Calcium permeability of the neuronal nuclear envelope: Evaluation using confocal volumes and intracellular perfusion

Abstract

In many calcium-imaging studies, the nuclear envelope appears to maintain a gradient of free calcium between the nucleus and cytosol. This issue was examined by loading amphibian sympathetic neurons with the calcium indicator fluo 3 via whole-cell patch clamping. Confocal optical sectioning allowed acquisition of independent calibration curves for the nucleus and cytoplasm. Cells were loaded with free calcium levels ranging from 10 nM to 50 μM, using 10 mM BAPTA to control free calcium. The nuclear fluorescence was usually about 130% brighter than the cytoplasmic fluorescence. Had the increased nuclear fluorescence been due to a calcium gradient, then, as fluo 3 was saturated with calcium in both compartments, the fluorescence gradient should have gradually disappeared. Instead, with free-calcium in the pipette set at 50 μM, about five times the level required to nearly saturate fluo 3, the nuclear/cytoplasmic (N/C) fluorescence ratio was not decreased but instead increased slightly. Perfusion of the patch pipette was used in conjunction with imaging to confirm that cytoplasmic fluo 3 was saturated with calcium. After loading cells with 10 nM free calcium, the patch pipette was perfused with high calcium (10 μM). Again, the N/C fluorescence ratio increased at high calcium. The effectiveness of patch-pipette perfusion in changing cellular free calcium levels was indicated by the degree of fluorescence increase-both nuclear and cytosolic compartments showed a roughly 20-fold increase in fluorescence, that is, most of the dynamic range observed in test droplets. To confirm further that cytoplasmic fluo 3 was saturated, cells were perfused with manganese, which binds with very high affinity to fluo 3. Manganese rapidly entered the cytoplasm and nucleus, causing a large increase in fluorescence, but the N/C fluorescence ratio remained relatively constant. Because free manganese in the pipette was 50,000 times the amount required to saturate fluo 3, the greater nuclear fluorescence probably results from additional fluo 3 in the nucleus rather than from calcium or manganese gradients. To gauge further the permeability of the nuclear envelope, the diffusion of calcium was visualized. Under voltage clamp, calcium channels were opened for periods ranging from 5 to 200 msec. Peak calcium levels were observed within 2 μm of the plasma membrane, and declined as calcium diffused into the cell. The nuclear fluorescence increased more than cytosolic fluorescence, but this apparent 'amplification' was eliminated by correcting for autofluorescence. Use of cells cultured on glass coverslips and a high-NA microscope objective allowed a satisfactory correction. The size of nuclear responses was proportional to the size of the calcium influx, with a 5 msec depolarization raising nuclear calcium by roughly 50 nM. In other experiments, nuclear calcium rose by about 10 nM in response to single action potentials. Line scans acquired at 2 msec intervals showed influx of calcium into the nucleus within 10-20 msec after the calcium transient arrived at the nuclear envelope. In each instance, out of 103 trials in 49 cells, cytosolic calcium transients spread directly into the nucleus. The simplest explanation is that calcium freely diffuses through the nuclear pores. This is consistent with numerous reports showing that molecules smaller than 20 kDa pass through nuclear pores in an unrestricted fashion. Diverse neural phenomena such as learning, ischemic damage, and Alzheimer's disease might directly involve actions of nuclear calcium. If so, the permeability of the nuclear envelope will influence their course.