CHIP2 Binding Experiments

It was noted that the structural biology of SIM’s had been investigated in a salt- free, 25mM phosphate pH7.0 (Conti et al., 2008; Hecker et al., 2006). Non-ionic detergent P20 was added for further SPR study. The next experiment used SUMO1 as the analyte over the ligands RGA, GID1a and GST, with variable dilutions of SUMO1, +/- gibberellin.

Figure 5-14c/d display the binding of SUMO1 to GID1a and GST. The sensorgram showing association between SUMO1 and GST (Figure 5-14d) shows no interaction between the proteins. This suggests that no binding of SUMO1 and GST occurs in a low salt buffer. In contrast, the sensorgram for SUMO1 and GID1a (Figure 5-14c) shows binding of the SUMO1 protein to GID1a and subsequent dissociation. The sensorgram has one trace that is slightly negative, which corresponds to a ⅛ dilution of SUMO1, and two traces that are horizontal, running along the baseline of the graph that are controls with no

SUMO1 protein in them. The rest of the traces had a curve, showing binding and dissociation as described. At a ¼ dilution of SUMO1, binding peaked at ~20 resonance units, and the two replicates of ½ dilution of SUMO1 peak at approximately 50 resonance units, demonstrating a weak, but clear concentration dependence to the association of SUMO1 to GID1a.

Figure 5-14 – SPR sensorgram traces showing SUMO1 protein binding to GID1a, not GST. All assays were done in 25mM phosphate buffer pH7 (+0.005% P20). Dilutions of SUMO1 and different concentrations of gibberellin were used as indicated: - SUMO1 (1/2) + 0, 1, 10 or 100μM GA or GID1a (a) or GST (b). SUMO1 (½, ¼ or ⅛) over GID1a (c) or GST (d). 10mM Glycine-HCl pH2.2 was used for chip regeneration. Each sensorgram shows data for test channel subtracting ch1. Concentration of SUMO1 (½) = 0.11mg/ml.

Having studied the interaction of the proteins at different quantities of SUMO1, gibberellin was introduced in order to see whether there was any level of GA dependence in this interaction. Figure 5-14b shows the interaction of SUMO1 with GST, at variable concentrations of GA from 0-100μM. The traces suggest no binding interaction between the proteins. By comparison, Figure 5-14a displays the interaction between SUMO1 and GID1a. As above (Figure 5-14c), the data all indicate binding between SUMO1 and GID1a, however no effect of GA can be discerned.

In order to test the effect was due to conditions, not variation in chip or proteins, the experiment was repeated to test the effect with low salt, phosphate buffer and with HBS-EP buffer. The results confirmed that the SUMO1-GID1a interaction was not supported in saline buffer.

Having also bound RGA onto the CM5 CHIP (Figure 5-13), we were able to test the interaction of RGA, as the ligand, with SUMO1 as analyte in low salt buffer.

Figure 5-15 – No convincing interaction was found between SUMO1 and RGA. 25mM phosphate buffer pH7 (0.005% P20). Reactions took place in the absence or presence of gibberellin at 0, 1, 10 and 100μM. 10mM Glycine-HCl pH2.2 was used for chip regeneration. Each sensorgram shows data for test channel subtracting ch1. Concentration of SUMO1 = 0.11mg/ml.

The interaction of SUMO1 with RGA was investigated at a series of different gibberellin concentrations, and also in the absence of gibberellin (Figure 5-15). Apart from controls lacking proteins in the analyte, all traces had a characteristic shape, building up in a curve to a peak at approximately 50-60 resonance units, then a plateau. Clustered in this second set of traces are all of the samples with SUMO1 included, with or without gibberellin. There is no GA- dependence and less than 10 resonance units difference between the highest and lowest samples. This means that the association of all the SUMO1 samples, whether with or without gibberellin, is fairly comparable and the very long dissociation rates again suggest that RGA is a ‘sticky’ protein.

Despite the evidence to suggest RGA was sticky, the next step was to test RGA and GID1a. This interaction was assessed, again using a low salt 25mM phosphate buffer at pH7 (Figure 5-16).

Figure 5-16 – RGA binds to GID1a, but fails to dissociate effectively. SPR sensorgram showing RGA binding to the ligands GID1a (a) and GST (b) in 25mM phosphate buffer pH7 (0.005% P20). Reactions took place in the absence or presence of 10μM gibberellin. 10mM Glycine-HCl pH2.2 was used for chip regeneration. Each sensorgram shows data for test channel subtracting ch1. Concentration of RGA (½) = 0.09mg/ml.

Figure 5-16bdemonstrates that there is no interaction between RGA and GST. Figure 5-16a shows interaction of the ligand GID1a with a series of dilutions of RGA (¼, ⅛, and 1/16) in the presence and absence of 10μM GA. Three samples gave positive binding associations, building up in a curve to a peak, and then reaching a plateau. The lowest of these traces is the sample of RGA ¼ + GA, with a maximum reading in resonance units of around 40. The middle trace is the sample RGA 1/8 + GA, which peaks higher at a reading of 80 resonance units. The highest curve is that of RGA ¼ on GID1a, with resonance units of approximately 100 for the interaction. These results give a mixed message, in that there appears to be an interaction between RGA and GID1a, and this binding may have some GA dependency, but the data do not show classical binding curves and the dissociation is, again, very slow.

5.5 Discussion