When rat myocytes are incubated in a Ca²⁺– and Na⁺-free (replaced with N-methyl-D-glucamine) medium containing 30 mM Mg²⁺ their cytosolic free Mg²⁺ concentration ([Mg²⁺]_i) increases to high levels. Superfusion of these Mg²⁺-loaded cells with normal Tyrode solution causes [Mg²⁺]_i to return to normal by a mechanism that requires external Na⁺ (Handy et al. 1996). Here we examine the effects of imipramine, an inhibitor of Na⁺/Mg²⁺ antiport, and KB-R7943 (KBR), an inhibitor of reverse Na⁺/Ca²⁺ exchange (Iwamoto et al. 1996), on the Mg²⁺ loading and recovery processes.

Hearts were collected from anaesthetised Sprague-Dawley rats (Na⁺ pentobarbitone, 120 mg kg^-1, I.P.). Cells were isolated and loaded at room temperature with mag-fura-2 as the AM ester (Handy et al. 1996). Changes in [Mg²⁺]_i were monitored at 37 °C by microfluorimetery using dual excitation (340/380 nm). Rates of [Mg²⁺]_i rise or reduction, with or without drugs, were compared within the same cell using a paired t test. Data are presented as means ± S.E.M.

In cells loaded to an [Mg²⁺]_i above about 2 mM, addition of imipramine (200 µM) to the loading medium did not significantly affect the rate of [Mg²⁺]_i rise (control: 0.39 ± 0.17 mM min^-1, with imipramine: 0.47 ± 0.20, n = 4, P > 0.05). The same imipramine concentration completely inhibited Na⁺-dependent [Mg²⁺]_i recovery in Mg²⁺-loaded cells (rate constant of [Mg²⁺]_i reduction, with imipramine: -0.003 ± 0.001, following removal of imipramine: 0.08 ± 0.01 min^-1, n = 3, P < 0.05).

In cells loaded to an [Mg²⁺]_i below 2 mM, 20 µM KBR almost entirely prevented further rise in [Mg²⁺]_i (control rate: 0.21 ± 0.06 mM min^-1, with KBR: 0.028 ± 0.01, n = 4, P < 0.05). However, the drug had little effect on [Mg²⁺]_i elevation when [Mg²⁺]_i was raised above 2 mM (control rate: 1.14 ± 0.01 mM min^-1, with KBR: 0.95 ± 0.08, n = 3, P > 0.05). In addition, 20 µM KBR had no effect on the rate of Na⁺-stimulated [Mg²⁺]_i reduction in Mg²⁺-loaded cells. [Mg²⁺]_i returned rapidly to pre-load levels in all experiments (control rate constant: 0.34 ± 0.03 min^-1, with KBR: 0.32 ± 0.05, n = 5, P > 0.05).

The results suggest that whereas [Mg²⁺]_i reduction in cardiac myocytes occurs mainly through an imipramine-sensitive transport process, probably Na⁺/Mg²⁺ antiport, Mg²⁺ loading occurs through several routes, and that some of the initial rise in [Mg²⁺]_i may occur through the Na⁺/Ca²⁺ exchanger working in reverse mode, carrying Mg²⁺ instead of Ca²⁺.

This work was supported by the Ministry of Health, Bahrain.