The mechanism of Gαq regulation of PLCβ3-catalyzed PIP2 hydrolysis

Significance For certain cellular signaling processes, the background activity of signaling enzymes must be minimal and stimulus-dependent activation robust. Nowhere is this truer than in signaling by PLCβ3 (Phospholipase Cβ), whose activity regulates intracellular Ca2+, phosphorylation by Protein Kinase C, and the activity of numerous ion channels and membrane receptors. In this study we show how PLCβ3 enzymes are regulated by two kinds of G proteins, Gβγ and Gαq. Enzyme activity studies and structures on membranes show how these G proteins act by separate, independent mechanisms, leading to a product rule of costimulation when they act together. The findings explain how cells achieve robust stimulation of PLCβ3 in the setting of low background activity, properties essential to cell health and survival.


Figures S1 to S8
Tables S1 to S3 SI Materials and Methods Appendix 1 Appendix 2 SI References Figure S1: Comparison of Gaq wildtype and hydrolysis deficient mutant, Q209L.A-B: PLCb3•Gaq complex formation with wildtype Gaq in the presence of GDP-AlF4 (A) or Gaq Q209L in the presence of GTP (B) with size exclusion chromatography profiles of the complex on a Superdex 200 10/300 increase column are on the left and accompanying SDS-PAGE gels are on right.In both cases, PLCb3 was mixed with a two-fold molar excess of Gaq, and Gaq comigrates with PLCb3.Inp refers to the input sample prior to size exclusion chromatography.C: SDS-PAGE gel of trypsin cleavage assay (1) (see methods) to evaluate the nucleotide state of purified Gaq wildtype (left) or Q209L (right).The time is in minutes and the 0 timepoint is prior to the addition of trypsin.Cleavage of GDP-bound Gaq produces a band ~ 20 kD and cleavage of GTP-bound Gaq produces a band ~37 kD (1).Purification of Gaq Q209L in the presence of GTP yields predominantly GTP-bound protein.D. GIRK current over time before and after addition of 50 nM Gaq Q209L GTP showing the current is not affected.

Figure S2:
Gaq-dependent activation of PLCb3.A: Raw PLCb3•Gaq-induced current decay and analysis scheme.The example shown is in the presence of 29 nM PLCb3 and 1.0 nM Gaq.The red bars in the bottom right corner of the trace demarcate the background leak current that remains in the bilayer when the current decay is complete.This is the background leak current that is subtracted prior to normalization and further analysis.To determine values for Vmax and KM, raw currents were leak subtracted, normalized to the maximum current before the decay, Imax, and scaled to the starting level of PIP2 in the bilayer, 1.0 mol%, which is 30% of the maximal GIRK-ALFA current with saturating PIP2.The normalized decays were fit to SI Appendix, Eq.S3 with free parameters Vmax, KM, and C. B-F: Representative normalized current decay (using 29 nM of wildtype PLCb3) fit to SI Appendix, Eq.S3 to determine Vmax and KM (red curves) in the presence of varying concentrations of Gaq, 1.0 nM (B), 10 nM (C), 50 nM (D), 100 nM (E), 300 nM (F).G-H: Representative normalized current for PLCb3 with the structured part of the X-Y linker deleted (DX-Y contact) using 290 pM enzyme fit to SI Appendix, Eq.S3 to determine Vmax and KM (red curves) in the absence (H) and presence (I) of 200 nM Gaq.Points in E and F were fit to Eq. 5 to determine Kx (dashed blue or pink curve).For wildtype PLCb3 in the presence of wildtype Gaq, Kx=3.1*10 4 , R 2 =0.81.For PLCb3 DX-Y all, Kx=5.8*10 4 , R 2 =0.96.The fit to Eq. 5 for wildtype PLCb3 alone is shown as a gray dashed curve for reference (2).PLCb3 is yellow, Gaq is pink, Gb 1 is dark teal, and Gb 2 is light blue.Gg 1 and 2 were omitted for clarity.Residues on PLCb3 that interact with G proteins are colored according to the corresponding G protein and residues on the G proteins that interact with PLCb3 are colored in yellow.Interface residues were determined using the ChimeraX interface feature using a buried surface area cutoff of 15 Å 2 .Interfaces are comparable to structures determined with each G protein on its own.
Table S3: PLCb3•G protein interface residue comparison.Interface residues were determined using the ChimeraX interface feature using a buried surface area cutoff of 15 Å
The PLCb3 construct was comprised of human PLCb3 residues 10-1234 (provided by Dr. Sondek (4), downstream of GFP with a 3C protease cleavage site between the GFP and PLCb3.For the X-Y linker deletion constructs, deleted residues (471-584 for DX-Y all, and 575-585 for DX-Y contact) were replaced with a 12 amino acid GSSG linker using the KLD enzyme mix (NEB).Expression and purification were carried out the same as for the wildtype protein.High Five insect cells were infected with 15-20 mL P3 virus per liter and harvested after 36-48 hours by centrifugation at 3,500 x g for 15 minutes.Protease inhibitors were used in all buffers, except the PLCb3 column wash buffer in which leupeptin, pepstatin A and PMSF were excluded.Cells were resuspended in PLCb3 lysis buffer (50 mM HEPES pH 8.0, 50 mM NaCl, 10 mM 2-mercaptoethanol, 5% glycerol (v/v), 0.1 mM EDTA, and 0.1 mM EGTA) and lysed by brief sonication.Lysate was clarified by centrifugation at 39,000 x g for 45 minutes and the supernatant was bound to GFP nanobody-coupled Sepharose resin for one hour.The resin was washed in batch with 10 column volumes of PLCb3 column wash buffer (20 mM HEPES pH 8.0, 400 mM NaCl, 10 mM 2-mercaptoethanol, 2% glycerol (v/v), 0.1 mM EDTA, and 0.1 mM EGTA), then loaded onto a column and washed with an additional 10 column volumes by gravity flow.Protein was eluted by cleavage with 3C PreScission protease for 1.5 hours and concentrated to ~10 mg/mL using a 15-mL Amicon concentrator with 100-kD molecular weight cutoff.Concentrated protein was subjected to size exclusion chromatography using a Superdex 200 10/300 increase column equilibrated with PLCb3 SEC buffer (20 mM HEPES pH 8.0, 100 mM NaCl, 5 mM Dithiothreitol (DTT), 2% glycerol (v/v), 0.1 mM EDTA, and 0.1 mM EGTA).Fractions containing PLCb3 were pooled, flash frozen, and stored at -80°C for later use.
To purify PLCb3 for non-specific cysteine labeling, the 2-mercaptoethanol was replaced with 2 mM tris(2-carboxyethyl)phosphine (TCEP) and peptide-based protease inhibitors were excluded from the labeling buffer.The protein-loaded resin was washed with PLCb3 labeling buffer (20 mM HEPES pH 7.4, 400 mM NaCl, 2 mM TCEP, 2% glycerol (v/v), 0.1 mM EDTA, and 0.1 mM EGTA) prior to 3C PreScission protease cleavage.Following elution, maleimide LD655 (5) was added in 5-fold molar excess and incubated overnight, protected from light.Labeled protein was concentrated to ~10 mg/mL using a 15-mL Amicon concentrator with 100-kD molecular weight cutoff and subjected to size exclusion chromatography using a Superdex 200 10/300 increase column in PLCb3 SEC buffer.Fractions with labeled PLCb3 were pooled and the labeling efficiency was evaluated.Aliquots were flash frozen and stored at -80°C for later use.Labeling efficiency was consistently 30-40%.
For lipidated Gbg, untagged human Gb1 was co-expressed with human Gg2 with an Nterminal His-YFP tag in High Five insect cells by infection with 12 and 8 mL of P3 baculovirus, respectively.Cells were harvested 36-48 hours after infection by centrifugation at 3,500 x g for 15 minutes.Pellets were resuspended in Gbg lysis buffer (25 mM Tris-HCl pH 8.0, and 125 mM NaCl) supplemented with 5 mM EGTA and 5 mM DTT and lysed by manual homogenization.Membranes were collected by centrifugation at 39,000 x g for 30 minutes, resuspended in fresh Gbg lysis buffer supplemented with dNase and protease inhibitors and manually homogenized again.Gbg was extracted using 1% sodium cholate for 1.5 hours and centrifuged at 39,000 x g for 30 minutes.The supernatant was bound in batch to TALON resin equilibrated with Gbg lysis buffer supplemented with 1% sodium cholate for one hour.The resin was washed in batch with 10 column volumes of Gbg lysis buffer supplemented with 1% sodium cholate then loaded onto a column and washed by gravity flow with 10 column volumes of high salt buffer (25 mM Tris-HCl pH 8.0, 500 mM NaCl, and 1% sodium cholate) and 10 mM imidazole buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, 1% sodium cholate, and 10 mM imidazole).Protein was eluted with Gbg lysis buffer supplemented with 1% sodium cholate and 200 mM imidazole and concentrated to ~2 mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff.The final protein was diluted to ~20 mL using Gbg lysis buffer supplemented with 1% sodium cholate and the His-YFP was removed by cleavage with 3C PreScission protease overnight.The free His-YFP was removed using TALON resin equilibrated with Gbg lysis buffer supplemented with 1% sodium cholate and 20 mM imidazole and the cleaved Gbg was concentrated to 1 mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff.Protein was purified further with size exclusion chromatography using a Superdex 200 10/300 increase column in Gbg lysis buffer supplemented with 1% sodium cholate and 5 mM DTT. Fractions containing Gbg were pooled and concentrated to 5-10 mg/mL using 4-mL Amicon concentrator with 30-kD molecular weight cutoff and immediately used for reconstitution.
Nonlipidated Gbg was generated by introducing the C68S mutation into the Gg construct, which prevents lipidation ((2, 6).In this work, the YFP was retained on the final protein.
Untagged human Gb1 was co-expressed with human Gg2 C68S with an N-terminal His-YFP tag in High Five cells by infection with 12 and 8 mL of P3 baculovirus, respectively.Cells were harvested 36-48 hours after infection by centrifugation at 3,500 x g for 15 minutes.Pellets were resuspended in Gbg lysis buffer and lysed by brief sonication.Lysate was clarified by centrifugation for 45 minutes at 39,000 x g and bound in batch to TALON resin equilibrated with Gbg lysis buffer.Resin was washed in batch with 10 column volumes of Gbg lysis buffer, loaded onto a column and washed by gravity flow with 10 column volumes of high salt buffer (25 mM Tris-HCl pH 8.0 and 500 mM NaCl) and 10 mM imidazole buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, and 10 mM imidazole).Protein was eluted with Gbg lysis buffer supplemented with 200 mM imidazole and concentrated to ~10 mg/mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff.Gbg-YFP was purified further via size exclusion chromatography using a Superdex 200 10/300 increase column equilibrated with Gbg lysis buffer supplemented with 5 mM DTT. Fractions containing Gbg-YFP were pooled, flash frozen, and stored at -80°C for later use.
For ALFA-nanobody tagged Gbg, the ALFA nanobody was inserted between the Nterminal His-YFP and the human Gg2 gene in the background of the C68S mutant.Untagged human Gb1 was co-expressed with the ALFA nanobody Gg2 construct in High Five insect cells by infection with 12 mL P3 baculovirus for each construct.Cells were harvested 36-48 hours after infection by centrifugation at 3,500 x g for 15 minutes.Cells were resuspended in Gbg lysis buffer supplemented with 5 mM DTT and lysed by brief sonication.Lysate was clarified by centrifugation at 39,000 x g for 45 minutes and bound to GFP nanobody-coupled Sepharose resin equilibrated with Gbg lysis buffer for one hour.The resin was washed in batch with 10 column volumes of Gbg lysis buffer then loaded into a column and washed with an additional 10 column volumes by gravity flow.Protein was eluted by cleavage with 3C PreScission protease for two hours, concentrated to 1 mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff, and further purified by size exclusion chromatography using a Superdex 200 10/300 increase column equilibrated with Gbg lysis buffer supplemented with 5 mM DTT. Fractions with nanobody-tagged Gbg were pooled, flash frozen, and stored at -80°C for later use.
Full-length mouse GIRK2 with a C-terminal ALFA peptide tag upstream of a GFP tag was expressed using HEK293S GnTI -by infection with 10% (v/v) P3 virus.10 mM sodium butyrate was added 12 hours after infection and the temperature was reduced to 30°C for 48 hours.Cells were harvested by centrifugation at 3,500 x g for 15 minutes.Cells were resuspended in GIRK lysis buffer (25 mM Tris-HCl pH 7.5, 150 mM KCl, and 2 mM DTT) and lysed by manual homogenization.Membranes were collected by centrifugation at 39,000 x g for 30 minutes, resuspended in fresh GIRK lysis buffer supplemented with protease inhibitors and dNase and manually homogenized.GIRK was extracted from membranes with 1.5% DDM/0.3%CHS for 1.5 hours and the extraction was centrifugated at 39,000 x g for 30 minutes.The supernatant was bound to GFP nanobody-coupled Sepharose resin equilibrated with GIRK wash buffer (20 mM Tris-HCl pH 7.5, 150 mM KCl, 2 mM DTT, and 0.05%/0.01%DDM/CHS) for one hour, washed in batch with 10 column volumes of GIRK wash buffer, then loaded into a column and washed with an additional 10 column volumes of GIRK wash buffer by gravity flow.Protein was eluted by cleavage with 3C PreScission protease for 1.5 hours, concentrated to ~10 mg/mL using a 15-mL Amicon concentrator with 100-kD molecular weight cutoff, and further purified by size exclusion chromatography using a Superose 6 10/300 increase column equilibrated with GIRK SEC buffer (20 mM Tris-HCl pH 7.5, 150 mM KCl, 10 mM DTT, and 0.025%/0.005%DDM/CHS).Fractions with GIRK were pooled, concentrated to ~2 mg/mL and used for reconstitution immediately.
Human Gaq was truncated to the methionine at position 7, which prevents palmitoylation (7), and inserted into the pFastBac vector downstream of an 8-His tag and a 3C protease cleavage site.This construct was previously shown to activate PLCb enzymes (7-10).Baculovirus was generated according to the manufactures protocol (Invitrogen).Gaq was coexpressed at a 2:1 ratio with untagged rat Ric8A to improve expression (11).High Five insect cells were infected with 7 mL of Ric8A and 15 mL Gaq p3 virus per liter of culture and harvested after 36-48 hours by centrifugation at 3,500 x g for 15 minutes.Cells were resuspended in Gaq lysis buffer ( and 30 µM GDP).Protein was eluted with elution buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, 5 mM MgCl2, 5 mM 2-mercaptoethanol, 250 mM imidazole, and 30 µM GDP) and concentrated to ~2 mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff and diluted to ~20 mL using Gaq dilution buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, 5 mM MgCl2, and 5 mM 2-mercaptoethanol) supplemented with 30 µM GDP.The His tag was cleaved by incubation with 3C PreScission protease for 1.5 hours and separated from the protein using Ni-NTA resin equilibrated with 25 mM imidazole buffer.Cleaved Gaq was concentrated to 1 mL using a 15-mL Amicon concentrator with 30-kD molecular weight cutoff, and further purified by size exclusion chromatography using a Superdex 200 10/300 increase column equilibrated with Gaq SEC buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, 5 mM MgCl2, and 5 mM DTT) supplemented with 30 µM GDP.Fractions with Gaq were pooled, flash frozen, and stored at -80°C for later use.
The Q209L mutation was generated in the same construct using the KLD enzyme mix (NEB).Expression and purification of Gaq Q209L were the same as for the wildtype protein with the following modifications.High Five insect cells were infected with 4 mL of Ric8A and 12 mL Gaq Q209L P3 virus (3:1 Gaq:Ric8A) per liter of culture and harvested after ~36 hours by centrifugation at 3,500 x g for 15 minutes.The GDP was replaced with 100 µM GTP in all buffers.The GTP-bound state of the protein was confirmed after every purification using the trypsin cleavage assay described below.
For PLCb3•Gaq Q209L complex formation (Fig. S1B), Gaq Q209L with GTP was mixed with PLCb3 in a 2:1 molar ratio in buffer containing 25 mM Tris-HCl pH 7.4, 125 mM NaCl, 3 mM MgCl2, 75 µM GDP, and 5 mM DTT and incubated on ice for one hour.The complex was purified via size exclusion chromatography using a Superdex 200 10/300 Increase column equilibrated with the same buffer.

Trypsin cleavage assay
The nucleotide state (GTP/GDP-AlF4 or GDP) of Gaq was evaluated using a trypsin cleavage assay as previously described (1).Purified Gaq was mixed with 3 µg of trypsin on ice.Samples for SDS-PAGE analysis were removed after 2 minutes and 10 minutes and immediately mixed with SDS loading dye to stop the reaction.Cleavage patterns were evaluated by SDS-PAGE.The GDP-bound state of Gaq produces a band at ~20 kD and the GTP/GDP-AlF4-bound state produces a band at ~36 kD (1)(Fig.S1C).

Nucleotide exchange of wildtype Gaq
For the structure of the PLCb3•Gbg(2)•Gaq complex and membrane partitioning studies, an additional purification step was added to ensure that all the Gaq was GDP-AlF4-associated. Gbg-YFP was bound to GFP nanobody-coupled Sepharose resin at 4°C for 2 hours to generate a Gbg column and unbound Gbg-YFP was removed by washing with 10 column volumes of Gaq SEC buffer supplemented with 30 µM GDP.Purified Gaq GDP was bound to the column by incubating at 4°C for 2 hours and unbound Gaq was removed by washing with 10 column volumes of Gaq SEC buffer supplemented with 30 µM GDP.Gaq GDP-AlF4 was eluted by incubating the column with GDP-AlF4 buffer (25 mM Tris-HCl pH 8.0, 125 mM NaCl, 50 mM MgCl2, 10 mM NaF, 30 µM AlCl3, 30 µM GDP, and 5 mM DTT) at 30°C for 2-3 hours (12).The GDP-AlF4 state of Gaq was confirmed using the trypsin cleavage assay.

Protein Reconstitution
For all reconstitutions, lipids in chloroform were mixed and dried under a stream of argon, washed with pentane, dried under a stream of argon and incubated under vacuum overnight.GIRK and lipidated Gbg were reconstituted for bilayer experiments using 3:1 ratio of 1palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE): 1-palmitoyl-2-oleoyl-snglycero-3-phospho-(1'-rac-glycerol) (POPG).For GIRK reconstitution, lipids were resuspended in 10 mM K2HPO4, 450 mM KCl, and 10 mM DTT at 20 mg/mL and sonicated to clarity.1% DM was added to the lipids and the mixture was sonicated again.GIRK was added to the mixture at a protein to lipid ratio of 1:10 (wt/wt), the lipid concentration was diluted to 10 mg/mL and incubated at 4°C for one hour.Detergent was removed at 4°C via dialysis against 10 mM K2HPO4 pH 7.4, 450 mM KCl, 10 mM DTT, and biobeads.Gbg was reconstituted in the same way but using different buffer conditions for the lipids and dialysis (10 mM HEPES pH 7.4, 150 mM KCl, and 10 mM DTT), 2% sodium cholate instead of DM, and a protein to lipid ratio of 1:5 (wt:wt).In both cases, dialysis buffer was changed every 12 hours for three changes with fresh DTT added at each change.After dialysis, liposomes were incubated with ~50% volume of biobeads for five hours at 4°C, flash frozen, and stored at -80°C until use.
For structural studies and partitioning experiments, liposomes were comprised of a 2:1:1 mixture of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE): 1-palmitoyl-2-oleoylglycero-3-phosphocholine (POPC): 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS).In reconstitutions for partition experiments, 0.1 mol% 1,2-dioleoyl-sn-glycero-3phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (18:1 Liss Rhod PE) was included in the lipid mixture.Lipids were resuspended at 25 mM in reconstitution buffer (25 mM HEPES pH 7.4, 150 mM KCl, 1 mM MgCl2, and 5 mM DTT) and sonicated to clarity.40 mM sodium cholate was added to the mixture and sonicated briefly.For protein-free liposomes for structural studies, the mixture was diluted to 20 mM lipid maintaining the 40 mM sodium cholate and incubated at 4°C for one hour.For Gbg-containing liposomes, Gbg was added at a protein to lipid ratio of 1:15 (wt/wt) and the lipid concentration was reduced to 20 mM maintaining 40 mM sodium cholate and incubated at 4°C for one hour.For protein-free liposomes for partitioning experiments, lipids were diluted to 10 mM, maintaining the 40 mM sodium cholate and incubated at 4°C for one hour.For all samples, detergent was removed using four exchanges of 200 mg/mL biobeads washed with reconstitution buffer after two hours, 12 hours, two hours, and two hours at 4°C.Liposomes for structural studies were used immediately following reconstitution to prepare grids.Liposomes for partitioning studies were flash frozen and stored at -80°C until use.
The Reference electrode and the Ground electrode were connected to the Trans and Cis chambers respectively via agarose salt bridges.A magnetic stir bar was included in the Cis chamber to facilitate continuous mixing; therefore all components of the experiment were added to the Cis chamber.Voltage across the lipid bilayer was controlled with an Axopatch 200Bamplifier in whole-cell mode.The analog current signal was lowpass filtered at 1 kHz (Bessel) and digitized at 10 kHz with a Digidata 1440A digitizer.Digitized data were recorded with the software pClamp (MolecularDevices).At +80 mV, current from channels with their PIP2-binding sites in the Trans chamber, which is inaccessible to PLCb3 added to the Cis chamber, was blocked by Mg 2+ (2).Therefore, current measured at +80 mV is only from channels sensitive to PLCb3-dependent degradation of PIP2, allowing the current to decay completely.
Gaq Q209L with GTP was used in all bilayer experiments.30 nM ALFA nanobodytagged Gbg was added to the Cis chamber at the beginning of each experiment.GIRK or Gbgcontaining liposomes were supplemented with KCl to a final concentration of 1 M, sonicated briefly at room temperature, and fused to the membrane from the Cis chamber.After fusing vesicles with GIRK, baseline current at +80 mV was measured for 3-5 minutes.Then the desired concentration of Gaq was added to the Cis chamber under continuous mixing and did not alter the baseline GIRK current (Fig. S1D).After at least 1 minute, 29 nM PLCb3 for wildtype enzyme was added to the Cis chamber under continuous mixing while recording.For experiments with wildtype PLCb3 and Gbg and Gaq, Gbg was incorporated into the bilayer via vesicle fusion and the PLCb3•Gaq complex was pre-formed on ice and added to the Cis chamber under continuous mixing to initiate the current decay.For experiments with PLCb3 DX-Y contact or DX-Y all constructs, a final concentration of 290 pM PLCb3 was used.In the presence of Gaq, the PLCb3•Gaq complex was pre-formed on ice and added to the Cis chamber under continuous mixing to initiate the current decay.After all current decays, a voltage family was measured to ensure integrity of the bilayer and then saturating C8PIP2 (32 µM) was added to recover the GIRK current.Any experiment where the current did not recover was discarded.We are confident that the starting PIP2 concentration in the membrane was not affected by vesicle fusion because the kinetics of PIP2 hydrolysis are not dependent on the number of channels that fused (2).
Analysis was carried out in Clampfit and qtiplot.To reduce the amplitude of highfrequency undesired signal owing largely to the stir bar in the recording chamber, current time series were low pass filtered at 10-20 Hz and exported to qtiplot.Decays lasting more than 100 s, i.e., at low concentrations of Gaq, were down sampled by a factor of 10 prior to exporting.We also analyzed our fastest decays without low pass filtering to ensure that the 10-20 Hz filter did not alter the determination of Vmax and KM to a significant degree (i.e., 10-20 Hz did not over filter the kinetic process under study).In qtipliot, decays were leak subtracted (where the baseline current remaining in the bilayer at the end of the decay, Ileak, was subtracted from each point in the decay) and normalized to the starting GIRK current scaled according to the starting PIP2 concentration, 1.0 mol%, which is ~30% of the GIRK current at saturating PIP2 concentrations (Fig. S2A).
To test the effects of DAG (Fig. S3), 1,2-Dioleoyl-rac-glycerol was mixed with bilayer lipids at 1.0 mol% and used to paint bilayers.This is the maximum concentration that would be present in our experiments after all the PIP2 is hydrolyzed.To test the effects of IP3 (Fig. S3), D-myo-Inositol 1,4,5-tris-phosphate trisodium salt was solubilized in water and added to the bilayer chamber at a final concentration of 1.0 µM.This concentration is several orders of magnitude higher than the maximal concentration of IP3 estimated to be present in the chamber after all the PIP2 has been hydrolyzed (~50 pM).The higher concentration should account for any possible higher local concentration present at the membrane following catalysis.
KM and Vmax for each experiment were determined as previously described (2).Briefly, we determined the relationship between normalized GIRK current, I/Imax, and PIP2 concentration in the bilayer using PIP2 titration experiments.The relationship is described by SI Appendix, Eq.S1, where A=0.88, k=1.5, and r=2.2 (2).We observed that PLCb3 operates according to Michaelis-Menten kinetics (Fig. 1D), described by Eq. 1 in the main text.The PIP2 concentration as function of time was determined by integrating Eq. 1 from t = 0 to t = t, (Eq. 2 in the main text).
Eq. 2 was substituted into SI Appendix, Eq.S1 and an additional term, C, was added to capture imperfections in the leak subtraction, yielding SI Appendix, Eq.S3, Normalized current decays were fit to SI Appendix, Eq.S3 to determine Vmax, KM, and C, which was very small, < 1% of the normalized current.
NMR experiments to measure lipid concentration 10-20 µL from each reconstitution was dissolved in ~550 uL of a mixture of deuterated methanol and chloroform (5:1) containing 100 µM of the standard sodium trimethylsilyl propionate (TSP) in 5 mm tubes.Proton spectra were measured on a Bruker 600 MHz instrument with a 5 mm HCN cryoprobe and an AVANCE NEO console.Spectra were collected at 298 K using a 30° flip angle, 16 scans, and 2.8 second acquisition time and a recycle delay of 18 µs.Spectra were processed using TopSpin 4.1.1 for phasing, baseline correction and line broadening.The lipid -CH3 peak at 0.875 ppm was integrated relative to the TSP peak at 0 ppm and normalized to the difference in protons (9 for TSP and 6 for the lipid CH3) (2).The normalized peak area was used to determine the lipid concentration using the known concentration of TSP.

PLCb3 and Gaq vesicle partition experiments
Reconstituted liposomes were subjected to 10 freeze-and-thaw cycles and extruded 21 times through a 200 nm membrane to produce LUVs.The reported lipid concentration is 50% of the total lipid concentration added in solution because partitioning proteins can only access the outer leaflet.Fixed concentrations of lipids were mixed with proteins of interest (wildtype Gaq alone, LD655 labeled PLCb3 DX-Y all alone, or wildtype LD655 labeled-PLCb3 and 200 nM wildtype Gaq or 200 nM Gaq Q209L) in partitioning buffer (25 mM HEPES pH 7.4, 150 mM KCl, 1 mM MgCl2, and 5 mM DTT) supplemented with 10 mM NaF, 30 µM GDP, and 30 µM AlCl3 for wildtype Gaq or 100 µM GTP for Gaq Q209L.Lipid-protein mixtures were incubated for one hour and centrifuged for one hour at 100,000 x g at room temperature.The supernatant was removed, and the membrane pellet was resuspended in an equal volume of buffer.The input, pellet, and supernatant samples were analyzed by SDS-PAGE.Samples with Gaq alone were stained using the Thermo Scientific silver stain kit and samples with PLCb3 were imaged using in-gel fluorescence to detect LD655-labeled PLCb3 (Fig. S4A-D).We previously showed that LD655-labeled PLCb3 has the same partition coefficient and unlabeled PLCb3 (2).Gel bands were quantified using Bio-Rad imagelab software.For experiments with PLCb3, input, supernatant, and pellet samples were solubilized in 5% Anapoe-C12E10 to eliminate scattering artifacts and the LD655 fluorescence (ext-649, em-666) and Rhodamine fluorescence (ext-560, em-583) were measured using a Tecan plate reader.The Rhodamine signal was used to estimate the fraction of lipids that were pelleted, and the measurements were corrected for this as well as the loss of material using the difference between the input and output (pellet and supernatant) LD655 signal.Each lipid concentration was repeated with two different protein concentrations, 400 and 200 nM for Gaq alone, 400 nM and 300 nM for PLCb3 DX-Y constructs, and 200 nM and 100 nM for PLCb3.Values for fraction of protein partitioned (Fp), were determined, plotted against lipid concentration, and fit to Eq. 5 to determine Kx (Fig. 2D, S4F-G).For experiments with PLCb3, values from the gels and the solution fluorescence measurements were consistent.The values from the solution measurements are reported.

Cryo-EM sample preparation and data collection
For the PLCb3•Gaq complex with reconstituted liposomes, Gaq was incubated on ice for 30 minutes in the presence of GDP-AlF4 (25 mM HEPES pH 7.4, 150 mM KCl, 5 mM MgCl2, 0.9 mM CaCl2, 10 mM NaF, 50 µM GDP, 30 µM AlCl3, and 5 mM DTT), then mixed with PLCb3 in a 2:1 molar ratio and incubated on ice for one hour.The complex was purified via size exclusion chromatography using a Superdex 200 10/300 Increase column equilibrated with 25 mM HEPES pH 7.4, 150 mM KCl, 5 mM MgCl2, 0.9 mM CaCl2, 10 mM NaF, 50 µM GDP, 30 µM AlCl3, and 5 mM DTT (Fig. S1A).Purified PLCb3•Gaq complex was mixed with vesicles at final concentrations of 3.6 µM complex and 17.5 mM lipids and incubated at room temperature for one hour.For the PLCb3•Gbg(2)•Gaq complex, Gaq GDP-AlF4 was mixed with PLCb3 in a 2:1 molar ratio, incubated on ice for one hour and exchanged into buffer containing 25 mM HEPES pH 7.4, 150 mM KCl, 5 mM MgCl2, 0.9 mM CaCl2, 10 mM NaF, 50 µM GDP, 30 µM AlCl3, and 5 mM DTT using a PD10 desalting column.Purified PLCb3•Gaq complex was mixed with Gbgcontaining vesicles (reconstituted at 1:15 wt/wt) at final concentrations of 4.1 µM complex and 16.7 mM lipids and incubated at room temperature for one hour.Both samples were supplemented with 3 mM Fluorinated Fos-Choline-8 ~5 minutes before grid preparation.Quantifoil R1.2/1.3 400 mesh holey carbon Au grids were glow discharged for 20s, and 3.5 µL of sample was applied, incubated for 5 minutes at 22°C and 100% humidity, and manually blotted from below.An additional 3.5 µL of sample was applied and after 30 seconds of incubation, grids were blotted for 3.5s with a blot force of 0 and plunge frozen in liquid ethane using a FEI Vitrobot Mark IV.For data acquisition, grids were loaded onto a 300-kV Titan Krios transmission electron microscope, located at the HHMI Janelia Research Campus, with a Gatan K3 Summit direct electron detector and a GIF quantum energy filter with a slit width of 20 eV.25,668 movies were collected for the PLCb3•Gaq complex and 29,832 movies were collected for the PLCb3•Gbg(2)•Gaq complex in superresolution mode with a pixel size of 0.4195 Å and a defocus range of 1.5 to 2.5 µm using SerialEM (13).The movies were recorded with 50 frames, a total dose of 60 e -/Å 2 (1.2 e -/Å 2 /frame) and a 4.26 second total exposure time (0.085s/frame) for the PLCb3•Gaq complex or a 3.86 second total exposure time (0.071s/frame) for the PLCb3•Gbg(2)•Gaq complex.

Cryo-EM data processing
For both complexes, motion correction was performed with 2x binning using the RELION implementation (in RELION 3.1) and CTF estimation was carried out using CTFfind4 (14)(15)(16).Particle picking was caried out using the model trained for PLCb3 on vesicles in crYOLO (2,17).For the PLCb3•Gaq complex, 3,543,739 particles were picked and extracted with 2x binning and a 260 Å box size, sorted to 2,825,696 using a resolution cutoff of 4 Å, and sorted to 2,342,176 using 2D classification in cryoSPARC.Iterative rounds of ab initio reconstruction and heterogenous refinement in cryoSPARC were carried out to generate an initial reconstruction with density for the membrane and protein protruding (Fig. S5F).This map was used as an input for heterogenous refinement in cryoSPARC to sort particles based on membrane alignment.592,280 particles with good membrane alignment were selected, subjected to refinement in RELION, and signal subtraction was applied to remove the membrane density.Subtracted particles were subjected to 2D classification in cryoSPARC and particles from the best 2D classes, 213,197, were used for ab initio reconstruction to obtain an initial map resembling the PLCb3•Gaq complex.All subtracted particles were subjected to iterative heterogenous refinement using this map as input.A final subset of 229,523 particles yielding a reconstruction with clear secondary structure features was un-subtracted and re-extracted without binning and a 289 Å box size.These particles were subjected to several cycles of Bayesian polishing and CTF refinement in RELION (18,19) and local refinement with a mask on the entire complex in cryoSPARC.After polishing, 3D classification without alignment was carried out to improve the resolution, search for alternate configurations of the X-Y linker and obtain density for the membrane.The best subset contained 67,454 particles and resulted in a 3.4 Å reconstruction from local refinement in cryoSPARC.Two subsets containing 20,261 and 15,486 particles yielded reconstructions with density for the membrane (Fig. 6B-C).No reconstructions with differing density for the X-Y linker were observed.
For the PLCb3•Gbg(2)•Gaq complex, 5,332,307 particles were picked and extracted with 2x binning and a 289 Å box size, sorted to 5,128,184 particles using a resolution cutoff of 5 Å, and sorted to 4,917,142 particles using 2D classification in cryoSPARC.271,175 particles from the best 2D classes were subjected to ab initio reconstruction and a map with density for the membrane and two blobs of protein protruding was produced.This reconstruction was used as an input for heterogenous refinement on all 4,917,142 particles to sort based on membrane alignment.Three classes with 1,120,862, 1,235,752, and 1,682,270 particles (4,038,884 particles total) were refined separately in RELION and subjected to signal subtraction to remove the membrane density.Subtracted particles were subjected to 2D classification in cryoSPARC.195,750 particles from the best 2D classes were subjected to ab initio reconstruction to obtain a map resembling the PLCb3•Gbg(2)•Gaq complex.This map was used as an input for iterative heterogenous refinement on all 4,038,884 subtracted particles.484,174 particles yielding a reconstruction with clear secondary structure features were un-subtracted and re-extracted without binning and a 289 Å box size.These particles were subjected to several rounds of Bayesian polishing and CTF refinement in RELION (18,19) and local refinement with a mask on the entire complex in cryoSPARC.After polishing, 3D classification without alignment was carried out to improve the resolution, search for alternate configurations of the X-Y linker and obtain density for the membrane.The best subset contained 359,215 particles and resulted in a 3.4 Å reconstruction from local refinement in cryoSPARC.Six subsets containing 19,266, 15,138, 27,763, 21,151, 17355, and 12,194 particles yielded reconstructions with density for the membrane (Fig. 6D).No reconstructions with differing density for the X-Y linker were observed.A subset of 48,367 particles was identified that only contained density for PLCb3 and Gbg.Refinement of these particles yielded a reconstruction that was very similar to the previously reported PLCb3-Gbg complex and these particles were excluded from the final subset.
For the conditions shown in Panel C, there is a deviation between the fitted and true values of KM and Vmax.These conditions correspond to our experiments in which a saturating quantity of Gbg is used, which increases the local enzyme concentration.Even in this most extreme case, Etot ≪S[0] + KM remains true, and the expected inaccuracy is small, likely within the uncertainty of our experimental data.
The reader should realize that the only real difference between measuring initial rates at various substrate concentrations and measuring the slope along the decay curve (which is then paired with the substrate concentration on the Y-axis) is the presence of product in the latter and the absence of product in the former.These approaches should give the same result if the products do not alter the function of the enzyme.For PLCb3 we do not observe product inhibition (Fig. S3).  the concentration profile for kcat = 1.7 s -1 (black) and 2000 s -1 (blue).Even at this lower PIP2 concentration, the reaction is far from diffusion limited.While several assumptions are made in this calculation, it presents a rational argument that the PLCb3 reaction under our assay conditions are far from diffusion limited.While a rate increase from 2000 s -1 to 35 times 2000 s -1 (The fold activation by Gaq) is not possible based on the diffusion limited example above, we should see a substantial rate increase, but we do not.For this reason, we hypothesize that the increased kcat, from 1.7 s -1 to ~60 s -1 , due to binding of Gaq to PLCb3, results from allosteric regulation mediated at least in part through the X-Y linker.

Figure S3 :
Figure S3: PLCb3 is not inhibited by the products of its reaction, IP3 and DAG, in our experimental setup.A-B: normalized PLCb3-dependent current decays using 29 nM enzyme in the presence of lipidated Gbg and 1.0 mol% DAG (A) or 1.0 µM IP3 (B).Red curves are fits to SI Appendix, Eq.S3. to determine Vmax and KM.R 2 =0.97 (A) and 1.0 (B).C-D: comparison of Vmax (C) or KM (D) in the presence and absence of 1.0 mol% DAG and 1.0 µM IP3.The bars are the mean, the error bars are standard error of mean, and the circles are values from individual experiments.Data for PLCb3 alone were reported in (2).

Figure S4 :
Figure S4: PLCb3 partitioning in the presence of Gaq.A: Silver stained SDS-PAGE gels of partitioning experiments for 200 or 400 nM wildtype Gaq with GDP-AlF4 (plotted in E).I is input, S is supernatant, and P is pellet.No membrane partitioning behavior is evident as protein in the pellet (P) does not increase with increasing lipid concentration.B-D: SDS-PAGE gels imaged for LD655 fluorescence of 100 or 200 nM wildtype PLCb3 partitioning experiments in the presence of 200 nM Gaq Q209L with GTP (B), 200 nM wildtype Gaq GDP-AlF4 (C) or 400 and 300 nM PLCb3 DX-Y all (D).E-G: Membrane partitioning curve for Gaq alone (E), wildtype PLCb3 in the presence of wildtype Gaq with GDP-AlF4 (F-blue) or PLCb3 DX-Y all (G-pink) for 2DOPE:1POPC:1POPS LUVs with Fraction Partitioned (Fp) on the Y axis.Points are the average from 2 repeats for each lipid concentration and error bars are range of mean.Points in E and F were fit to Eq. 5 to determine Kx (dashed blue or pink curve).For wildtype PLCb3 in the presence of wildtype Gaq, Kx=3.1*10 4 , R 2 =0.81.For PLCb3 DX-Y all, Kx=5.8*10 4 , R 2 =0.96.The fit to Eq. 5 for wildtype PLCb3 alone is shown as a gray dashed curve for reference (2).

Figure S5 :
Figure S5: Structure of the PLCb3•Gaq complex.A: Representative micrograph.B: Fourier shell correlation (FSC) curves for the unmasked (black) and masked (red) maps and between the map and model (blue).The 0.143 and 0.5 thresholds are denoted by dashed lines.C: 3D FSC plot for the final masked map (3).D: Final masked, sharpened map colored by local resolution determined by cryoSPARC.E: Angular distribution plot for the final masked map from cryoSPARC.F: Summary of data processing steps, see methods.Maps shown are unsharpened.

Figure S6 :
Figure S6: Structure of the PLCb3×Gbg(2)×Gaq complex.A: Representative micrograph.B: Fourier shell correlation (FSC) curves for the unmasked (black) and masked (red) maps and between the map and model (blue).The 0.143 and 0.5 thresholds are denoted by dashed lines.C: 3D FSC plot for the final masked map (3).D: Final masked, sharpened map colored by local resolution determined by cryoSPARC.E: Angular distribution plot for the final masked map from cryoSPARC.F: Summary of data processing steps, see methods.Maps shown are unsharpened.

Figure S7 :
Figure S7: PLCb3-G protein interfaces.A: Hydrogen bonds at the PLCb3•Gaq interface.PLCb3 is yellow and Gaq is pink.Relevant side chains are shown as sticks and colored by heteroatom.Hydrogen bonds are denoted by dashed black lines.All hydrogen bonds shown are less than 3.5 Å apart.B-D: Surface representation of the PLC-Gaq (B), PLCb3-Gb 1 (C), or PLCb3-Gb 2 (D) interfaces in the PLCb3•Gbg(2)•Gaq complex peeled apart to show extensive interactions.PLCb3 is yellow, Gaq is pink, Gb 1 is dark teal, and Gb 2 is light blue.Gg 1 and 2 were omitted for clarity.Residues on PLCb3 that interact with G proteins are colored according to the corresponding G protein and residues on the G proteins that interact with PLCb3 are colored in yellow.Interface residues were determined using the ChimeraX interface feature using a buried surface area cutoff of 15 Å 2 .Interfaces are comparable to structures determined with each G protein on its own.

Figure S8 :
Figure S8: Potential role of the Ha2' in membrane association of the PLCb3 catalytic core.A-B: Catalytic core of apo (A) or Gaq-bound (B) PLCb3 colored by hydrophobicity highlighting a hydrophobic patch covered by Ha2' in the apo conformation.C-D: Hydrophobicity-colored apo (C) or Gaq-bound (D) PLCb3 catalytic core fit into the density of the membrane associated PLCb3•Gaq complex.The exposed hydrophobic patch is close to the position of membrane association in the Gaq-bound conformation and the Ha2' associated with the catalytic core could hinder membrane association.

Table S1 :
Cryo-EM collection parameters and model statistics