1. We used the patch-clamp technique, in conjunction with membrane capacita
nce measurement, fluorescence measurement of intracellular calcium concentr
ation ([Ca2+](i)), and flash photolysis of caged Ca2+ to study exo- and end
ocytosis in identified rat corticotrophs.
2. Exocytosis stimulated by depolarization pulses was typically followed by
a 'slow' endocytosis that retrieved the membrane with a time constant of s
imilar to6 s. The efficiency (the endocytosis/exocytosis amplitude ratio) o
f 'slow' endocytosis was similar to1.2 at [Ca2+](i) < 3 muM and increased t
o similar to1.6 at [Ca2+](i) > 3 muM.
3. Whole-cell dialysis through a patch pipette did not affect the kinetics
and the efficiency of 'slow' endocytosis, but the amplitude of exocytosis w
as reduced.
4. 'Slow' endocytosis did not require sustained [Ca2+](i) elevation and its
kinetics was only weakly [Ca2+](i) dependent. Our results suggest that 'sl
ow' endocytosis involves a Ca2+ sensor with a high Ca2+ affinity (similar t
o 500 nM).
5. At high [Ca2+](i) (> 10 muM), the 'slow' endocytosis was frequently prec
eded by a 'fast' endocytosis that comprised multiple steps of rapid decreas
e in membrane capacitance.
6. Neither calmodulin nor calcineurin appeared to be the Ca2+ sensor for en
docytosis because the two forms of endocytosis were not affected by the cal
modulin inhibitor calmidazolium (500 muM) or the calcineurin inhibitors cyc
losporin A (1 muM) and calcineurin autoinhibitory peptide (1 mg ml(-1)). Ba
2+, a poor activator of calmodulin, could support both forms of endocytosis
but slowed the kinetics of 'slow' endocytosis similar to2-fold.
7. Non-hydrolysable analogues of GTP (GDP-beta -S) and ATP (ATP-gamma -S) a
lso failed to inhibit either form of endocytosis, indicating that neither G
TP nor ATP was essential for endocytosis.
8. We suggest that the high Ca2+ affinity of 'slow' endocytosis may be impo
rtant for maintaining continuous cycles of exocytosis-endocytosis during su
stained adrenocorticotropin secretion in corticotrophs.