FIGURE 3.

Slopes of nearly perfect linear correlations between charge and burst length reveal limited signaling for frequencies higher than 300 Hz as a result of insufficient glutamate release. (A) Correlations between relative charge and burst length at 100, 300, 500, and 700 Hz; no differences were seen in the slopes between 300, 500, and 700 Hz (p > 0.05; n = 6). (B)Similar experiment as shown in A repeated in current clamp. Recording conditions had no effect on the outcome (n = 6). (C) Similar experiment as shown in A, now including 200 Hz. Note that the slope measured for 200 Hz differed significantly from that at 100, 300, and 500 Hz (p < 0.05; n = 8). (D) When CX546 was applied, correlations were less linear as a result of small facilitations for bursts smaller than 20 ms (see text). However, bursts that were shorter than 20 ms in duration still displayed a linear relationship and the slopes at 300, 500, and 700 Hz did not differ significantly from each other (p > 0.05; n = 6). (E) Application of γDGG (1 mM) did not affect the observed limitation of signaling at frequencies higher than 300 Hz (n = 5). (F) γDGG (1 mM) effectively suppressed all responses by more than 35%. Relative suppression was significantly smaller after five stimuli at 200, 300, and 500 Hz, although no further decrease was observed for longer stimulus trains. This indicates that release might be facilitated within the first few stimuli, but quickly reaches its optimum (n = 5).