Properties of human glycine receptors containing the hyperekplexia mutation alpha1(K276E), expressed in Xenopus oocytes

Abstract

1. Inherited defects in human glycine receptors give rise to hyperekplexia (startle disease). We expressed human glycine receptors in Xenopus oocytes, in order to examine the pharmacological and single-channel properties of receptors that contain a mutation, alpha1(K276E), associated with an atypical form of hyperekplexia. 2. Equilibrium concentration-response curves showed that recombinant human alpha1(K276E)beta receptors had a 29-fold lower glycine sensitivity than wild-type alpha1beta receptors, and a greatly reduced Hill coefficient. The maximum response to glycine also appeared much reduced, whereas the equilibrium constant for the glycine receptor antagonist strychnine was unchanged. 3. Both wild-type and mutant channels opened to multiple conductance levels with similar main conductance levels (33 pS) and weighted mean conductances (41.5 versus 49.8 pS, respectively). 4. Channel openings were shorter for the alpha1(K276E)beta mutant than for the wild-type alpha1beta, with mean overall apparent open times of 0.82 and 6.85 ms, respectively. 5. The main effect of the alpha1(K276E) mutation is to impair the opening of the channel rather than the binding of glycine. This is shown by the results of fitting glycine dose-response curves with particular postulated mechanisms, the shorter open times of mutant channels, the properties of single-channel bursts, and the lack of an effect of the mutation on the strychnine-binding site.

Figure 1. Examples of glycine-activated currents

Homomeric α1 or heteromeric α1β (ratio 1: 4) glycine receptors produced large inward currents in response to the application of glycine, reaching a maximum near 1 mM glycine (top two traces). Mutant α1(K276E) and α1(K276E)β (ratio 1: 4) receptors required higher concentrations of glycine to activate inward currents, reaching a 10- to 40-fold smaller maximum near 50 mM glycine.

Figure 2. Averaged concentration-response curves

Shown here are the averaged data for wild-type α1 (▪) and α1β (▴), and for mutant α1(K276E) (•) and α1(K276E)β (▾). The continuous line is the fit of the Hill equation. In each case the corresponding open symbols represent the response used to normalize the data and therefore not included when fitting the data.

Figure 3. Strychnine antagonism of glycine responses

A and B show examples of responses to glycine in the absence of strychnine (□), and in the presence of 50 nM (•), 100 nM (▴), 300 nM (▾) and 1000 nM (♦) strychnine; obtained from oocytes expressing either α1β (A) or α1(K276E)β (B) glycine receptors. Continuous lines show the simultaneous fits of power functions constrained to be parallel, and dotted lines show separate unconstrained fits. From results such as these, dose ratios were estimated for each concentration of strychnine. C, Schild plot obtained from the dose ratio data for α1β (K_B= 28.1 ± 2.7 nM; slope, 0.949 ± 0.14; n= 3) and α1(K276E)β (K_B= 23.1 ± 2.2 nM; slope, 1.06 ± 0.10; n= 3) glycine receptors.

Figure 4. Examples of single-channel openings of α1β and α1(K276E)β glycine receptors

Records are currents obtained from outside-out patches (held at -100 mV; 3 kHz filter) in response to the application of glycine (10 μM and 100 μM for wild-type and α1(K276E)β receptors, respectively). Note that mutant receptors open to a range of conductance levels similar to that of wild-type receptors, but that the openings are shorter.

Figure 5. Amplitudes of wild-type and mutant single-channel currents

The histograms in A and C show the distribution of amplitudes of channel openings longer than three times the rise time of the filter in a wild-type α1β patch (A) and in a α1(K276E)β patch (C). The curves superimposed on the histograms are the result of the fit of four Gaussian components to these distributions (inset: fitted parameters for each component). The raster graph in B shows the amplitude of the components fitted in wild-type and mutant patches (filled symbols represent the most common current amplitude level). Most patches have a main component between 3 and 4 pA, plus a higher component around 6 pA; the α1(K276E) mutation does not shift conductance to lower levels.

Figure 6. Apparent open periods for α1(K276E)β and α1β glycine receptors

The histograms display the distribution of the log duration of apparent open periods (frequency plotted on a square root scale) for an α1β (A) and an α1(K276E)β (B) outside-out patch. The latter are shorter than the former. The continuous lines show the result of fitting each distribution with a mixture of three exponential components (estimated parameters in the inset). τ, time constant. Agonist (glycine) concentration was 3 μM (α1β) and 100 μM (α1(K276E)β).

Figure 7. Fit of a mechanism to the concentration-response curves for heteromeric receptors

The data in both parts are the same as those shown in Fig. 2 for the heteromeric receptors, except that the values have been multiplied by the mean maximum currents given in Table 1. In this case the results have been fitted with the three binding site mechanism described in the Discussion (see Scheme 1). Equation (A1) was fitted, rather than the Hill equation used in Fig. 2, with the simplification that we took R=K₁/K₂=K₂/K₃ (see description in the text). Thus seven parameters were simultaneously fitted to the two curves; namely separate values of K₁, R and E for each and a common maximum, y_max (which is not the maximum for either curve, but the current that would be achieved if all channels were open). A, response plotted against log(concentration). B, the same data and fit as in A, but plotted using log(response), to show more clearly the quality of the fit to the mutant receptor data. The result of the fit was K₁= 0.47± 0.88 mM, R= 0.93± 1.4 and E= 42± 123 for the wild-type channel, and K₁= 0.14± 0.11 mM, R= 0.27± 0.13 and E= 0.103± 0.010 for the mutant, with a maximum log likelihood of -2.11. The weights for the fit were calculated from the standard deviations of the points, as indicated by the error bars on the graph, and these were the basis for the calculation of the errors (from the observed information matrix). These values imply, from eqn (3), that for wild-type, K₂= 0.50 mM and K₃= 0.54 mM; and for mutant, K₂= 0.52 mM and K₃= 1.9 mM.

Scheme 1

Scheme 2