2.1.

System Design

The microscope that was modified is a spectral confocal scanning Olympus FluoView1000. It has a spectral scan unit connected to an inverted microscope IX81. A laser combiner module is used to deliver light from several lasers (multiline Argon, green He–Ne) for confocal illumination. An external mercury lamp (100 W Olympus) is used for conventional wide field fluorescence imaging. Both illumination systems are connected to the scan unit via fiber optic ports.

In our modification, the UV light is provided by the external mercury lamp that is used for conventional fluorescence microscopy (100 W Olympus) but instead of having it connected to the microscope via the conventional fluorescence mercury lamp fiber optic port, the light is introduced via the SIM port (see Fig. 1). The SIM port is supposed to be used to incorporate a second scanner unit dedicated for laser light stimulation, in addition to the primary scanner. In our modification we have added some optical elements to direct the light from the fiber optic to the SIM port. To that end, we have used the same optics as that of the conventional fiber optic port plus a fluorescence filter 377/50 (Brightlines) to select the short wavelengths that are necessary to perform the uncaging. All these elements are placed in a black tube inside a container that can be moved vertically so that the circular output aperture of the fiber cable can be positioned at the microscope's image plane, producing a disk of UV energy at the object plane of the microscope. The UV spot's size depends on the fiber optic diameter and the magnification of the microscope objective. One adjustment screw allows the axial location of the UV spot to be changed relative to the microscope focal plane. As the fiber cable is composed of several fiber optics, if the circular output aperture of the fiber cable is placed exactly at the microscope's image plane, the image of the different fibers can be seen at the scanned plane obtaining a nonuniform illumination. In order to prioritize homogeneity, we slightly defocus the image so that a more uniform UV illumination can be obtained at the scanning plane. Figure 1a shows a lateral view of the confocal scan unit, with the laser combiner fiber port, the conventional fluorescence mercury lamp fiber port and the adaptation tube connected to the SIM port. Figure 1b shows a more detailed photograph of the adaptation tube for the uncaging mode.

Fig. 1

Photograph of the spectral confocal unit with the modification. (a) Lateral view showing the three illumination ports: the conventional fluorescence port for the mercury lamp light, the laser combiner port for confocal imaging, and the modified SIM port in which the mercury lamp light enters for photolysis purposes allowing simultaneous uncaging and confocal imaging. (b) Detailed view of the modified SIM port showing the vertical displacement that allows the axial location of the UV spot to be changed relative to the microscope focal plane.

All the different illumination beams enter through the scan unit to the microscope IX81. The scan unit has a motorized three position rail (see Fig. 2). The first position has a mirror that redirects the light of the mercury lamp in the conventional wide field fluorescence observation mode (in this case light from the laser combiner cannot pass to the microscope). The second one is empty, so that the light from the laser combiner can reach the microscope (but the light from the mercury lamp cannot pass). In our modification we have used the third position for the uncaging placing a dichroic mirror to reflect the short wavelengths (<450 nm) and transmit the long ones (>450 nm). In this way the reflected beam can be used for the photolysis. This last configuration allows the transmission of the light from the laser combiner to occur simultaneously with the UV illumination from the mercury lamp. It also allows the transmission of the light from the sample to the confocal detectors, so that confocal imaging can be performed at the same time.

Fig. 2

Scheme of the different illumination paths that may be used by changing the rail position: (a) Position 1 has a mirror that redirects the light of the mercury lamp for conventional wide field fluorescence observation (A: conventional fluorescence port; B: to IX81 microscope). (b) Position 2 is empty so that the light from the laser combiner can reach the microscope for confocal imaging (c) Position 3 has a dichroic mirror that reflects the short wavelengths (<450 nm) of the mercury lamp light to do the photolysis and allows the transmission of the lasers to do confocal imaging at the same time.

In order to control the exposure time and UV flash intensity, we have placed a fast shutter (Uniblitz) and a motorized filter wheel (Olympus) with neutral filters immediately after the mercury lamp and before the exit of the optical fiber (see Fig. 3). The motorized filter wheel, which can be commanded from the Olympus software, has six places where we have put five neutral density filters (New Focus) of optical density between 0.1 and 2, leaving one place empty to have maximum intensity. To control the flash duration, we have connected a home made pulse generator to the fast shutter driver unit (VCM-D1, Uniblitz) which sends 5 V transistor transistor logic (TTL) pulses of the desired duration starting at 10 ms with 10 ms steps. The Olympus microscope has a TTL output signal that is commanded with the Olympus software FW10-ASW (version 01.07.03) through the time controller window. This enables the user to program the acquisition routine (scanning mode, laser intensity, etc.), together with the delivery of the output signal. To synchronize the UV flash with the imaging scanning, the outgoing TTL signal is connected to the pulse generator. Employing the time controller routine it is possible to choose the time at which the TTL output signal is sent and the UV flash is delivered.

Fig. 3

Photograph of the external mercury lamp used for conventional fluorescence observation and uncaging. The mercury lamp house, fast shutter, and motorized neutral filter wheel, together with the fiber delivery system, are shown.

2.2.

Calcium Signaling Experimental Procedure

3.1.

Characterization of the UV Illumination

3.1.1.

UV spot size

One of the most common scanning aperture techniques to determine beam width is the scanning-slit profiler that measures the transmitted intensity as a thin slit cuts through the beam.¹⁵ As the intensity is integrated over the slit width, the resulting measurement is equivalent to the original cross section convolved with the profile of the slit. To estimate the spot size of the UV light in our case, instead of measuring the transmitted intensity, we have measured the reflected intensity as a thin strand passed through the beam. By scanning the strand through the beam the profile of the reflected intensity can be obtained as a function of position. Assuming a Gaussian field distribution, the beam waist can be estimated by fitting the profile with a function of the form [TeX:] \documentclass[12pt]{minimal}\begin{document}$ {\bm E}(x,y,0)= {\bm E}_{\bm 0}\,e^{-(x^{2}+y^{2})/{\omega _{0}^{2}}}$\end{document} $\mathit{E}(x,y,0)={\mathit{E}}_{\mathit{}\mathbf{0}}{e}^{-(x^2+y^2)/{\omega }_0^2}$ , where [TeX:] \documentclass[12pt]{minimal}\begin{document}$ {\bm E}_{\bm 0}$\end{document} ${\mathit{E}}_{\mathit{}\mathbf{0}}$ is a constant field vector in the transverse (x, y) plane, z = 0 at the beam waist and ω₀ denotes the beam waist radius.

In our setup, the strand was mounted onto a coverslip and was illuminated only by the UV spot. Employing a motorized stage (Scientific Roper) the strand was scanned with 10 μm steps, while the intensity was registered employing the confocal detectors (highly sensitive photomultipliers, Hamamatsu). Figure 4 shows the plot of the reflected intensity as a function of position together with the fitting curve. The experiments displayed here were done using a 60×, 1.35 N.A. oil-immersion objective (UPlanSAPO, Olympus) and a pinhole aperture of 105 μm. This lens produces a UV spot with a 212 ± 7 μm diameter.

Fig. 4

Normalized intensity profile as a function of the strand position. Squares: experimental data points. Solid line: Gaussian fit.

3.1.2.

UV power and attenuators

The mercury lamp power is 100 W, distributed over a continuous spectrum; it has an important peak at 365/366 nm which is useful for photolysis purposes. The optical components of the uncaging system and the microscope optics are expected to attenuate the light intensity. For this reason we have measured the power of the uncaging light at the output of the 60× objective after it has gone through different neutral filters. Measurements were performed with an Advantest TQ8210 power meter, at 364 nm showing that the maximum power was ∼400 μW. In Fig. 5 we show the plot of power as a function of transmission. Given that optical density (OD) and transmission (T) follow the relation T = 10^OD, it can be inferred from the linear dependence in Fig. 5 that the filters are attenuating equally all wavelengths in the 350 to 400 nm range, a positive feature for uncaging that had not been informed by the manufacturer.

Fig. 5

UV power as a function of the neutral filter transmission specified by the manufacturer. Solid curve: linear fit of the data forced to pass through zero.

3.2.

Calcium Caged Uncaging

As a first check of the uncaging capabilities of our setup we used NP-EGTA,^{16, 17} a photolabile chelator that exhibits a high selectivity for Ca^{2 +} and which dissociation constant increases from 80 nm to more than 1 mm upon UV illumination. Therefore, starting with a solution that mainly contains Ca^{2 +}-bound NP-EGTA, the application of a photolysis pulse induces a rapid delivery of Ca^{2 +}. Employing the calcium indicator Fluo 4 dextran it is possible to monitor this calcium release and verify the effectiveness of the uncaging system.

For this purpose we have used a solution containing [NP-EGTA] = 1.5 mm, Ca = 8.5 μm and the calcium indicator Fluo 4 dextran, [F4] = 0.1 mm. Single point scanning measurements at a 100 kHz sampling rate were performed employing a 488 nm line with a 60× objective. No neutral filter was used for the UV illumination. An increase in the fluorescence signal was observed after the UV flash delivery in response to the release of Ca^{2 +} from the caged NP-EGTA. We repeated the experiment using UV flashes of different durations that ranged between 100 and 400 ms. We computed the fluorescence amplitude (ΔF) as the difference between the maximun fluorescence obtained during the time course of each experiment and the basal fluorescence before the UV flash delivery.

We show the plot of maximum fluorescence amplitude as a function of UV flash duration in Fig. 6. In this figure it is possible to observe that the longer the duration of the UV flash, the greater the amount of Ca^{2 +} that is released. Furthermore, both quantities are linearly related as shown by the linear regression that is superimposed on the experimental data points. This is the expected behavior: the amount of uncaged compound is linearly proportional to the concentration of caged compound, flash intensity, and duration. This first set of experiments shows that the modification that we have introduced allows caged components to be photolyzed having full control of the time duration of the UV flash that is delivered producing results that follow the expected dependence with this duration.

Fig. 6

Maximum fluorescence amplitude (ΔF) as a function of UV flash duration. Solid line: linear regression forced to pass through the origin.

3.3.

InsP3 Evoked Calcium Signals

We illustrate in this section the capabilities of the modification that we describe in this paper to obtain confocal images of InsP3 evoked calcium signals. To this end experiments were performed in Xenopus Laevis oocytes that were previously microinjected with a mix of Fluo 4 dextran (high affinity) and caged InsP3 as explained in Sec. 2. Figure 7 shows confocal linescan images of representative signals obtained after the delivery of a 10 ms UV photolysis flash. In the case of Fig. 7a, EGTA was added to the mix to prevent Ca^{2 +}-wave propagation and increase the probability of observing localized signals such as puffs. Two different release sites can be identified in Fig. 7, denoted by the letters A and B. The time dependence of the intensity fluorescence profiles at these two sites are also displayed in Fig. 7. It is clear that both sites act independently of one another in this record with Ca^{2 +} release starting at disparate times. Site B, on the other hand, shows the occurrence of three consecutive Ca^{2 +} puffs. It is evident, particularly in the intensity profile plot, that the first event has a greater amplitude than the others. The time elapsed between events is ∼3.5 s, which is consistent with the interpuff time estimates presented in Ref. 18, 19. A more global signal is shown in Fig. 7b which corresponds to an experiment in which no EGTA was added to the mix. In this case, all release sites become active almost simultaneously and liberate Ca^{2 +} during a longer time. The intensity profile is also displayed in this case showing that the average Ca^{2 +} concentration decays with a time scale that is consistent with previous observations.²⁰

Fig. 7

Representative confocal linescan images showing Ca^{2 +} signals after a 10 ms UV flash employing Fluo 4 dextran and caged InsP3. In the panels, (F − F _o)/F _o is plotted using a color code, with distance along the scanned line in the vertical axis and time running from left to right in the horizontal axis. Increasing fluorescence ratios (increasing free Ca^{2 +}) are denoted by increasingly “warmer” colors (as indicated by the color bar to the right of the figure). The white vertical lines superimposed on the images indicate the UV flash delivery time. (a) Top: representative localized Ca^{2 +} signals (puffs). Two release sites are indicated by the letters A and B. In this case EGTA was also microinjected in the cell. Bottom: Temporal profile of the puffs that occurred at sites A and B. (b) More global signal (Ca^{2 +} wave) Bottom: Temporal profile of the wave.

In Fig. 8 a typical calcium wave is presented. As it can be observed, after one release site is activated the Ca^{2 +} that enters the cytosol diffuses and induces the opening of IP3R's in neighboring sites, generating the wave. It is known that the velocity of Ca^{2 +} waves varies with InsP3 concentration and ranges between ∼10 μm/s at low concentrations and ∼50 μm/s at larger ones.¹⁹ In order to estimate the wave velocity for the experiments of Fig. 8 we aligned a straight line with the apparent wavefront, as shown in the figure. The velocity obtained in this way is 18 ± 2 μm/s which is within the expected range of values.

Fig. 8

Confocal linescan image showing Ca^{2 +} waves after a 10 ms UV flash. White lines are aligned to the wavefronts in order to estimate the wave velocity from their slopes.

The behaviors that we observe in Figs. 7, 8 and the quantitative estimates that we can draw from them are in full agreement with previous observations of IP3R-mediated Ca^{2 +}-signals. There is no doubt then that the system is effectively photolyzing the caged InsP3 that was previously injected in the cells and is simultaneously acquiring confocal images of the signals evoked.