Li+ Influx and Binding, and Li+/Mg2+ Competition in Bovine Chromaffin Cell Suspensions as Studied by 7Li NMR and Fluorescence Spectroscopy

Li+ influx by bovine chromaffin cells, obtained from bovine adrenal medulla, was studied in intact cell suspensions using 7Li NMR spectroscopy with the shift reagent [Tm(HDOTP)]4-. The influx rate constants, ki, were determined in the absence and in the presence of two Na+ membrane transport inhibitors. The values obtained indicate that both voltage sensitive Na+ channels and (Na+/K+)-ATPase play an important role in Li+ uptake by these cells. 7Li NMR T1 and T2 relaxation times for intracellular Li+ in bovine chromaffin cells provided a T1/T2 ratio of 305, showing that Li+ is highly, immobilized due to strong binding to intracellular structures. Using fluorescence spectroscopy and the Mg2+ fluorescent probe, furaptra, the free intracellular Mg2+ concentration in the bovine chromaffin cells incubated with 15 mM LiCl was found to increase by about mM after the intracellular Li+ concentration reached a steady state. Therefore, once inside the cell, Li+ is able to displace Mg2+ from its binding sites.


INTRODUCTION
The molecular and cellular mechanisms underlying the clinical use of lithium salts in the treatment of manic-depression (also called bipolar disease) as well as other psychiatric and non-psychiatric conditions are still poorly understood [1][2][3]. Several interrelated hypotheses have been formulated to clarify the action of Li+. Among these are: 1) Li + inhibition of the enzymes inositol-l-monophosphatase and adenylate cyclase, causing depletion in brain inositol and cyclic AMP (cAMP) levels [4][5][6][7]; and 2) competition between Li + and Mg 2+ ions for Mg 2+ binding sites in biomolecules, in particular guanine-nucleotide binding proteins which are involved in the signal transduction cascade [8][9]. It is possible that the pharmacological action of Li + cannot be explained by one single mode of action. In an attempt to contribute towards a better understanding of this problem, at the cellular and molecular levels, recent studies using fluorescence spectroscopy with the Mg 2+ 31 indicator furaptra 10] as well as Li and P NMR spectroscopy [11 have been undertaken. These techniques proved to be useful to investigate Li + transport [12], Li + binding [13] and Li+/Mg 2+ competition [11,14]. Li+/Mg 2+ competition was demonstrated for the phosphate groups of small phosphorylated molecules involved in second messenger systems, such as ATP/ADP, GTP/GDP and IP3 [11,15,16]; for the phosphate groups of erythrocyte membrane phospholipids [17]; in Li+-loaded human erythrocytes [10,18] and in human SH-SY5Y neuroblastoma cells 14]. Chromaffin cells from the adrenal gland medulla are excitable endocrine cells which can be used as good neuronal models [19,20] and whose neurotransmitter release has been found to be stimulated by Li + loading [21][22][23]. Therefore, in this study we used 7Li NMR and fluorescence spectroscopic methods to ,+ 2+ examine Li / uptake, intracellular Li / binding and competition between L and Mg ons in bovine chromaffin cells in suspension. The results obtained were compared with those previously obtained with other cell types. These studies aim to establish the generality of the ionic competition model described and may also contribute to the understanding of the pharmacological action of Li / at the molecular level.

Isolation of bovine chromaffin cells
Chromaffin cells were isolated from bovine adrenal medulla as described by Brocklehurst and Pollard [24] and were purified by centrifugation in a continuous Percoll gradient. The Neutral Red dye test [25] indicated that 65-80% of the cell preparation contained chromaffin cells. Cell viability was tested by the 357 Li-Influx and Binding, and Li-/Mg 2-Competition in Bovine Chromaffin Cell Suspensions as Studied by 7Li NMR and Fluorescence Spectroscopy Trypan Blue dye exclusion method [26]. The cells were maintained in a 1:1 mixture of Dulbecco's Modifed Eagle's Medium Ham's Nutrient Mixture F-12 (DMEM/F-12) (1.56%) containing 5% of heat-inactivated fetal calf serum, penicillin (10,000 units/ml), streptomycin (100 mg/ml), amphotericin B (25 tg/ml) and 10 laM of cytosine 13-D-arabinofuranoside, at 37C, in a humidified CO2 (5%) and air (95%) atmosphere. The cells were cultured up to a density of 106 cells/ml in 100 mm Petri dishes, 2-4 days before use for NMR and fluorescence experiments. For the fluorescence experiments, the cell preparation was further purified by differential platting in order to reduce contamination by endothelial cells.
The NMR experiments were performed with a Varian-Unity 500 NMR spectrometer equipped with a multinuclear 10 mm broad-band probe and a controlled temperature unit. 7Li NMR spectra were acquired at 194.3 MHz at 25C + 0.5C. The samples were spun at 16 Hz to prevent cell sedimentation. 7Li NMR spectra were obtained from 64 transients (total accumulation time of minutes), a spectral width of 5500 Hz, a pulse width of 15 lus, an interpulse delay of 10 s and an acquisition time of 0.362 s. The signal-to-noise ratio was enhanced by exponential multiplication with a line broadening of 30 Hz. 7Li T1 and T2 relaxation measurements for intracellular Li were conducted by use of inversionrecovery and Carl-Purcell-Meilboom-Gill (CPMG) pulse sequences, respectively, after the intracellular Li + concentration reached a steady state at the end of the influx experiments.
The kinetics of Li influx is defined by the equation A=-At A= e "kit (1) where At and A= are, respectively, the integrals of the intracellular Li + signal at time (t) and when the Li concentration inside the cells reaches the steady state. Thus, the apparent influx rate constants, ki, were determined using a graphical representation of time dependence of the percent of intracellular Li from plots of areas of intracellular Li + signal (Ai) over the total area (Ai + Ae) (Ae is the area of extracellular Li signal), [Ai/(Ai + Ae)] vs. time [12]. The integrals of the NMR signals were calculated using a signal deconvolution program.
Calculation of total intracellular [Li+]i from 7Li NMR spectroscopy 7Li NMR spectroscopy was used to determine the total intracellular Li + concentration where [Li+]i and [Li+]T are the total intracellular lithium concentrations within the cell and added to the sample, respectively, and Ai and Ae are the areas of the intracellular and extracellular NMR resonances taken at various times. CT is the cytocrit, i.e., the total percentage volume of cells in the sample used. The volume of cells can be calculated from a cell count and the known volume of a chromaffin cell. In this study, the sample volume was 1.5 ml with a cell concentration of 108. Since these cells are spherical in shape [27] and the diameter is 21.7 tm [27], the volume can be estimated to be-5400 tm .

Fluorescence Spectroscopy experiments
The fluorescence experiments were performed on a SPEX FluoroMax fluorimeter, at 30C, using the Mg 2+ fluorescent dye furaptra.
The free fluorescent probe furaptra has an excitation ,max at 370 nm and, as free Mg "+ concentration increases, an induced shift to lower frequencies is observed. The excitation ,,nax for the furaptra-Mg -+ complex is at 330 nm. The fluorescence intensities were measured at 335 nm and 370 nm and were monitored simultaneously over time with the emission wavelength fixed at 500 nm.
Similar studies were carried out using chromaffin cells permeabilized with digitonin (final concentration 0.2 mM) at 30C, in a Krebs medium without Ca + and Mg 2+ (in mM: NaCI 140, KCI 5, glucose 10, HEPES 20, EGTA 0.5, pH=7.35); increasing concentrations of Mg -+ and Ca -+ were added to investigate the behavior of furaptra in these conditions.
For the experiments with intact chromaffin cells, 3-4x107 cells were incubated in a Krebs medium (in mM: NaCI 140, KCI 5, CaClz 2, MgCI2 1, glucose 10, HEPES 20, pH 7.35) containing 1% BSA and 5 tM of the probe in the acetoxymethylester form (furaptra/AM) at 37C, in a humidified CO_ (5%) air (95%) atmosphere over 30-40 minutes to allow probe loading. Before addition to the cells, Pluronic (R) F-127 (0.1%) was added to the previous medium and sonicated over 3 minutes. This detergent was used to help dispersion of the fluorescent indicator in the loading medium [28]. After probe loading, the cells were diluted with Krebs medium with 1% BSA and incubated at 37C for 30 minutes to allow ester hydrolysis. The cells were then washed in a Krebs medium with 0.1% BSA and finally ressuspended in Krebs medium without BSA.

RESULTS AND DISCUSSION
Li + transport across the chromaffin cell membrane The influx of Li + into chromaffin cells was studied using 7Li NMR spectroscopy. The paramagnetic shift reagent [Tm(HDOTP)] 4 [12] was used to separate 7Li NMR signals corresponding to Li + inside and outside the cells. Due to its negative charge, this compound is impermeable to cell membranes and shifts the resonance of the extracelullar Li + to higher frequencies relative to the resonance of the intracellular Li+.   33 The results with TTX indicate that voltage-sensitive Na + channels are involved in the uptake of Li + by these cells, confirming the results already obtained with other types of cells [12,30,31]. In the presence of ouabain, an inhibitor of (N + + a ,K )-ATPase, an inhibition of 33% in Li + influx was observed. In human + + these cells erythrocytes, Li + is not transported by (Na ,K )-ATPase under physiological conditions [30], in i+ Li + influx is only detected through this pump in the absence of extracellular Na+. However, L influx is detected through (Na+,K+)-ATPase under physiological conditions in some other cell types such as neuroblastoma x glioma hybrid cells [32]. Therefore, the results obtained in this work could be due to a direct or an indirect effect on (Na+,K+)-ATPase. An indirect effect of this inhibitor could occur by changing the Na equilibrium or the membrane potential, and therefore indirectly causing a change in Li + influx. Thus, it should be further investigated if (Na+,K+)-ATPase has a direct or an indirect role in the Li + transport in this cell type.
To qualitatively compare the amount of Li + accumulated by this cell type as compared to other cells, we used previously reported studies on human erythrocytes and human neuroblastoma cells. For a meaningful comparison, the cytocrit, the extracellular Li + concentration and the periods of incubation should be approximately equal. In this study, bovine chromaffin cells at 25 % cytocrit after 80 min accumulate an 360 Metal Based Drugs Vol. 7, Nr. 6,2000 intracellular Li concentration of 1.67 + 0.43 mM with an extracellular Li + concentration of 15 mM LiCI as obtained by equation (2). With human neuroblastoma cells (10-15 % cytocrit), an intracellular Li concentration of 3.1 + 0.1 mM was obtained with an extracellular Li + concentration of 15 mM at 90 min .+ [33]. Unfortunately, no studies with an extracellular L concentration of 15 mM and a similar hematocrit have been performed on human erythrocytes. However, in one study, an extracellular Li concentration of 150 mM resulted in an intracellular Li + concentration of 3 mM after 75 min (10% hematocrit) [13]. Even though the experimental conditions were different in our study as compared to the other two [13,33], it is apparent that human erythrocytes accumulate the least amount of Li+; human neuroblastoma cells and bovine chromaffin cells are more similar in the amount of Li + accumulated.

Intracellular Li / binding
At the end of the Li + influx experiments, when the Li + concentration inside the cell reached the equilibrium, 7Li NMR Ti and T2 relaxation times for the Li+i resonance were determined using the inversionrecovery and CPMG pulse sequences, respectively. Because 7Li NMR T/T2 ratio is a measure of the relative degree of immobilization of the Li + ion, the higher the value of this ratio is, the lower the mobility of Li [13]. Therefore, 7Li NMR relaxation measurements give information about the physical state of the ion inside the cell. A comparison of the 7Li NMR T/T2 ratios obtained in this work with those for other types of cells and some of their components is shown in Table 11. From these values which were obtained with similar intracellular Li concentrations [12,13,33], it is possible to conclude that Li is more immobilized within chromaffin cells that in other cellular systems already studied [12,13], which is in agreement with the higher degree of complexity of these cells. Li + is strongly bound to the membrane, intracellular structures in the cytosol and probably inside the chromaffin granules. The high 7Li NMR T/T2 ratio of the plasma membrane of the human neuroblastoma cells and human erythrocytes as compared to LiCI free in solution ( Table !I) indicates that these are major Li + binding sites. A more detailed study with various chromaffin cell components, in particular plasma membrane suspensions, could be useful to better define which are the most probable Li + binding sites in these cells.  [13] T (s) (+_ SD) 6 from the cells with time, and its binding to divalent cations present in the extracellular medium, a control experiment was made in the same conditions but with the cells suspended in a Krebs medium without Li (data not shown). The fluorescence changes observed during this control experiment were taken into account and subtracted to the Li+-Ioaded cells data. The fluorescence data was converted into intracellular free Mg concentrations through equation (3), using a value of 1.9 mM for K (furaptra-Mg2+) [14], and the result is represented in Fig. 2B were calculated using eq. 3 and a value of 1.9 mM for Kd (furaptra-Mg+) [14].
As shown in Fig. 2B, there is an increase in the intracellular free Mg2. + concentration with time, from its basal value of 1.1 _+ 0.20 mM (n=3), determined in the absence of Li+, to 2.1 _+ 0.14 mM (n=2), after the incubation of the cells in a 15 mM LiCI modified Krebs medium for 60 minutes.
As the intracellular space of the Li+-loaded cells contains Mg2+, Ca 2+ and Li+, their binding to the dye has to be taken into account. In vitro studies of the excitation fluorescence spectra of furaptra, in its salt form, in the presence of different concentrations of Mg 2+ (in the absence of Ca2+), and of different concentrations of Ca 2+ (in the absence of Mg2+) have been described in the literature [29,34]. The behavior of the probe in the presence of Ca z+ or Mg z+ is very similar, with an excitation max for the furaptra-Ca 2+ complex also at 330 nm and the dissociation constants Kd for furaptra-Ca 2+ and furaptra-Mg 2+ complexes are 53 IaM [29,34] and 1.5 mM [29] (or 1.9 mM [14]), respectively. Fluorescence excitation spectra obtained with 2 10 6 chromaffin cells previously incubated with furaptra/AM, suspended in a free-Ca "+ and free-Mg 2+ Krebs medium and then permeabilized with digitonin, were found to be very similar to those obtained in solution, 2+ when increasing amounts of Ca 2+ and Mg were added (data not shown). Thus, the fluorescent probe has the same photophysical properties inside the cells and free in solution.

2+
Ca has a higher affinity to furaptra than Mg + according to the respective Kd values [29,34]. However, it was observed that the intracellular free Ca 2+ levels.during Li+-loading experiments with chromaffin cells using the Ca 2+ sensitive fluorescent probe fura-2, increase slowly from its basal value (100-200 nM) up to around 600 nM (data not shown). Once the induced shift of furaptra to lower frequencies occurs when Ca z+ concentration is higher than IM [29], this ion do not interfere with the spectroscopic properties of this probe.
Since Li + has also been shown to bind furaptra (Kd 250 mM at 25 C and 237 mM at 37 C [10]) it was necessary to determine if the contribution of Li + binding to the dye was significant and if a correction in eq. 3 was necessary [11]. In this study it was already mentioned that the intracellular Li concentration was estimated to be 1.67 + 0.43 mM. This amount of Li + would only affect furaptra fluorescent properties by 0.7 % (1.67 mM/250 mM), which can not account for the changes shown in Fig. 2B.
These data show that the increase in the ratio of fluorescence intensities R(335/370) of furaptra observed when Li goes into the cells is due to an increase in the intracellular free Mg 2+ concentration, without interference of cytosolic Ca 2+ or Li + in the fluorescent properties of this probe, confirming the ability of Li + to displace Mg + from its binding sites within the cells.

CONCLUSIONS
The present study investigates Li + influx pathways in bovine chromaffin cell suspensions by 7Li NMR methods. The data indicate that voltage-sensitive Na + channels, which have been shown to be active in Li influx in human neuroblastoma SH-SY5Y cells [12] and other cells [30,31], play an important role in Li uptake by this cellular system. Also, (Na+,K+)-ATPase affects Li influx rates when inhibited by ouabain.
However, the exact role of this pathway for Li + should be further explored with this cell type. Comparing the total accumulation of Li + under similar conditions, it is apparent that human erythrocytes accumulate much less Li + than human neuroblastoma and chromaffin cells. 7Li NMR T/Tz ratios for intracellular Li in chromaffin cells show that Li + is highly immobilized due to strong binding to intracellular structures as compared to other cell types. A more detailed study of the Li binding sites will help elucidate the location of major Li + binding sites in bovine chromaffin cells. Once inside the cell, Li is able to displace Mg 2+ from its binding sites. Thus, in this work, we confirm that Li+/Mg 2+ competition also occurs in bovine chromaffin cells, in agreement with previous results published in the literature with other cellular systems [14,18], which is one of the hypotheses to explain the therapeutic action of Li + at the molecular and cellular levels.