Fig. 2. Selective labeling of axons and glial cells. A, Calcium transients evoked by single action potentials in a P6 rat optic nerve, in which the axons were selectively labeled with the calcium indicators by diffusion through the cut ends. For selective labeling of axons, the suction pipette enclosing the nerve end contained calcium green-1 hexapotassium salt (membrane impermeant) dissolved to a final concentration of 10 mm in a buffer containing (in mm): KCl, 150; MgCl2, 4.6; CaCl2, 0.1; EGTA, 1; HEPES, 10; Na-GTP, 0.4; and Na-ATP, 4, pH 7.3. A loading time of 4–5 hr was found to give adequate labeling of axons for detection of evoked calcium transients. In these experiments, the glial cells at the immediate vicinity of the loading pipette were stained, but unlike the axons, no transfer of the dye through glial gap junctions to the middle portion of the nerve (at which calcium imaging was performed) was seen. This may be caused either by closing of the gap junctions at the damaged end of the nerve or by a much slower diffusion rate of the dye through the gap junction channels compared with that along the axoplasm of axons. The calcium transient shows properties similar to that shown in Figure 1 (left), in which both axons and glial cells were labeled. The calcium transient was inhibited by Ni2+ (middle) and recovered after washing (right). B, A stimulation experiment in which only the glial cells were stained by impalement with a sharp microelectrode. Brief electrical stimulation failed to elicit a calcium transient (left). The sharp microelectrode contained calcium green-1 hexapotassium salt dissolved to a final concentration of 12 mm in 140 mm KCl plus 10 mm HEPES, pH 7.2. The glial cells were responsive to bath application of adenosine (middle,right).