|Home | About | Journals | Submit | Contact Us | Français|
The mainstay of assessing guanosine diphosphate release by the α-subunit of a heterotrimeric G-protein is the [35S]guanosine 5′-O-(3-thiotriphosphate) (GTPγS) radionucleotide-binding assay. This assay requires separation of protein-bound GTPγS from free GTPγS at multiple time points followed by quantification via liquid scintillation. The arduous nature of this assay makes it difficult to quickly characterize multiple mutants, determine the effects of individual variables (e.g., temperature and Mg2+ concentration) on nucleotide exchange, or screen for small molecule modulators of Gα nucleotide binding/cycling properties. Here, we describe a robust, homogeneous, fluorescence polarization assay using a red-shifted fluorescent GTPγS probe that can rapidly determine the rate of GTPγS binding by Gα subunits.
Seven transmembrane-domain G-protein coupled receptors (GPCRs), along with their associated heterotrimeric G-proteins (Gα·guanosine diphosphate [GDP]/Gβ/Gγ), serve to transduce signals from diverse extracellular stimuli, such as photons, tastants, hormones, and neurotransmitters, to the intracellular compartment.1–3 Agonist binding to the GPCR elicits guanine nucleotide exchange factor (GEF) activity, resulting in receptor-catalyzed release of GDP by Gα and subsequent binding of guanosine triphosphate (GTP).4 The GTP-bound Gα subunit is the active signaling species, yet has an intrinsic ability to hydrolyze GTP back to GDP, which can be accelerated by a family of regulators of G-protein signaling (RGS proteins5). Historically, both receptor-catalyzed and spontaneous nucleotide release by Gα subunits has been measured using the radioactively labeled, nonhydrolyzable nucleotide [35S]guanosine 5′-O-(3-thiotriphosphate) (GTPγS).6 These radionucleotide binding assays typically involve incubation of [35S]GTPγS with the Gα subunit, followed by vacuum filtration, buffer washes, membrane dessication, and then quantification of protein-bound [35S]GTPγS by liquid scintillation.7 While producing reliable and accurate results, this method is tedious, generates radioactive waste, and is not easily amenable to automation. The use of membrane-immobilized scintillation proximity assay (SPA) beads has allowed several groups to develop high-throughput-screen (HTS)-compatible [35S]GTPγS assays8–10; however, inherent to the use of radionucleotides is the generation of unwanted radioactive waste. In an attempt to develop a nonradioactive, HTS-compatible GTPγS binding assay, others have reported using an europium-labeled GTPγS probe either in time-resolved fluorescence resonance energy transfer (TR-FRET)11–13 or in quenching resonance energy transfer (QRET).14 The ability to use QRET in a homogenous format (i.e., without the need for separation of bound and unbound Eu-GTPγS) represents an advance over the earlier TR-FRET-based assays.15 Additional, non-lanthanide-based fluorescent GTP analogs have also led to the establishment of nonradioactive assays to quantify the nucleotide cycling properties of G-proteins16–18; however, these alternate assays involve monitoring changes in the absolute intensity of the fluor as its local solvating environment changes upon binding or hydrolysis events. The need to measure absolute intensity change, coupled with the use of fluors in the green range, prevents these assays from being suitable for screening small molecule libraries for nucleotide-state modulators.19,20 Advances in plate readers capable of detecting fluorescence polarization (FP) and the commercial availability of red-shifted fluorescent-GTP analogs recently allowed Evelyn et al.21 to measure nucleotide exchange by small, Rho-subfamily GTPases using a BODIPY-Texas Red (TR)-GTPγS (Invitrogen). Here, we describe using this fluor-labeled GTPγS in a homogenous FP assay for measuring the rate of spontaneous nucleotide exchange by Gα subunits.
BODIPY-TR-GTPγS was purchased from Invitrogen and all other chemicals were purchased from Sigma at the highest quality obtainable. Two human Gα subunits were each separately expressed in Escherichia coli and purified in their GDP-bound forms by affinity chromatography exactly as previously described7,19: namely, wildtype Gαil and a double-point-mutant Gαi1(R178M/A326S) that we recently developed to have accelerated spontaneous GDP release and slowed GTP hydrolysis.7 FP experiments were conducted on the POLARStar Omega plate reader (BMG Labtech) containing a dichroic mirror and a dual emission beam splitter to measure fluorescence intensity parallel (F||) and perpendicular (F) to the excitation plane. Samples were excited at 584nm (excitation filter range of 566–588nm) and emission was detected at 630nm (cutoff±5nm). The photomultiplier tubes were calibrated so that 25nM TR-GTPγS in assay buffer (10mM Tris-HCl pH 7.5, 50mM NaCl, 10mM MgCl2, and 0.05% (v/v) NP40 alternative) had a polarization of~35mP. Polarization was calculated as P=(F||−F)/(F||+F) and expressed as mP (“milliP” or 1000*P); fluorescence intensity was calculated as I=F||+2F. Trials were conducted at 26°C using Corning Black Polystyrene 96-well plates (cat# CLS3875; Sigma). Gαil was diluted to 500nM in assay buffer and plated at an initial volume of 180μL/well. Experiments were initiated upon addition of 20μL of 250nM TR-GTPγS to each well (25nM TR-GTPγS final concentration). All experiments were conducted at least in triplicate. Nonlinear regression was used to fit the data to a single exponential association curve without constraints to calculate the kobs using Prism version 5.0c (GraphPad).
To compare results obtained by the FP assay with the traditional radioactivity-based assay, radionucleotide binding assays were performed in parallel as previously described.22 Briefly, assays were initiated by addition of [35S]GTPγS to 100nM wildtype Gαi1 or Gαi1(R178M/A326S) mutant, either in assay buffer or in assay buffer containing 100μM GTPγS. At indicated time points, aliquots were filtered by vacuum through nitrocellulose membranes and washed with ice-cold buffer. Assays were conducted in duplicate and error bars represent standard error of the mean. Nonspecific binding was subtracted from all time points. Nonlinear regression and statistical analyses were performed in Prism version 5.
Using the TR-GTPγS FP assay, observed rates of GTPγS binding (kobs) were determined to be 0.0013s−1 (95% CI 0.0011–0.0015s−1) and 0.0078s−1 (95% CI 0.0063–0.0092s−1) for wildtype Gαi1 and the double-mutant Gαi1(R178M/A326S), respectively (Fig. 1A); the greater kobs for the latter Gα subunit is wholly consistent with its known increased rate of spontaneous GDP release and thus faster GTP binding.7 The change in total intensity (Fig. 1B) was independently used to determine the kobs for both Gαi1 subunits: rates of GTPγS binding for wildtype Gαi1 and Gαi1(R178M/A326S) were determined to be 0.0016s−1 (95% CI 0.0015–0.0016s−1) and 0.010s−1 (95% CI 0.009–0.011s−1), respectively. Both the FP and fluorescence intensity results are consistent with the data obtained using the [35S]-GTPγS radionucleotide binding assay (Fig. 1C), performed exactly as previously described.22 Values of kobs using [35S]-GTPγS binding were determined to be 0.0019s−1 (95% CI 0.0015–0.0023s−1) and 0.0090s−1 (95% CI 0.007–0.0100s−1) for wildtype Gαi1 and Gαi1(R178M/A326S), respectively. While measuring fluorescence intensity and FP both allow monitoring of TR-GTPγS binding to an α-subunit of a heterotrimeric G-protein, measuring FP is considered superior given that it is a ratiometric measurement and thus less sensitive to interference from compounds that absorb or fluoresce in the same spectral region.
To validate the sensitivity of the TR-GTPγS FP assay with respect to changes in spontaneous GDP release, we used the Gαi1·GDP-binding peptide AGS3Con, derived from the consensus of the 4 GoLoco motifs of AGS3 (TMGEEDFFDLLAKSQSKRMDDQRVDLAG; ref.23) and known to exhibit GDP dissociation inhibitory activity toward Gαi·GDP subunits.24,25 A dramatic decrease in the rate of TR-GTPγS binding was observed using FP upon addition of 10μM AGS3Con peptide to wildtype Gαi1, as expected (Fig. 2).23 The rate of TR-GTPγS binding decreased from 0.0013s−1 (95% CI 0.0012–0.0014s−1) for wildtype Gαi1 alone, to 0.0004s−1 (95% CI 0.0003–0.0005s−1) for Gαi1 plus AGS3Con peptide. The addition of a 200-fold excess of nonfluorescent GTPγS was able to compete away completely any change in FP signal over the indicated time interval (Fig. 2), indicating that the TR-GTPγS does not bind nonspecificially to Gαi1. To establish that this TR-GTPγS FP assay was not specific to a single plate reader, we also performed this assay on an EnVision Alpha HTS plate reader (Perkin-Elmer) and observed similar results (data not shown).
We have described a robust, “mix-and-measure,” fluorescence-based assay system for measuring GTP binding activity by an α-subunit of a heterotrimeric G-protein. The TR-GTPγS FP assay should work universally for any Gα subunit, requires no radioactivity, and produces results consistent with radionucleotide binding assays. For researchers without the ability or desire to use radionucleotide binding assays, the TR-GTPγS FP assay offers a safer alternative. Additionally, performing this assay with the 96-well plate reader offers the advantage of automation (e.g., screening chemical libraries, profiling hits, and performing these assays on multiple Gα isoforms or multiple Gα point mutants). On the basis of its ability to detect differences in the rates of spontaneous GDP release by wildtype Gαi1 versus the Gαi1(R178M/A326S) mutant, we expect this assay to be readily utilized in assessing the GEF activity of GPCRs and nonreceptor Gα-GEFs such as Arr4 and RIC-8.26,27
Work in the Siderovski lab was supported by NIH grant R01 GM082892. R.E.M., K.R.K., and A.J.K. acknowledge early support from NIH training grant T32 GM008719, and A.J.K. acknowledges current support from NIH fellowship F30 MH074266.