Simultaneous Sodium and Calcium Imaging from Dendrites and Axons

Abstract

Dynamic calcium imaging is a major technique of neuroscientists. It can reveal information about the location of various calcium channels and calcium permeable receptors, the time course, magnitude, and location of intracellular calcium concentration ([Ca(2+)]i) changes, and indirectly, the occurrence of action potentials. Dynamic sodium imaging, a less exploited technique, can reveal analogous information related to sodium signaling. In some cases, like the examination of AMPA and NMDA receptor signaling, measurements of both [Ca(2+)]i and [Na(+)]i changes in the same preparation may provide more information than separate measurements. To this end, we developed a technique to simultaneously measure both signals at high speed and sufficient sensitivity to detect localized physiologic events. This approach has advantages over sequential imaging because the preparation may not respond identically in different trials. We designed custom dichroic and emission filters to allow the separate detection of the fluorescence of sodium and calcium indicators loaded together into a single neuron in a brain slice from the hippocampus of Sprague-Dawley rats. We then used high-intensity light emitting diodes (LEDs) to alternately excite the two indicators at the appropriate wavelengths. These pulses were synchronized with the frames of a CCD camera running at 500 Hz. Software then separated the data streams to provide independent sodium and calcium signals. With this system we could detect [Ca(2+)]i and [Na(+)]i changes from single action potentials in axons and synaptically evoked signals in dendrites, both with submicron resolution and a good signal-to-noise ratio (S/N).

Keywords: calcium; imaging; sodium.

Figure 1.

Design of dichroic and emission filters for the simultaneous detection of sodium and calcium changes. a, Excitation and emission spectra of SBFI and OGB-1 in cuvettes (from the Chroma Technology website). The spectra have been arranged so the same wavelengths are aligned vertically. The excitation spectrum of SBFI shows the sodium bound form. The spectra of OGB-1 are much narrower. b, Design of custom dichroic and emission filters to detect responses from these indicators. The dichroic cuts on at 390 nm to allow excitation of SBFI at 380 nm. The notch at 460 allows excitation of OGB-1, which has a significant shoulder at that wavelength. The emission filter passes most of the emitted fluorescence from both indicators, missing only a part of the SBFI fluorescence around 460 nm. c, Spectra of bis-fura-2 and ANG-2 (from the Molecular Probes website and Roder and Hille, 2014). d, Design of custom dichroic and emission filters to detect responses from these indicators.

Figure 2.

a, Arrangement of LEDs at the excitation port of the microscope fluorescence illuminator. The outputs of the two LEDs are combined with a dichroic mirror centered at 425 nm. b, Spectra of LEDs used in our experiments (from Prizmatix catalogue). They are nominally centered at the indicated wavelengths but each emits some light over a 50–100 nm range. Below the spectra are the bandpass ranges of excitation filters used to limit the excitation light. The sequence of frame illumination in a typical experiment; at 500 Hz each frame has duration of 2 ms (dotted vertical lines). The LED turns on 0.1 ms after the start of the frame and turns off 0.1 ms before the end of the frame. In the next frame, the second LED is on for the same duration. This alternating sequence continues for the duration of the trial. A similar arrangement was used when pairing the 385 and 520 nm LEDs.

Figure 3.

Simultaneous sodium and calcium imaging from an apical dendrite of a hippocampal CA1 pyramidal neuron. The image shows a small region of the apical dendrites (32 × 32 µm²). The fluorescence is from ANG-2 injected from a patch electrode on the soma. The black ROI is ∼3 × 5 µm²; the smaller orange ROI is ∼0.8 × 0.8 µm². Five pulses at 100 Hz from a stimulating electrode on the SC near the ROI generated a synaptic response that evoked a single AP. The optical recording on the left shows the fluorescence levels in the black ROI when the cell was alternately illuminated at 500 Hz at 385 and 520 nm. The last 20 ms are stretched out to show the alternating response more clearly. There is an envelope to the signals at the top and bottom of the recording (arrows). The traces on the right show the records when the two channels are separated. Traces were corrected for bleaching (see Discussion). There are clear signals in both the sodium and calcium channel. Note that there is a delay in the onset of the sodium signal compared to the spike time, whereas the calcium signal is synchronous with the spike. This is clearer in the black ROI, which has less noise because it integrates over a larger area. The calcium signal goes down because bis-fura-2 fluorescence decreases when [Ca²⁺]_i increases.

Figure 4.

Simultaneous recording of spike evoked sodium and calcium signals from the axon of a CA1 pyramidal neuron. a, Left, pyramidal neuron filled with ANG-2 (200 µM) and bis-fura-2 (200 µM) from a patch electrode on the soma. Intrasomatic current injection evoked a single spike that generated a sharp fluorescence change in each channel at the ROI (yellow box, single pixel) on the axon. The sodium signal appears to rise within one frame, the calcium signal within two frames. There is no delay from the time of the AP. Traces have been corrected for bleaching. b, Plot of the relative ΔF/F of the two signals along the axon, normalized to the peak values (average of 2 cells). These measurements were made with 3 × 3 µm² ROIs to improve the S/N of the measurements. The distributions are clearly different, with the calcium signal peaking in the soma and the sodium signal peaking ∼30 µm from the center of the soma.

Figure 5.

Simultaneous detection of sodium or calcium changes with two different indicators. a, Pyramidal neuron filled with both ANG-2 and SBFI from a patch electrode on the soma. The ANG-2 filter set (Fig. 1d ) was used with excitation filters centered at 380 and 520 nm. A single AP was evoked with a 2 ms intrasomatic current pulse. ANG-2 and SBFI fluorescence changes in the axon were detected in the two channels. The downward slope indicates strong bleaching due to the high light intensity. The combined trace shows the SBFI trace (1.5×) subtracted from the ANG-2 trace. Most of the bleaching is cancelled and the signal enhanced. Typical result from three cells. b, Similar recordings in the dendrite when two calcium indicators were used. The OGB-1 filter set (Fig. 1b ) was used with excitatory filters centered at 387 and 470 nm. The combined trace is bis-fura-2 (BF-2; 1.3×) subtracted from the OGB-1 trace. Typical result from four cells.

Figure 6.

The switching protocol does not introduce additional noise, which is dominated by shot noise. a, Recording of fluorescence levels from two positions in a field containing a pipette filled with ANG-2 while the illumination at 520 nm was switched on and off on alternate frames. Only the signals from the illuminated frames are shown. There clearly is a higher noise level from the position on the pipette. The downward slope is due to bleaching of ANG-2. b, Plot of the SD of the noise level at five different illumination intensities at the red ROI. The yellow diamonds show the results from the condition in a when only alternate frames were illuminated (“half”). The black squares show the SD when the same measurement was made (only alternate frames) but when the illumination was the same in all frames (“uniform”). There is essentially no difference between the two conditions. Both the diamonds and squares appear to lie along a straight line, the result expected when the system is shot noise limited.