Small-angle neutron scattering studies on the AMPA receptor GluA2 in the resting, AMPA-bound and GYKI-53655-bound states

In this study, the behaviour of the detergent-solubilized tetrameric, full-length ionotropic glutamate receptor GluA2 in solution was investigated using small-angle neutron scattering. It was found that the GluA2 solution structure is preferentially in a compact form in the resting state as well as in the presence of AMPA and of the negative allosteric modulator GYKI-53655.


Table S1
Information about the samples, the SANS measurements and the software used for the data analysis. with BCA assay for GluA in the resting state and GluA2 AMPA-bound at pH 5.5. * 4 It was not possible to obtain fully linear region at < 1.3 (Fig. S6C), so the the values may be incorrect.
* 5 determined with the Fischer method (Fischer et al., 2011) with parameters given in Table S3.
* 6 The dummy atom model for GluA2 apo is approximate, since aggregation was not taken into account. For the same reason, a dummy atom model was only generated for the GluA2 apo sample, where the aggregation scattering contribution was very minor. Supporting information, sup-5

Table S2
Fischer and Petoukhov determination (Petoukhov et al., 2012;Fisher et al., 2010), where is determined via. the scattering "invariant" (Porod, 1982) The upper integration limit used to determine was 8/ (Petoukhov et al., 2012). is the apparent volume, and is the same for the two methods. In the Fischer method, linear coefficients and given in the table are used to convert to the Porod volume , and the weight-to-volume conversion constant of 0.83 kDa/nm 3 to obtain . In the Petoukhov method, is determined directly from the using the conversion constant 0.625 kDa/nm 3 . The constant subtracted backgrounds were used to assure a constant plateau in the Porod plots (Fig. S2) and the data sets were extrapolated to = 0 by simple linear extrapolation. An implementation in MATLAB of the methods was used.
The value for obtained with the Fisher method is given in Table S1 and used in the paper, since this method takes the size of the particle into account, which adds an important correction for large proteins such as GluA2. Values of and (0) from the ( ) analysis were used (Table S1).  (Fischer et al., 2010, p. 106), and a 20% uncertainty on (Petoukhov et al., 2012, p. 344

Figure S1
Guinier plots and residual plots for GluA2 in the resting state (A), in the AMPA bound state at pH 7.5 (B), in the AMPA bound state at pH 5.5 (C) and in the GYKI-53655 bound state (D). Residuals show the difference between log( ) and the fit, weighted with the errors on log( ). Resulting values for (0) and are given in Table S1. The AMPA bound state at pH 5.5 (panel C) does not have a fully linear Guinier region at < 1.3, meaning that the values for (0) and may be wrong. The values of (0) and from the ( ) funciton was therefore used for determination.

Figure S2
Porod plots (black) for GluA2 in the resting state (A), in the AMPA bound state at pH 7.5 (B), in the AMPA bound state at pH 5.5 (C) and in the GYKI-53655 bound state (D). Additional constant backgrounds were subtracted to give a constant behavior at high-(red). The constants are listed in Table   S2.

Figure S3
Kratky plots for GluA2 in the resting state (A), in the AMPA bound state at pH 7.5 (B), in the AMPA bound state at pH 5.5 (C) and in the GYKI-53655 bound state (D). Constant backgrounds were subtracted, and listed in Table S2.