File Name : membrane_protein_interactions_sifig1.eps

Caption : fig s1 {{\bf (a)} schematic of the microfluidic device design used in this study. the channel passes through the detection region four times in order to collect data for diffusion at multiple time points within a single image. {\bf (b)} observed $d$ for 1~$\mu$m cam as a function of aqp0 concentration. fitting the data with one binding site per tetramer results in a $k_d$ of 1.3~$\mu$m. {\bf (c)} fit to the diffusion data with two $k_d$ values and two sequential binding events to allow for cooperativity gives $k_{d1}$\,=\,2.4~$\mu$m and $k_{d2}$\,=\,2.4~$\mu$m. the shaded areas covers a factor 2 in the k$_{\rm d}$ values and the fitted diffusion coefficients $\pm$ the mean percentage error on the measured d.

File Name : membrane_protein_interactions_sifig2.eps

Caption : fig s2. the observed $\mu_{\rm obs}$ of cam against an increasing concentration of aqp2 tetramer. errorbars represent the standard deviation of three independent measurement repeats for each sample concentration.

File Name : membrane_protein_interactions_sifig3.eps

Caption : fig s3. {\bf (a)} data fits for 1 binding site per aqp0 tetramer. left, the observed $\mu_{\rm obs}$ for 1~$\mu$m cam in the presence of increasing aqp0. right, the observed $\mu_e$ for 1.25~$\mu$m aqp0 tetramer as a function of cam concentration. {\bf (b)} data from {\bf (a)} fitted to a model with sequential binding to two sites per aqp0 tetramer and two $k_d$ values. the shaded areas covers a factor 2 in the k$_{\rm d}$ values and the fitted diffusion coefficients $\pm$ the mean percentage error on the measured $\mu_{obs}$.

File Name : membrane_protein_interactions_sifig4.eps

Caption : fig. s4{\bf (a)} using the adair equation to fit the observed $\mu_{\rm obs}$ for 1.25~$\mu$m aqp0 tetramer as a function of cam concentration. {\bf (b)} zoom of 0\,-\,5~$\mu$m cam in {\bf (a)}. the shaded areas covers a factor 2 in the k$_d$ values and the fitted diffusion coefficients $\pm$ the mean percentage error on the measured $\mu_{\rm obs}$.}