Factors affecting guanine nucleotide binding to rat AMPA receptors

Abstract

Glutamate receptors are competitively inhibited by guanine nucleotides. Insight into the physiological function of this inhibition would be greatly advanced if nucleotide binding could be eliminated through mutations without altering other aspects of receptor function, or if compounds were discovered that selectively prevent nucleotide binding. It was previously reported that a lysine in the chick kainate binding protein (cKBP) is specifically involved in guanine nucleotide binding. In the present study we mutated the equivalent lysine in the rat AMPA receptor subunit GluR1 flip to alanine (K445A) and assessed changes in nucleotide affinity from the displacement of [(3)H]fluorowillardiine. As in the cKBP, the affinity for nucleotides was greatly reduced while the binding affinity for agonists remained unchanged. The reduction in affinity was largest for GTP (factor of 5.8) and GDP (4.4) and minor for GMP and guanosine. This suggests that K445 is involved in stabilizing the second phosphate of the nucleotide. Given that bulkier analogs like GDP-fucose are also accommodated at this site, it seems likely that nucleotides bind in such a way that their phosphates project out of the cleft. In excised-patch recordings using short pulses of glutamate, the K445A mutation increased the EC(50) for the peak response 1.8-fold and accelerated desensitization and deactivation. This indicates that the effects of this mutation are not as specific as previously suggested. Efforts to selectively eliminate inhibition by nucleotides may therefore depend on mapping out further the docking site. In a first attempt using point mutations we ruled out several amino acids around the cleft as being involved in nucleotide binding. Also, the AMPA receptor modulator PPNDS which competitively inhibits nucleotide binding to purinergic receptors did not affect nucleotide inhibition, suggesting that there are major differences in the topography between purinergic and glutamate receptors. Thus new approaches, including crystallography, may be called for to identify residues uniquely involved in nucleotide binding.

Figure 1. Agonist binding and guanine nucleotide inhibition in wildtype GluR1 flip versus GluR1i-K445A

A. Binding of [³H]FW was measured at 0°C at radioligand concentrations of 1–100 nM. Non-specific binding was determined by inclusion of 10 mM glutamate and subtracted from total binding. Specific binding was transformed into the Scatchard format; linear regression provided the K_D values shown as insets. Representative experiment; for summary data see table 1. B. Inhibition of [³H]FW binding (5 nM, 0°C) to wildtype (wt) R1i receptors by guanine nucleotides and guanosine. Binding was expressed as percent of that in the absence of nucleotide and averaged across experiments. The averaged data (mean and sem) were then fitted through non-linear regression with a sigmoidal inhibition curve (n_Hill = 1; bottom asymptote fixed at 0). C. Comparison of nucleotide inhibition in wildtype R1i and R1i-K445A (mut). The data for the wildtype receptor are the same as in B. The inhibition curves were fitted to the averaged data. The insets show the averaged inhibition constants K_i as taken from table 1.

Figure 2. Physiological properties of GluR1-K445A

Membrane patches were excised from HEK293 cells transfected with GluR1-flip wildtype (wt) or GluR1-K445A (K445A). Typically, 5–10 consecutive traces were averaged. The holding potential was −70 mV. A. Concentration response relations for glutamate. Pulses of various concentrations of glutamate were applied for 100 ms, and peak amplitude at each concentration were normalized to that at 10 mM glutamate. The data points represent values obtained in individual experiments (n=5 for wildtype, n=6 for K445A). The EC50 value was calculated by fitting the data points with a sigmoidal function. The insets show representative traces (0.3–10 mM). The horizontal bar indicates application of glutamate. B. Changes in kinetic parameters. Desensitization and deactivation kinetics were examined in currents induced by 100 ms and 1 ms pulses of glutamate (10 mM), respectively. The time constants were determined by fitting the decay phase of the response with a single exponential function. %SS/Peak denotes the percentage of steady-state current relative to the peak current. Only responses with a peak current higher than 100 pA were analyzed. The bars show the mean and s.e.m. of 30 (K445A) and 21–22 (wild type) experiments; current density for R1i-K445A was determined in 40 experiments. *, p < 0.05; **, p < 0.01; ***, p < 0.001. C. Concentration-response relations for GDP. The nucleotide was included at the indicated concentrations in both the background flow line and in the glutamate pulse (10 mM). Left: representative traces; horizontal bars indicate glutamate application. Right: Summary data for the effect of GDP on peak amplitude. The data points represent the mean and sem from 7 (K445A) and 8 (wildtype) experiments. The data were fitted with a sigmoidal function; IC50 values are shown in the insets.

Figure 3. Lack of interactions between PPNDS and guanine nucleotides

Left. Enhancement of [³H]AMPA binding to wildtype GluR1i and GluR1i-K445A by PPNDS. Membranes were incubated for 30–60 min at 0°C with 20 nM [³H]AMPA (without potassium thiocyanate) and the PPNDS concentrations indicated on the abscissa. For each experiment, binding was normalized to that in the absence of PPNDS (1.25 pmol/mg protein for wt and 0.52 pmol/mg for mutant receptors); non-specific binding was 10–20% of total binding. Averaged data (mean and sem, n=2) were then fitted with a four-point logistic equation. EC₅₀ values for PPNDS are shown as insets; Hill coefficients were 1.4 for both receptors. Right. Inhibition of [³H]FW binding to brain AMPA receptors by GDP in the absence and presence of 100 μM PPNDS. Membranes were incubated for 60 min at 0°C with 20 nM [³H]FW and the indicated GDP concentrations. PPNDS was added to the membranes together with GDP and [³H]FW. The data were normalized, averaged (mean and sem, n=3), and fitted with a logistic equation (n_Hill = 1). The IC₅₀ values are shown as insets. The ~1.5-fold shift in the IC₅₀ for GDP was also seen in tests with 1 mM PPNDS (not shown) and is due to the increase in affinity for [³H]FW caused by PPNDS. If PPNDS and GDP interactions were competitive then the IC₅₀ for GDP should have shifted about 25-fold at 100 μM PPNDS and about 250-fold at 1 mM PPNDS. Absence of competition was similarly noted in tests with the bulkier nucleotides GTP and GDP-mannose (not shown).