Chemicals and solutions
The fluorescent Ca2+ dyes fluo-3, fluo-4 and fluo-5F as pentapotassium salts or acetoxymethyl (AM) esters, as well as the calcium standards used to measure the dissociation constants of these dyes were purchased from Molecular Probes (Eugene, OR, USA). All other chemicals were obtained from Sigma Chemical Co. (St Louis, MO, USA). The dissection and incubation with AM ester dye was carried out in Locke solution (including mm): 140 NaCl, 3.6 KCl, 2.4 MgCl2, 1.2 CaCl2, 3 Hepes, 10 glucose, 5 Na ascorbate and 20 μm EDTA, pH 7.4). During fluorescence and suction pipette measurements, photoreceptors were continuously superfused with oxygenated Dulbecco's modified Eagle's medium (DMEM) without bicarbonate or phenol red (Sigma Chemical Co., D-2902), supplemented with (mm): 20 Hepes as well as 5 succinic acid, 0.5 glutamic acid and 5 gluconic acid (as Na+ salts). The pH was adjusted to pH 7.4, and temperature was maintained near 37 °C as described previously (Matthews, 1991). The suction pipette and bath ground electrode were filled with Locke solution without glucose or ascorbate.
Dissection and dye loading
All procedures were approved by the Chancellor's Animal Research Committee (ARC no. 93-230).
Mice previously maintained in total darkness for 3 h were killed by cervical dislocation. The eyes were removed from the head and trimmed of fat and extra-ocular muscle, and they were then washed with 1–2 ml of cold Locke solution. Killing, eye removal, trimming and washing were done in dim red light. The rest of the dissection was done under infrared illumination with an IR-sensitive TV camera (CCD camera model OS-40D, Mintron USA, Fremont, CA, USA) or an IR-sensitive viewer (FJW Optical Systems, Inc., Palatine, IL, USA).
Each eye was hemisected with a razor blade. The front of the eye (with lens) was discarded, and the eyecup (with retina) was cut in two by slicing through or close to the optic nerve head. Three pieces of eye were reserved on ice in total darkness, and one was placed in Locke solution in a 35 mm petri dish whose bottom had been coated with Sylgard (Dow Corning, Midland, MI, USA). From this piece the retina was isolated and finely chopped with a piece of razor blade. The resulting suspension was incubated for 30 min at room temperature in Locke solution containing fluo-3 AM, fluo-4 AM or fluo-5F AM (see Results). After completion of the incubation period, the cells were superfused at 37 °C for 20 min with DMEM, supplemented as described above, to wash out excess dye and to allow completion of the intracellular hydrolysis of the AM ester.
Methods for measuring calcium have been described previously (Sampath et al. 1998a, 1999; Matthews & Fain, 2000, 2001). Calcium dye fluorescence was excited by an intense 10 μm spot of light imaged on the rod outer segment through a 40 × oil-immersion fluorescence objective (1.3 NA). Excitation light was provided by an argon laser (American Laser Corporation, Salt Lake City, UT, USA), which for fluo-3 was tuned to 514 nm, and for fluo-4 and fluo-5F, to 488 nm. For fluo-3 we used a 525 nm dichroic filter and a 530 nm long pass filter, but for fluo-4 and fluo-5F, a 505 nm dichroic and a 510 nm long pass filter (Omega Optical, Brattleboro, VT, USA). Ca2+ dye fluorescence was recorded from the outer segments of single mouse rods either fully isolated or protruding from small chunks of retina; the saturated electrical response to the intense light of the laser was normally not recorded.
In control experiments, fluorescence was measured for the pentapotassium salt of each of the dyes immobilised in agarose gel (see Fig. 3A). To prevent dye-bleaching, the unattenuated intensity of the laser was normally reduced to around 5–6 × 1010 photons μm−2 s−1 by reflective neutral density filters (Newport Corporation, Irvine, CA, USA). Fluorescence was measured with a low-dark-count photomultiplier tube with a restricted photocathode (Model 9130/100A, Electron Tubes Ltd, Ruislip, UK), and the current was amplified by a low-noise current-to-voltage converter (PDA-700, Terahertz Technology, Oriskany, NY, USA), low-pass filtered with an eight-pole Bessel filter (Frequency Devices, Haverhill, MA, USA) and digitised (PCLAMP, Axon Instruments, Foster City, CA, USA) with a PC-compatible computer over the bandwidths and at the sampling rates specified in individual figure legends.
Figure 3. Bright light evokes a rapid decline in fluo-4 fluorescence in wild-type mouse rods without dye bleaching
A, photomultiplier current elicited by fluorescence from a film of fluo-4 pentapotasssium salt immobilised in a high-Ca2+ agarose gel, to a 30 s exposure at a laser intensity comparable to that used for measurements from mouse rods. Trace is the average of four records obtained from different areas of the dye film. B, fluorescence response from a dark-adapted mouse rod loaded with fluo-4 evoked by the first exposure to laser light. C, the same data on an expanded time base illustrating the initial decline in fluorescence, together with a superimposed trace evoked by a second laser exposure 60 s later. The decay in fluorescence was fitted with a double exponential decay (eqn 2) with time constants for this cell of τ1/ 159 ms and τ2= 508 ms. Inset, mean change in circulating current as a function of time for 12 rods recorded with a suction pipette and exposed to the same laser intensity used for the calcium measurements. For clarity, only the response to the first laser exposure is shown; there was no response to the second, indicating that the circulating current remains suppressed between the two laser presentations. The arrow indicates Na+/Ca2+-K+ exchange current, which could be fitted with two constants of about 40 and 690 ms (continuous curve).
Download figure to PowerPoint
Measurement of dye dissociation constant
Dissociation constants for the pentapotassium salts of fluo-3, fluo-4 and fluo-5F were measured as described by Tsien & Pozzan (1989). Two sets of standards were used, one without Mg2+ and one containing 1 mm Mg2+. The standards also both contained 100 mm KCl and 10 mm Mops buffer (pH 7.2). The free Ca2+ concentration in these solutions was estimated by the Max Chelator program (version 4.62, Chris Patton, Hopkins Marine Station, Stanford University, CA, USA) following well-established procedures (Martell & Smith, 1974; Fabiato & Fabiato, 1979; Harrison & Bers, 1989), to correct estimates of free Ca2+ for temperature, Mg2+ concentration and ionic strength. The concentration of Ca2+ in the standards was confirmed by Ca2+ electrode measurements. The final fluorescent dye concentration in all calibration standards was 10 μm. All dye calibration measurements were carried out with the same microscope, laser wavelength, filters and spot diameter as for measurements from photoreceptors.
To determine the temperature dependence of the dissociation constant, a special chamber was constructed. A triple layer of Parafilm (American National Can, Neenah, WI, USA) was pressed onto both sides of a rectangular glass coverslip (45 mm × 50 mm × 150 μm thick). A 10 mm hole was cut in the parafilm on the underside of the coverslip, and a second, smaller coverslip was placed over this hole, forming a chamber that contained the dye solution, held in place by surface tension. Slits were cut in the parafilm to either side to permit the passage of two small-diameter (0.051 mm) thermocouple wires (T-type, Cu-CuNi; Omega Engineering, Stamford, CT, USA). The thermocouple junction was positioned in the middle of the chamber, and melted paraffin wax held the wires in place and prevented leakage of solution. The temperature of the dye solution was monitored with a digital thermometer at 0.1 °C resolution (Model HH-25TC, Omega Engineering, Stamford, CT, USA).
The other side of the coverslip facing upward on the microscope stage contained a larger, 14 mm diameter well for the perfusion of distilled water to control the temperature of the dye chamber. Two streams of water fed into the upper well, one from an ice bath and the other from a heated water bath also passing through a heated aluminium block. Solution inflow from these two inlets was controlled by solenoid values (Lee Company, Westbrook, NJ, USA). The path length of the dye-containing solution in the lower well was approximately 0.4 mm, much larger than the vertical dimension of the laser spot focused at the image plane of the microscope.
Fluorescence was recorded via a 40 × dry objective (0.65 NA) while moving the plane of focus smoothly up and down through the sample by rotating the fine focusing control of the microscope with a small DC motor (Motor Mike, Oriel, Stratford, CT, USA). The fluorescence was initially low but increased to a plateau as the focal plane of the laser entered the dye solution (Fig. 1A, inset). The motor was then reversed and the spot moved back through the solution. This procedure was repeated to measure fluorescence from the same sample at four temperatures: 20, 22, 30 and 37 °C. All measurements were taken within 1 mm of the thermocouple junction to ensure that the temperature at the laser spot would be close to the measured temperature. At the higher free Ca2+ concentrations (from 500 nm to 40 μm), when the fluorescence signal was relatively high, a measurable change in fluorescence could be detected for a change of temperature of only 0.1 °C. For each dye at each temperature, three independent determinations were made, each with freshly made standard solutions.
Figure 1. Temperature changes within the physiological range alter the Ca2+ affinity of the fluorescent indicator dyes fluo-3, fluo-4 and fluo-5F
A, effect of temperature on the titration curve for Ca2+ binding to fluo-4 measured at 20 (×), 22 (□), 30 (▵) and 37 °C (□). Ordinate plots photomultiplier current evoked by dye fluorescence as a function of Ca2+ concentration (see Methods). Ca2+ concentrations have been corrected for effects of temperature on Ca2+-EGTA binding, leading to a slight displacement of the values at each temperature for individual solutions. Continuous curves determined according to a sigmoidal logistic model (eqn 1) fitted to the data at each temperature by a least-squares algorithm. Inset, superimposed photomultiplier current traces during fluorescence measurements with fluo-5F and nominally 60 nm Ca2+ at 20, 30 and 37 °C as the plane of focus was moved upward (first maximum) and then downward (second maximum) through the liquid dye film (see Methods). B, plot of dissociation constant against temperature determined for each of the three dyes as in A in the presence and absence of 1 mm Mg2+. Each point represents the mean of three independent determinations; error bars are s.e.m. Thermodynamic parameters calculated from the regression lines fitted to these data are collected in Table 1.
Download figure to PowerPoint
Increasing temperature shifted the dissociation constant (Kd) of the dye to a lower Ca2+ concentration and decreased the maximum fluorescence (Fmax see Fig. 1A). Little effect of temperature was observed on minimum fluorescence (Fmin). Data at each temperature were fitted with a sigmoidal logistic model (eqn 1):
The apparent cooperativity (n) of Ca2+ binding was not significantly different from unity for any of the three dyes used. Values for both Kd and Fmax are summarised in Table 1. The addition of Mg2+ caused an increase in the Kd of about 100–200 nm for all of the dyes at all temperatures.
Table 1. In vitro dissociation constants, fluorescence ratios and thermodynamic constants for fluo indicator dyes
| ||–||22||1000||19||7.6||2260||310|| ||–8100||74|
| ||–||30||540||6||7.2||1860||260|| ||–8690||74|
| ||–||37||310||11||6.9||1570||230|| ||–9230||74|
| ||–||22||520||14||14.8||1530||100|| ||–8480||71|
| ||–||30||310||9||13.8||1430||100|| ||–9030||71|
| ||–||37||190||9||13.1||1300||100|| ||–9540||71|
| ||–||22||590||23||1.9||400||260|| ||–8410||72|
| ||–||30||330||24||1.8||350||230|| ||–8990||72|
| ||–||37||200||12||1.8||320||200|| ||–9500||72|
| ||+||22||1280||117||9.3||2230||240|| ||–7960||74|
| ||+||30||700||55||9.6||1850||190|| ||–8540||74|
| ||+||37||400||25||10.4||1560||150|| ||9080||74|
| ||+||22||670||18||14.3||1510||110|| ||–8340||72|
| ||+||30||370||10||14.3||1460||100|| ||–8920||72|
| ||+||37||230||12||14.4||1350||100|| ||–9410||72|
| ||+||22||723||57||1.9||437||272|| ||–8290||69|
| ||+||30||420||50||2.1||387||225|| ||–8840||69|
| ||+||37||266||30||2.2||344||176|| ||–9330||69|
In Fig. 1B, the logarithm of the Kd is plotted against temperature for each dye, with and without Mg2+. The slopes of the regression lines through the data are virtually the same for all three dyes and were unaffected by Mg2+. The average enthalpy (ΔH) of Ca2+ binding was determined from a van't Hoff analysis (log Kavs. 1/T) and is given in Table 1. The enthalpy was about +12 to +13 kcal mol−1 (50 to 54 kJ mol−1) for all of the three fluo dyes tested. Table 1 also gives calculated values for the changes in free energy (ΔG) and entropy (ΔS).
Determination of free Ca2+ concentration
We estimated the absolute value of [Ca2+]i in darkness and after exposure to bright light in the same way as previously for salamander rods (Sampath et al. 1998a). The minimum fluorescence Fmin was determined by exposing the rod to a zero Ca2+/ionomycin solution (mm): 140 NaCl, 3.6 KCl, 3.08 MgCl2, 2.0 EGTA, 3.0 Hepes, with 10 μm ionomycin, pH 7.4). When the fluorescence reached a steady minimum value (Fmin), the rod was exposed to a high Ca2+/ionomycin solution (mm): 96.9 CaCl2, 3.6 KCl, 3.0 Hepes with 10 μm ionomycin, pH 7.4) to estimate Fmax. The calcium concentration was then calculated from the previously determined value of dye Kd and the Michaelis-Menten equation (Grynkiewicz et al. 1985; Sampath et al. 1998a).
Light stimulation and electrical recording
In experiments to determine the effects of dye loading on the electrical responses of wild-type rods or to investigate the degree of stimulation by the laser, the photocurrent was recorded using the suction pipette technique. Suction pipette currents were recorded with a patch-clamp amplifier (Model PC501-A, Warner Instruments Co, Hamden, CT, USA), low-pass filtered at 35 Hz and sampled at 100 Hz with the same apparatus used for recording the dye signals. In these experiments dye fluorescence was typically not recorded, since the outer segment was normally drawn into the suction pipette precluding spatially precise fluorescence measurements.
Light stimuli were delivered from a dual-beam optical bench via a 500 nm interference filter; stimulus intensity was controlled with absorptive neutral density filters and calibrated with a calibrated silicon photodiode (Graseby Optronics, Orlando, FL, USA). In some experiments, instead of 500 nm light rods were stimulated with white light, whose effective intensity was estimated by comparing the relative sensitivity of the electrical response to dim flashes of white and 500 nm light. Rhodopsin bleaching was estimated from the photosensitivity for a vitamin-A2-based pigment in solution (Dartnall, 1972), corrected for the difference in dichroism in free solution and in disk membranes (Jones et al. 1993).