Skip to main content
. 2021 Apr 21;10:e67027. doi: 10.7554/eLife.67027

Figure 1. mGluR2 and mGluR3 transmembrane domains show different dimerization propensities in a single molecule pulldown assay.

(A) Left, cartoons of SNAP-mGluR2-TMD (top) and SNAP-mGluR3-TMD (bottom) labeled with fluorophore LD555. Right, representative images showing expression and surface labeling of SNAP-mGluR2-TMD (top) and SNAP-mGluR3-TMD (bottom) in HEK 293 T cells before lysis. (B) Top, schematic showing the SiMPull setup. Bottom, representative image of single molecules with representative fluorescence time course for an individual protein complex (red circle) demonstrating two-step photobleaching. Scale bar = 10 µm. (C) Histogram summarizing the photobleaching step distribution for SNAP-mGluR2-TMD (n = 1598 total spots from 14 movies) and SNAP-mGluR3-TMD (n = 1435 total spots from 10 movies). Dashed line on the SNAP-mGluR3-TMD plot shows the normalized photobleaching step distribution for SNAP-mGluR2-TMD for comparison. (D) Bar graph showing percent dimerization for SNAP-tagged constructs. * indicates statistical significance (one-way ANOVA, p=8.9E-30; Tukey-Kramer for mGluR2 vs. mGluR3 p=0.78, for mGluR3 vs. mGluR2-TMD p=6.9E-8, for mGluR2-TMD vs. mGluR3-TMD p=8.2E-13, for mGluR3-TMD vs. β2AR p=1.7E-12). (E) Bar graph showing surface expression for constructs in (D). Values are normalized to SNAP-mGluR2. Expression is not significantly different between constructs (one-way ANOVA, p=0.64). (F) Bar graph showing the percent dimerization for SNAP-tagged constructs. * indicates statistical significance (one-way ANOVA, p=2.1E-17; Tukey-Kramer for mGluR2 vs. mGluR2-C121A p=0.27, for mGluR2-C121A vs. mGluR2-C121A-mGluR3TMD p=2.5E-5, for mGluR3 vs. mGluR3-C127A p=1.1E-12, for mGluR3-C127A vs. mGluR3-C127A-mGluR2TMD p=4.1E-8). (G) Bar graph showing surface expression for constructs in (F). Values are normalized to SNAP-mGluR2. Expression is not significantly different between constructs (one-way ANOVA, p=0.12). (H) Schematic illustrating the effect of inter-TMD interactions in mGluR2 and mGluR3 dimer assembly. The '<' and '~' symbols refer to relative differences in dimerization propensity. Number of movies analyzed for each condition is shown in parenthesis above each bar. Error bars are s.e.m. Associated figure supplements include Figure 1—figure supplements 1–5.

Figure 1—source data 1. Data and statistics for Figure 1 and associated supplemental figures.

Figure 1.

Figure 1—figure supplement 1. Further characterization of single-molecule pulldown of mGluR TMDs: lack of subunit exchange in detergent.

Figure 1—figure supplement 1.

(A) Cartoons of mGluR2-TMD constructs. Left, a SNAP-tagged TMD is labeled with LD655 and has an HA tag for antibody pulldown. Right, a CLIP-tagged TMD is labeled with CLIP-Surface 547. (B) Schematic of experimental design in which cells are co-transfected with CLIP-mGluR2-TMD and SNAP-mGluR2-TMD constructs. (C) Schematic of experimental design in which cells are transfected with either CLIP-mGluR2-TMD or SNAP-mGluR2-TMD and combined just prior to cell lysis. (D) Representative single-molecule images of co-transfected HA-SNAP-mGluR2-TMD labeled with LD655 (red circled spots) and CLIP-mGluR2-TMD labeled with CLIP-Surface 547 (green circled spots). Molecules were imaged with a 640 nm laser (left) and a 561 nm laser (right). Scale bar = 10 µm. (E) Representative single-molecule images of HA-SNAP-mGluR2-TMD labeled with LD655 (red circled spots) and CLIP-mGluR2-TMD labeled with CLIP-547 (green circled spots) mixed prior to cell lysis. Molecules were imaged with a 640 nm laser (left) and a 561 nm laser (right). Scale bar = 10 µm. (F) Bar graph showing the total number of SNAP- or CLIP-mGluR2-TMD spots for the co-transfected condition and the condition in which cells were mixed just prior to lysis. Error bars represent s.e.m.
Figure 1—figure supplement 2. Inter-TMD dimerization propensities are maintained with C-terminal pulldown of mGluR2 and mGluR3 TMDs.

Figure 1—figure supplement 2.

(A) Top, schematic showing SiMPull strategy with C-terminal pulldown. Bottom, representative single-molecule images of SNAP-mGluR2-TMD and SNAP-mGluR3-TMD pulled down via the C-terminal antibody. Scale bar = 10 µm. (B) Bar graph showing that dimer propensities characteristic of SNAP-mGluR2-TMD and SNAP-mGluR3-TMD are maintained upon C-terminal pulldown. * indicates statistical significance (unpaired t-test, p=0.0011). (C) Bar graph showing that dimer propensities characteristic of SNAP-mGluR2-TMD and SNAP-mGluR3-TMD are maintained in 0.1% IGEPAL detergent or 0.1% DDM detergent with 0.01% CHS. Number of movies analyzed for each condition is shown in parenthesis above each bar. Error bars represent s.e.m.
Figure 1—figure supplement 3. Ensemble FRET dequenching measurements show higher levels of inter-TMD FRET for mGluR2 compared to mGluR3.

Figure 1—figure supplement 3.

(A) Representative donor and acceptor images of HEK 293 T cells before and after acceptor channel bleaching for SNAP-mGluR2-TMD. (B) Summary bar graph showing a significantly higher donor recovery for mGluR2-TMD compared mGluR3-TMD and the prototypical class A GPCR, ß2AR (unpaired t-tests; for ß2AR vs. mGluR2-TMD, p=1.3E-6; for mGluR2-TMD vs. mGluR3-TMD, p=0.00060). Error bars represent s.e.m.
Figure 1—figure supplement 4. CG MD analysis of mGluR TMD dimerization.

Figure 1—figure supplement 4.

(A) Coarse-grained (CG) representations of pairs of mGluR2 or mGluR3 simulated in a hydrated POPC bilayer. Scale bar = 3 nm. (B–C) Reactive flux between the different macrostates identified by the PCCA +analysis of the MSM analysis. Macrostates encompassing asymmetric dimers (i.e. tm1/tm4 and tm4/tm1) have been aggregated for the sake of clarity. The size of the nodes is proportional to the probability of each macrostate, and the thickness of the edges is proportional to the logarithm of the total reactive flux between the nodes. Microstates were defined by labels that contained the names of helices and loops of each protomer that formed more than 20 contacts with the other protomer, with contacts defined by a cutoff of 10 Å on the minimal distance over the beads of the residue. Total probability and fraction within a macrostate were calculated aggregating microstates if their labels contained the same helices but different loops, or if their labels were equivalent after swapping protomers. Microstates with total probability above 0.1% and macrostate fraction above 0.6% in either mGluR2 or mGluR3 are indicated below as part of the top macrostate they belong to for each receptor separately. (B) Reactive flux between the different macrostates identified for mGluR2; a consisting of 33.8% microstate (tm1,tm7)/tm4, 18.2% tm3/tm7, 7.9% (tm1,tm7)/tm3, 7.9% tm4/tm7, 5.1% (tm3,tm5)/tm7, 5.0% tm1/tm4, 4.5% tm5/tm7, 4.1% tm1/tm3, 3.9% tm1/tm5, 1.9% (tm3,tm5)/tm1, 1.5% (tm1,tm7)/(tm3,tm4), 1.0% (tm1,tm7)/(tm3,tm5), 0.6% (tm3,tm4)/tm1, 0.4% tm2/tm3, 0.3% tm4/tm4; b consisting of 32.2% microstate (tm1,tm7)/tm5, 8.5% (tm1,tm7)/(tm5,tm6), 6.1% tm2/tm5, 3.0% (tm2,tm7)/tm5, 2.4% (tm1,tm5)/tm2, 1.6% (tm5,tm7)/tm7, 1.0% (tm1,tm2)/(tm5,tm6), 1.0% (tm5,tm7)/tm2, 0.8% (tm5,tm6)/tm1; c consisting of 13.0% microstate (tm1,tm7)/(tm1,tm7), 8.8% (tm1,tm7)/tm7, 8.1% tm2/tm7, 6.0% tm7/tm7, 5.1% tm1/tm7, 4.4% (tm2,tm7)/tm7, 4.3% tm1/tm1, 3.8% (tm1,tm7)/tm1, 3.5% (tm1,tm2)/(tm1,tm7), 2.6% tm1/tm2, 1.9% (tm1,tm2)/tm7, 1.7% (tm2,tm7)/tm2, 1.5% (tm2,tm7)/(tm2,tm7), 1.5% (tm2,tm3)/tm7, 1.3% tm1/tm6, 1.2% (tm1,tm7)/tm2, 1.1% (tm1,tm2)/tm1, 0.8% (tm1,tm7)/(tm2,tm7), 0.8% (tm1,tm6)/tm1, 0.7% (tm3,tm4)/tm7, 0.7% tm2/tm2, 0.6% tm6/tm6, 0.3% (tm1,tm7)/(tm2,tm3); d consisting of 47.5% microstate tm3/tm6. (C) Reactive flux between the different macrostates identified for mGluR3; a' consisting of 17.6% microstate tm2/tm3, 10.9% tm3/tm7, 9.7% (tm1,tm7)/tm4, 9.1% (tm2,tm7)/tm3, 6.2% (tm1,tm7)/tm3, 5.2% (tm3,tm5)/tm2, 4.7% tm1/tm4, 4.3% (tm1,tm2)/tm4, 3.8% tm1/tm3, 3.5% (tm2,tm7)/tm4, 2.7% (tm3,tm5)/tm7, 2.4% tm4/tm7, 1.5% (tm3,tm5)/tm1, 1.4% (tm2,tm7)/(tm3,tm5), 1.2% (tm3,tm4)/tm1, 1.1% tm2/tm4, 1.1% (tm1,tm7)/(tm3,tm4), 0.8% (tm1,tm7)/tm2, 0.8% (tm1,tm7)/(tm2,tm3), 0.7% (tm2,tm7)/tm5, 0.7% (tm1,tm7)/(tm3,tm5), 0.5% (tm3,tm4)/tm7, 0.4% tm1/tm5, 0.4% tm2/tm5, 0.3% (tm2,tm3)/tm7, 0.3% (tm2,tm7)/tm2; b' consisting of 53.7% microstate (tm1,tm7)/(tm1,tm7), 15.1% (tm1,tm7)/tm7, 10.7% tm7/tm7, 4.3% tm1/tm7, 3.7% (tm1,tm7)/tm1, 3.3% tm2/tm7, 3.1% tm2/tm2, 1.4% (tm1,tm7)/(tm2,tm7), 1.3% (tm1,tm2)/(tm1,tm7), 0.6% (tm2,tm7)/tm7; c' consisting of 32.0% microstate tm3/tm4, 20.4% tm4/tm4, 16.0% (tm3,tm5)/tm4, 8.4% tm4/tm5; d' consisting of 56.6% microstate (tm1,tm7)/(tm5,tm6), 18.7% (tm5,tm6)/tm1, 8.6% (tm1,tm7)/tm5; e' consisting of 54.8% microstate (tm1,tm7)/tm4. (D–G) Representative structures of the largest macrostates identified by MSM analysis of simulations of mGluR2 ((D) macrostate c, (E) macrostate a, and (F) macrostate b) and mGluR3 ((G) macrostate a’). Representative coarse-grained configurations of the macrostates of mGluR2 or mGluR3 were clustered based on the RMSD of the corresponding interfaces. For a given interface, the least different configuration from all others was converted into an all-atom model using backward (Wassenaar et al., 2015). Interfacial residues within 4 Å are depicted as spheres. TM1: green, TM3: yellow, TM4: cyan, TM5: orange, TM7: red.
Figure 1—figure supplement 5. Implied timescales as a function of lag time for CG MD simulation.

Figure 1—figure supplement 5.

The slowest five relaxation timescales are indicated by continuous lines. The limit of the smallest timescale that can be described at a given lag is shown in red.