on the AT-2 cells. These results are consistent with functional voltage-gated Na⁺ channel expression occurring specifically in highly metastatic cells. This is also the first demonstration of control of galvanotaxis, in any cell type, by voltage-gated Na⁺ channel activity. The possible underlying mechanisms and the in vivo relevance of these findings are discussed.