Perforated patch recordings of neuronal calcium currents (I(Ca)) with amphotericin B or nystatin reduce dialysis of intracellular constituents and current rundown, but can be difficult and frequently unsuccessful. We investigated the saponin beta-escin as a putative ionophore for perforated patch I(Ca) recordings in acutely dissociated, rat dorsal root ganglion neurons. I(Ca) was recorded in time-course studies after including either beta-escin (50 microM), or amphotericin B (240 microg/ml) as perforating ionophores in the internal pipette solution, in comparison to standard ruptured-patch technique, using suction. Perforated patches were allowed to take place spontaneously. The percentage loss of I(Ca) per min (within the first 20 min) was significantly less after beta-escin (0.0518%) (n = 18), versus either amphotericin (1.82%) (n = 12) or standard patch (4.52%) (n = 7), (P < 0.001). The slope of the rundown after linear fit was also less after beta-escin (P < 0.001). Minimal "steady-state" access resistance (R(a)) of 6.6 +/- 1.6 MOmega was achieved within 7.1 +/- 9.3 min following perforation with beta-escin, 7.9 +/- 3.5 MOmega within 44 =/- 14 min after amphotericin B, and 6.8 +/- 1.9 MOmega with standard patch (P < 0.05 for R(a), and P < 0.01 for permeabilization time, respectively). Success rates were 59% with beta-escin versus 27% with amphotericin. Leak >10% of peak I(Ca) was present in 25% of cells after beta-escin versus 20% after amphotericin, and 12% after standard technique. Perforated patches using beta-escin were stable for 15-60 min. We conclude that beta-escin may be used as an alternative ionophore for perforated patch-clamp studies in neurons, and results in minimal rundown that can facilitate long-term recordings of I(Ca). Limited rundown may be due to better preservation of cytosolic ATP content.