Glutamate uncaging

For uncaging experiments, the setup was integrated with an ultra-fast pulsed titanium-sapphire laser tuned to 720 or 725 nm (Chameleon Ultra, Coherent Inc.). The microscope was fixed in position and the stage with mounted micromanipulators was controlled with differential micrometers (Newport). The expanded beam of the laser was reflected off a dichroic mirror (FF700/SP, Semrock) and overfilled the back aperture of the 60× objective. The light intensity and duration of the laser pulse (typically 0.5–1.0 ms) were controlled by a Pockels Cell (Model 350–80, Conoptics Inc.) gated by a mechanical shutter with a 4-ms window. The shutter blocked the small amount of light (<15%) that came through the Pockels cell in the “off” setting. The stationary laser spot [∼0.7 μm, full-width at half-maximum (FWHM)] was positioned in the center of the field of view. Using the fluorescence of coinjected Alexa Fluor-594 (50 μm), we positioned a dendritic spine just next to this location. We estimate that the center of the uncaging spot was ∼1 μm from the edge of the spine. Caged MNI-glutamate (∼2.5–5.0 mm; R&D Systems) added to normal ACSF was bath applied through a recirculating system that had a volume of ∼6 ml. The flash intensity was adjusted to generate an EPSP of ∼1 mV. The amplitudes of the calcium responses were not critical for these experiments since we mostly were interested in the time course of the calcium transients. Transients were measured using either 150 or 300 μm OGB-5N. No difference was observed or expected since backpropagating action potential (bAP)-evoked transients measured with 600 μm OGB-5N had the same fast recovery times as those measured using 150 μm (see Results), indicating that there was no significant buffering at these concentrations.

Typically, recordings were made from one to four pyramidal neurons per animal. Cells were accepted for analysis only if dendrites were close to the surface of the slice and resting potential was below −55 mV. All example traces are from single trials (with one exception) with no temporal filtering. Pseudocolor images of the locations of putative spine signals were spatially smoothed using Matlab subroutine imfilter. The number of trials for each experiment was limited by photodynamic damage. Errors in the measurements are presented as SD. p-values of comparisons were calculated using paired t test in Figures 3 and ​and66 and unpaired t test in Figure 7.

The effect of APV (100 μm) on synaptically evoked Ca²⁺ transients in dendritic spines. APV reduced the magnitude to 0.33 ± 0.09% of control (p = 0.00,006) and half decay times from 30.2 ± 8.6 to 18.7 ± 9.2 ms (p = 0.003) but had no significant effect on the rise times (91% of control; p = 0.5). Changes for each of ten spines are shown.

The effect of NBQX on synaptically evoked Ca²⁺ transients detected with indicators of different K_d values. A, Examples of responses with and without NBQX. The effect of NBQX is greatest when OGB-5N was used. All records are single trials except the recording of NBQX on the OGB-5N cell (average of 3 traces). The effect of NBQX on the somatically recorded EPSP is shown below. B, Histogram of the effect of NBQX on the amplitudes of the synaptic transients detected with the three indicators. The fractional reduction is greater with OGB-5N than with fluo-5F (p = 0.004) or with OGB-1 (p = 0.001; OGB-5N, 150 μm, n = 5; fluo-5F, 300 μm, n = 5; OGB-1, 50 μm, n = 5).

Time courses of Ca²⁺ transients evoked by uncaging MNI-glutamate on a dendritic spine detected with 150 μm OGB-5N. A, Image of a dendritic segment with a targeted spine indicated with a red arrow. A 1-ms flash at 720 nm generated a small EPSP and a calcium transient from the red ROI. The image in the inset shows the same segment without the ROI to make the spine clearer. The pseudocolor difference image (between the times at the ends of the black arrow) shows that the [Ca²⁺]_i increase was only over the spine. B, Enlargement of a part of an image of a fluorescent bead excited by an 800-nm two-photon flash. Each pixel is 0.2 μm. The FWHM of a profile through the image (data not shown) was 0.7 μm. C, Cumulative frequency of rise times from many synaptically activated (n = 63) and uncaging (n = 16) calcium transients. D, Similar cumulative frequencies of half decay times. There is a clear difference between the two profiles.