Stoichiometry of the human glycine receptor revealed by direct subunit counting

Abstract

The subunit stoichiometry of heteromeric glycine-gated channels determines fundamental properties of these key inhibitory neurotransmitter receptors; however, the ratio of α1- to β-subunits per receptor remains controversial. We used single-molecule imaging and stepwise photobleaching in Xenopus oocytes to directly determine the subunit stoichiometry of a glycine receptor to be 3α1:2β. This approach allowed us to determine the receptor stoichiometry in mixed populations consisting of both heteromeric and homomeric channels, additionally revealing the quantitative proportions for the two populations.

Figure 1.

β-subunit stoichiometry of GlyR. A, Placement of the VFP in the intracellular loop between TM3 and TM4 transmembrane domains. B, An example of the first frame from a movie of a Xenopus oocyte plasma membrane expressing single GlyR channels. The expression level is low enough that the probability for two channels to coincide within the same diffraction-limited spot is small, as shown in the high-magnification region. C, Representative intensity–time trace showing two photobleaching steps. Transfection scheme with β-VFP and untagged α-subunits is shown in the inset. Below the graph, 10 images corresponding to the channel from which the time trace was extracted are shown in pseudocolor. When the first VFP photobleaches, the color of a channel changes from yellow to magenta. Photobleaching of the second VFP results in a change from magenta to green (background). Box size is 1.1 × 1.3 μm. D, Histogram of photobleaching steps counted from intensity–time traces after correction for missed events (black bars), and the best fits to the binomial distribution for two (dark gray bars) and three (white bars) β-subunits. The fit to three β-subunits is poor (p < 0.01) and gives a low percentage of fluorescent VFP (50%).

Figure 2.

α-subunit stoichiometry of GlyR. A, Representative intensity–time trace showing photobleaching steps observed in a homomeric GlyR channel. Five steps are clearly visible (arrows). B, Histogram of the number of photobleaching steps counted from intensity–time traces (black bars) after correction for missed events and the best fit to the binomial distribution for five subunits (gray bars) (goodness of fit, p = 0.82). C, Representative intensity–time trace showing three photobleaching steps observed in heteromeric channels where α1-subunit is labeled. Inset, Proposed composition of the heteromer. D, Histogram of the number of photobleaching steps counted from intensity–time traces after correction for missed events (black bars) in a mixed population of homomeric and heteromeric channels and the binomial distribution for three subunits (light gray bars) and five subunits (dark gray bars). The gray bars are calculated from the best fit to the data, in which the proportion of homomeric and heteromeric channels and the probability for the VFP to be fluorescent were left as free parameters (goodness of fit, p = 0.52). The corresponding observed and calculated percentage of traces that show one to five photobleaching steps are given in Table 2.