Measurement of cytosolic calcium using 19F NMR.

Fluorine-19 nuclear magnetic resonance (NMR) studies of cells and perfused organs loaded with fluorinated ion chelators represent a new approach to determining cytosolic free calcium levels in situ. The molecular basis for this approach and the relative advantages and disadvantages of the NMR technique are discussed in this paper. Results obtained on perfused normoxic and ischemic rat hearts are presented, indicating that ischemia is associated with an elevation in the level of cytosolic free calcium before the onset of irreversible cell injury. The results are therefore consistent with this elevation playing a causative role in the mediation of myocardial cell injury resulting from ischemia.


Introduction
As discussed elsewhere (1,2) cytosolic free calcium [Ca2]i 'is an important regulator of cell physiology, and perturhaations of [Ca2+]i may be important in mediating cell injury resulting from chemical or physical agents. Accurate methods for the measurement of [Ca2+], and other ions are central to elucidating the role of ions in these processes. Many of the methods used for making such measurements involve the introduction of chelators into the cytosol so that changes in [Ca2+]i can be followed by the corresponding spectral changes that accompany calcium-chelator complexation. Several types of [Ca2+]i indicators including fluorescent chelators (3), photoproteins (4), and metallochromic dyes (5)

Materials and Methods
Adult, male Sprague-Dawley rats weighing 200 to 300 g were anesthetized with pentobarbital. The heart was excised and the aorta was cannulated within 15 sec. Retrograde perfusion was started from a reservoir 90 cm above the aortic cannula. The perfusate was Krebs-Henseleit buffer containing 120 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO, 1.2 mM KH2PO4, 1.25 mM CaCl2, 25 mM NaHCO3, and 10 mM glucose. The buffer was continuously aerated with humidified 95%02 to 5% CO2 and was maintained at 370C.
For assessment of contractile function, a rubber ballon on the tip of a polyethylene catheter was inserted through the left atrium into the left ventricle. The catheter was connected to a Statham P23d pressure transducer that was outside the magnet located at the same height as the heart. The balloon was inflated to give an end-diastolic pressure of 5 to 15 cm of water.
After 15 min of control, nonrecirculating perfusion, the heart was perfused with 150 to 400 mL of a Krebs-Henseleit solution containing 5 iM of the acetoxymethyl ester of the appropriate 19F NMR indicator. After the indicator loading was completed, the heart was placed in a standard 20-mm NMR tube with the apex of the heart approximately 1 cm from the bottom of the tube. The perfusate was evacuated by PE tubing connected to a variable speed Masterflex peristaltic pump that extended to the bottom of the tube.
NMR studies were performed on a Nicolet NT 360 NMR spectrometer at 37°C, using either a 20 mm 19F probe tuned to 339.7 MHz (Doty Scientific, Columbia, SC) or a 20-mm broad-band Nicolet probe with decoupler tuned to 339.7 MHz for the 19F studies and the observe coil tuned to 31P. The sample was shimmed on H20, and we routinely obtained a (nonspinning) line width at half height of approximately 0.25 ppm. For the '9F studies we used a 40°pulse angle, a 500 us delay, and 127 ms acquisition time. A 70°pulse angle, a 1-sec delay, and a 205-msc acquisition time were used for the 31P studies.

Results and Discussion
The calcium chelator, 5FBAPTA is introduced into cells and perfused organs as the acetoxymethylester (AME), a form that is readily permeable across the plasma membrane (9). Once inside the cell, naturally occurring esterases cleave the AME, leaving the negatively charged 5FBAPTA trapped in the cytosol. The 5FBAPTA undergoes an NMR chemical shift upon binding calcium, the basis for which is apparent from the X-ray crystal structure (10). The chelator 5FBAPTA forms an octadentate complex with calcium ions, as illustrated in Figure 1. This complex involves the coordination of one of the carboxyl oxygen atoms from each of the four carboxyl groups, the two ether oxygen atoms, and the two anilinic nitrogen atoms. The formation of a nearly tetrahedral, sp3 bonding structure for the nitrogens alters the conjugation with the aromatic system and is responsible for the large perturbation of the 19F chemical shift that accompanies complexation. As discussed by Tsien (3), analogous structural changes accompanying calcium ion complexation with the fluorescent indicators can provide the basis for the observed spectral changes as well.
The 19F NMR spectrum of 5FBAPTA in the presence of a nonsaturating level of calcium exhibits two resonances corresponding to the uncomplexed and to the calcium complexed species (Fig. 2). The simultaneous observation of both resonances reflects the fact that the rate of dissociation of the calcium-chelator complex falls into the slow exchange limit in which the lifetime of the Ca-5FBAPTA complex is longer than the reciprocal of the chemical shift difference (GK X > 1/6 ppm = 4.9 x i0-' sec). As illustrated in The reported values for the dissociation constant of Ca-FBAPTA have ranged from 285 nM measured at 30°C, pH 7.2 (no Mg) (11); to 337 nM at 37°C, pH 7.2 (no Mg) (11); to 451 nM at 300C, pH 7.1 (0.5 mM Mg) (12); to 635 nM at 370C, pH 7.05 (no Mg) (8); to 708 nM at 37°C, pH 7.1 (no Mg) (6). These measurements illustrate the trend of higher Kd values associated with increased temperature. However, the discrepancies are currently somewhat larger than expected and further work is necessary to arrive at a consensus value.
The recently developed a fluorinated magnesium ion indicator (13). Just as the EGTA molecule forms a useful starting point for the develpment of calcium-sensitive indicators (3), the development of magnesium-sensitive indicators has proceeded by modifying the structure of EDTA. Specifically, the replacement of the central ethylene moiety by a fluorinated benzene, and the substitution of one of the N,N diacetic acid groups with an oxoacetic acid, results in a fluorinated 0aminophenol-N,N,O-triacetate (APTRA) structure, which we have found useful for carrying out intracellular magnesium ion measurements. These molecules exhibit magnesium Kd values that are closely matched to the anticipated cytosolic magnesium concentrations. Although the calcium Kd values are somewhat lower, they are still 100to 1000-fold greater than typical cytosolic calcium values. With the possible exception of extreme pathological states, the presence of calcium will not interfere with the magnesium determination. For two of the fluorinated derivatives studied, 5-fluoro APTRA and 4-methyl, 5fluoro APTRA, the kinetics of the magnesium chelator interaction again fall into the slow exchange limit at the field strength used in this study. Therefore, two resonances can be observed, and the level of magnesium determined is analogous to the calcium determination illustrated in Figure 2. Finally, it is worth noting that the APTRA structure readily lends itself to furthur modification to include fluorophores analogous to those utilized for intracellular calcium ion indicators (14). Cytosolic free magnesium levels measured with 5FAPTRA in perfused rat heart average 0.85 mM. The relative advantages and disadvantages of the use of NMR-sensitive indicators are summarized in Table 1. Briefly, the absence of fluorine-containing metabolites in the cell makes it possible to carry out