FRACTIONAL CALCIUM CURRENTS THROUGH RECOMBINANT GLUR CHANNELS OF THE NMDA, AMPA AND KAINATE RECEPTOR SUBTYPES

1. Simultaneous fluorescence and whole-cell current measurements using the calcium indicator dye fura-e were made in HEK 293 cells expressing recombinant glutamate receptor (GluR) channels, and fractional Ca2+ currents (the proportion of whole-cell current carried by Ca2+; P-f) were determined. 2. Cells expressing N-methyl-D-aspartate receptor (NMDAR) channels showed glutamate-activated Ca2+ inflow in the voltage range -60 to 40 mV in normal extracellular solution. Ca2+ inflow decreased in a voltage-dependent manner at membrane potentials more negative than -30 mV due to Mg2+ block. Voltage dependence of block at negative potentials was stronger in cells expressing the NR1-NR2A as compared with cells expressing NR1-NR2C subunits. 3. Fractional Ca2+ currents through NMDARs were independent of extracellular Mg2+ and varied between 8.2% (NR1-NR2C subunits) and 11% (NR1-NR2A subunits) in normal extracellular solution (1.8 mM Ca2+) at -60 mV membrane potential. P-f values increased with increasing [Ca2+](o) in the range of 0.5-10 mM [Ca2+](o) in a saturating fashion. 4. In cells expressing alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate receptor (AMPAR) subunits which were unedited at the Q/R site of the putative transmembrane segment TM2 (Q-form), or in cells coexpressing unedited and edited subunits (R-form), the glutamate-evoked Ca2+ inflow increased from 20 to -80 mV in an almost linear way. 5. Fractional Ca2+ currents through AMPAR channels depended on subunit composition. P-f values of Q-form homomeric channels at -60 mV and 1.8 mM [Ca2+](o) were between 3.2 and 3.9%. They were slightly voltage dependent and increased with [Ca2+](o) in the range 1.8-10 mM. P-f values in cells co-expressing Q- and R-form subunits were almost one order of magnitude smaller (0.54%). 6. Relative concentrations of Q-form and R-form GluR-B subunit-specific cDNAs used for cell transfection determined the expression of functionally different heteromeric AMPARs. P-f decreased with increasing relative concentration of R-form encoding cDNAs from 3.4 to 1.4%, demonstrating that editing of the Q/R site of GluR-B subunits decreases Ca2+ inflow through heteromeric AMPARs. 7. Cells expressing the GluR-B subunit of the kainate receptor (KAR) family were characterized by P-f values which depended on the editing in the TM1 and TM2 segments. P-f values were largest for the Q-form (1.55-2.0%) and lowest for R-form channels (< 0.2%), suggesting that Q/R site editing also decreases Ca2+ inflow through KAR channels. Cells co-expressing both subunit forms showed an intermediate value (0.58%). 8. Fractional Ca2+ currents measured with normal [Ca2+](o) were different from P-f values predicted with constant field assumptions from reversal potentials measured in bi-ionic (Ca2+-Cs+) conditions with high [Ca2+](o). 9. The results indicate that the size of P(f)s mediated by recombinant GluR channels in normal extracellular solution varies over a 50-fold range between less than 0.2 and 11%, depending on the combination of GluR channel subunits expressed. Assembly of different subunit combinations, relative abundance of subunit specific mRNAs and editing of mRNA are major mechanisms which control this wide range of Ca2+ inflow through different versions of GluR channels under physiological conditions.