HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 280, Issue 45, 38071-38080, November 11, 2005

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 280/45/38071    most recent M505259200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Abdulaev, N. G. Articles by Marino, J. P. Search for Related Content PubMed PubMed Citation Articles by Abdulaev, N. G. Articles by Marino, J. P. Social Bookmarking                      What's this?

Heterotrimeric G-protein -Subunit Adopts a "Preactivated" Conformation When Associated with -Subunits^*

Najmoutin G. Abdulaev, Tony Ngo, Cheng Zhang, Andy Dinh, Danielle M. Brabazon¶, Kevin D. Ridge1, and John P. Marino2

From the Center for Advanced Research in Biotechnology, University of Maryland Biotechnology Institute and National Institute of Standards and Technology, Rockville, Maryland 20850, the Center for Membrane Biology, Department of Biochemistry and Molecular Biology, University of Texas Health Science Center, Houston, Texas 77030, and the ^¶Department of Chemistry, Loyola College in Maryland, Baltimore, Maryland 21210

Received for publication, May 12, 2005 , and in revised form, August 23, 2005.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Activation of a heterotrimeric G-protein by an agonist-stimulated G-protein-coupled receptor requires the propagation of structural signals from the receptor binding interface to the guanine nucleotide binding pocket of the G-protein. To probe the molecular basis of this signaling process, we are applying high resolution NMR to track structural changes in an isotope-labeled, full-length G-protein -subunit (G) chimera (ChiT) associated with G-protein -subunit (G) and activated receptor (R^*) interactions. Here, we show that ChiT can be functionally reconstituted with G as assessed by aluminum fluoride-dependent changes in intrinsic tryptophan fluorescence and light-activated rhodopsin-catalyzed guanine nucleotide exchange. We further show that ¹⁵N-ChiT can be titrated with G to form stable heterotrimers at NMR concentrations. To assess structural changes in ChiT upon heterotrimer formation, HSQC spectra of the ¹⁵N-ChiT-reconstituted heterotrimer have been acquired and compared with spectra obtained for GDP/Mg²⁺-bound ¹⁵N-ChiT in the presence and absence of aluminum fluoride and guanosine 5'-3-O-(thio)triphosphate (GTPS)/Mg²⁺-bound ¹⁵N-ChiT. As anticipated, G association with ¹⁵N-ChiT results in ¹HN, ¹⁵N chemical shift changes relative to the GDP/Mg²⁺-bound state. Strikingly, however, most ¹HN, ¹⁵N chemical shift changes associated with heterotrimer formation are the same as those observed upon formation of the - and GTPS/Mg²⁺-bound states. Based on these comparative analyses, assembly of the heterotrimer appears to induce structural changes in the switch II and carboxyl-terminal regions of G ("preactivation") that may facilitate the interaction with R^* and subsequent GDP/GTP exchange.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Heterotrimeric G-proteins are intracellular signaling partners for the large family of seven transmembrane helix G-protein-coupled receptors (GPCRs).³ An agonist-activated GPCR (R^*) binds to the inactive G(GDP)·G heterotrimer to promote GDP release and GTP uptake. The binding of GTP results in conformational changes in G (reviewed in Ref. 1), ultimately leading to the dissociation of G. Both GTP-bound G and G are capable of initiating signaling cascades through interactions with downstream effectors such as adenylyl cyclase, cGMP phosphodiesterase, and various phospholipases or ion channels (reviewed in Ref. 2). The intrinsic GTPase activity of G results in the hydrolysis of GTP to GDP, returning G to its inactive GDP-bound ground state. G(GDP) eventually reassociates with G, thereby terminating all effector interactions.

Activation of the retinal heterotrimeric G-protein transducin (G_t)by the GPCR rhodopsin represents an excellent model system to probe the molecular details of activated GPCR/G-protein interactions. Rhodopsin, the rod cell photoreceptor involved in dim light vision, is by far the best studied GPCR in terms of structure and function (3). Light triggers the cis trans isomerization of the retinal chromophore to initiate structural changes in the transmembrane helices that results in the formation of the light-activated signaling state, metarhodopsin II or R^*. This is accompanied by small, yet functionally significant changes in the solvent-exposed cytoplasmic loops that lead to the formation of binding and activation sites for several signaling proteins, including G_t (2-4). Importantly, crystal structures of the dark (inactive) state of bovine rhodopsin have been solved and refined providing valuable insights into the overall structural organization of this and other GPCRs (5-9).

Crystal structures of G, including G_t (10-15) and G (16) subunits, as well as G heterotrimeric complexes (17, 18) have provided important insights into the structural rearrangements accompanying guanine nucleotide exchange and the GTPase cycle. The G subunit (Fig. 1) is composed of two domains: a guanine nucleotide binding domain with high structural homology to the Ras family of GTPases and an all -helical domain that, in combination with the GTPase domain, helps to form a deep pocket for binding the guanine nucleotide (reviewed in Ref. 3). G subunits contain three flexible regions designated switch I, switch II, and switch III that have been shown to adopt different conformations in the presence of GDP, GTPS, and other nucleotide adducts and analogs (11, 13). The GTP-bound form of G, which can be mimicked by the nonhydrolyzable GTP analog GTPS, has decreased affinity for G and increased affinity for G effectors. The GTPase domain of G (Fig. 1), a variation on the nucleotide-binding fold (19), adopts a similar conformation to that observed in crystal structures of elongation factor Tu and Ras (20, 21). The helical domain represents an insertion between the ₁ helix and the ₂ strand of the core GTPase domain and folds into a six--helix bundle. Interactions among residues that span the two-domain interface are thought to be involved in R^*-catalyzed guanine nucleotide exchange and subsequent G-protein subunit dissociation (22). G subunits also contain an extended amino-terminal region of 26-36 amino acid residues. The first 23 residues appear to be disordered in the structure of G_t in both the GDP- and the GTPS-bound states (11, 13). In contrast, structures of the G(GDP)·G heterotrimer indicate that this region forms an -helix that interacts with G (17, 18). Recent findings suggest that in addition to G binding, N-myristoylation of G also imparts structural rigidity to the amino terminus (23, 24). The extreme carboxyl-terminal region is also unresolved in the various crystal structures of G_t (11, 13, 14, 17) and in the corresponding structures for the -subunit of the inhibitory G-protein, G_i1 (10, 12, 18). High resolution NMR methods, however, have provided insights into the structures of carboxyl-terminal G_t peptides encompassing this region when bound to light-activated rhodopsin (25-27).

View larger version (32K): [in this window] [in a new window]   FIGURE 1. A ribbon representation of the GDP/Mg2+-bound form of the G-protein heterotrimer crystal structure. The GTPase and helical domains of the -subunit are colored in aquamarine and green, and the -subunits are shown in yellow and blue ribbon, respectively. Switch I, II, and III regions of the -subunit are labeled and shown with red ribbons, and Trp residues 127, 207, and 254 (W127, W207, and W254) are labeled and highlighted in white (shown in CPK). The amino and carboxyl termini of G are labeled, and the bound GDP is in purple and shown in CPK. Note that portions of the amino and carboxyl termini of G are not observed in the various crystal structures of G or G. In G, residues 6-23 in the amino terminus of G are observed as an -helix, whereas residues 1-5 in the amino terminus and 344-350 in the carboxyl terminus are unresolved. The image was generated using Protein Data Bank code 1GOT [PDB] .

Despite this wealth of available structural data, many questions remain surrounding R^*/G-protein interactions and, specifically, the functional role of G in facilitating G(GDP) binding to the activated receptor and the subsequent release and exchange of bound GDP for GTP. Previous work has demonstrated that G-protein binding to an activated GPCR requires the presence of both G and G subunits, suggesting both a direct and indirect role for G in enabling R^* interactions (33). In addition, R^*-catalyzed GDP release from G has been proposed to be mediated by G through two different mechanisms (the "lever arm" (59) and "gear shift" (60) models) based on available structural and biochemical data, as well as analogies to proposed mechanisms for guanine exchange factor (GEF)-stimulated guanine nucleotide exchange in monomeric G-proteins. Clearly, a structural understanding of the conformational changes accompanying these processes poses many unique challenges given the inherently dynamic nature of the interactions. Thus, whereas the crystallographic studies have been instrumental in obtaining "snapshot" structures of dark state rhodopsin and have revealed guanine nucleotide-dependent structural rearrangements in the switch-I, -II, and -III regions of G, conformational changes in G that accompany R^*/G interactions remain unknown. Moreover, elucidation of the structural mechanism for R^*-stimulated guanine nucleotide is severely hampered by a limited knowledge of the structures of the amino- and carboxyl-terminal regions of G, which are critical for a functional interaction with R^*.

As an initial step toward applying high resolution NMR methods to study the structure and dynamics of G in the heterotrimer state and in R^*·G complexes, we previously showed that a G chimera composed of amino acids 1-215 and 295-350 from G_t, with an intervening sequence from G_i1 (Chi6) (29), can be expressed to high levels in a soluble form using a subtilisin prodomain fusion construct (proR8FKAM) and purified in a single step using an immobilized "slow cleaving" mutant form of subtilisin (30). The mature, full-length chimera protein, which we call ChiT, exhibits a molecular mass of 40 kDa and has biochemical properties similar to those of G_t isolated from bovine retina. Most importantly, milligram quantities of isotope-labeled ChiT can be purified using this expression/purification approach, enabling functional studies under solution NMR experimental conditions. In this paper, we show that ChiT can be reconstituted with G_t subunits to form a functional heterotrimer that is amenable to structural analysis by high resolution NMR, thereby allowing us to probe differences in the conformation of G in various states as well as the structural basis for signal transfer from R^* to the heterotrimeric G-protein. Surprisingly, analysis of the chemical shift patterns in the heteronuclear single quantum correlation (HSQC) spectrum of the ¹⁵N-ChiT-reconstituted heterotrimer are found to be very similar to the spectra acquired for the - and GTPS/Mg²⁺-bound states of ChiT. In particular, chemical shifts for resonances associated with residues that report on changes involving switch II, ₃, and the carboxyl terminus are found to respond similarly to heterotrimer reconstitution and formation of the "transition/activated" states (i.e. - and GTPS/Mg²⁺-bound forms of ChiT). The functional implications of these observations with respect to the role of G-protein -subunits in "preactivating" G for interaction with R^* and GDP/GTP exchange are discussed.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Materials—Complete^TM EDTA-free protease inhibitor tablets and GTPS were from Roche Applied Science, and [³⁵S]GTPS was from PerkinElmer Life Sciences. Phenylmethylsulfonyl fluoride, isopropyl--D-thiogalactopyranoside, GDP, and D₂O were from Sigma, and blue Sepharose CL-6B was from Amersham Biosciences. ¹⁵NH₄Cl and d₁₁-Tris were from Spectra Stable Isotopes. The QuikChange II site-directed mutagenesis kit was from Stratagene, and Cymal-5 was from Anatrace. Bovine retinas were from W. Lawson Co. (Lincoln, NE). The pG58 expression vector, a fusion vector encoding a modified 77-amino acid prodomain region of subtilisin BPN' (proR8FKAM), the pG58-derived expression vector encoding a G_t/G_i1 chimera (Chi6) as a proR8FKAM fusion, and preparation of the S189 subtilisin BPN' HiTrap NHS column have been described (30, 31).

Construction of ChiT Mutants—The F350A and F350W mutants were generated using the QuikChange II mutagenesis kit according to the manufacturer's instructions with the pG58 expression vector encoding the prodomain/Chi6 fusion serving as the template. The resulting mutations were verified by DNA sequencing. Construction of the W127F, W207F, and W254F mutants has been described (30).

Expression and Purification of Subtilisin Prodomain/Chi6 Fusions—Detailed protocols for the inducible expression and purification of isotope-labeled G have been described (30). Briefly, Escherichia coli BL21 cells harboring the pG58 expression vector encoding the proR8FKAM fusion upstream of a chimeric G gene, Chi6 (29), were grown in minimal medium supplemented with 1 g/liter ¹⁵NH₄Cl as the sole nitrogen source and 100 µg/ml ampicillin at 26 °C to A₅₅₀ 0.3 and then induced with 30 µM isopropyl--D-thiogalactopyranoside for 12 h at 26 °C. The cell pellet was resuspended in 50 mM Tris-HCl, pH 8.0, containing 50 mM NaCl, 150 µM GDP, 5 mM MgCl₂, 5 mM -mercaptoethanol, 0.1 mM phenylmethylsulfonyl fluoride, and a protease inhibitor tablet and then disrupted by sonication. The supernatant obtained by centrifugation of the cell lysate at 100,000 x g for 45 min was collected and loaded onto a S189 subtilisin BPN' HiTrap NHS column (31). The prodomain-released ChiT was eluted after 12 h in 10 mM Tris-HCl, pH 7.5, containing 100 mM NaCl, 5 mM MgCl₂, 2.5 mM DTT, 50 µM GDP, and 0.1 mM phenylmethylsulfonyl fluoride (Buffer A). For NMR analysis, the purified, isotope-labeled protein was concentrated and dialyzed against 25 mM d₁₁-Tris-HCl, pH 7.5, containing 100 mM NaCl, 5 mM magnesium acetate, 2.5 mM DTT, 50 µM GDP, and 5% glycerol (Buffer B), or eluted directly from the column in Buffer B. All ChiT constructs were purified in a similar manner.

Reconstitution of ChiT with G_t to Form the G Heterotrimer—G_t was prepared from bovine retina by the method of Fung et al. (32). To separate the G_t and G_t subunits, purified G_t in Buffer A containing 50% glycerol was diluted three times with the same buffer without glycerol and applied to blue Sepharose CL-6B equilibrated with 10 mM MOPS, pH 7.5, containing 5 mM magnesium acetate and 2.5 mM DTT. G_t does not bind to the resin and is obtained from the flow-through. The column was washed with equilibration buffer, and G_t eluted in the same buffer containing 2 M NaCl. The purified G_t and G_t subunits were concentrated and stored at -20 °C in Buffer A containing 50% glycerol (plus 50 µM GDP in the case of G_t). The G-protein heterotrimer was reconstituted from ChiT and G_t essentially as described (33). Briefly, equimolar amounts of purified ChiT (or G_t) were incubated with G_t at room temperature for 15 min in 100 mM HEPES, pH 8.0, containing 1 mM EDTA, 10 mM MgSO₄, 10 mM DTT, 10% glycerol, and 50 µM GDP. Heterotrimer formation was verified by analyzing an ali-quot of the mixture by gel filtration chromatography and then isolated in large quantities by preparative gel filtration chromatography. Prior to NMR experiments, the ¹⁵N-ChiT-reconstituted heterotrimer was concentrated and dialyzed against Buffer B.

Fluorescence Assay for Measuring -dependent Changes in G_t, ChiT, ChiT-reconstituted Heterotrimer, and G_t—The tryptophan fluorescence of the GDP/Mg²⁺-bound form of ChiT and G_t in isolation and in heterotrimer were determined in signal/reference mode essentially as described (34) using a FluoroMax-2 spectrofluorometer (Instruments SA, Edison, NJ) with a 0.3-cm square cuvette at 20 °C. Emission spectra were recorded over the wavelength range of 310-450 nm with an excitation wavelength of 290 nm. The spectral excitation and emission band pass was 5 nm for all spectra, with a signal integration time of 1 s. The 150-µl assay mixture contained a 150 nM concentration of either ChiT, G_t, ChiT-reconstituted heterotrimer, or G_t in 50 mM Tris-HCl, pH 8.0, containing 50 mM NaCl, 2 mM MgCl₂, and 1 mM DTT. The aluminum fluoride-induced changes were measured by adding 5-10 µl of AlCl₃ (300 µM) and NaF (10 mM) separately or as a premixed solution. The fluorescence data were analyzed as previously described (35).

Detergent Solubilization and Purification of Rhodopsin—Rhodopsin containing rod outer segments (ROS) were prepared from bovine retina using a standard protocol (36). Rhodopsin was solubilized in phosphate-buffered saline, pH 7.0, containing 1% Cymal-5 and purified on immobilized rho-1D4 using methods previously described (37, 38) in phosphate-buffered saline, pH 7.0, containing 0.1% Cymal-5. Rhodopsin concentrations were determined by UV-visible absorption spectroscopy using a 6 spectrophotometer.

Filter Binding Assay for Measuring G-protein-mediated Guanine Nucleotide Exchange—The functionality of the ChiT-reconstituted heterotrimer was examined by following the light-activated rhodopsincatalyzed uptake of [³⁵S]GTPS by G-protein using a nitrocellulose filter binding assay (39, 40). The filter-binding assay is based on the property that G-protein and its bound [³⁵S]GTPS are retained on nitrocellulose filters, whereas free [³⁵S]GTPS passes through the filters. The 500-µl assay mixtures initially contained 100 nM ROS rhodopsin solubilized and purified in Cymal-5 detergent and 4 µM G_t, G_t reconstituted heterotrimer, or ChiT reconstituted heterotrimer in 10 mM Tris-HCl, pH 7.5, containing 100 mM NaCl, 5 mM MgCl₂, and 2 mM DTT. The samples were illuminated (>495 nm) for 1 min or allowed to remain in darkness, and the reactions were initiated by the addition of 5 µM [³⁵S]GTPS. The final concentrations of Cymal-5 detergent and glycerol in the assay mixture were 0.08 and 5%, respectively. After various time intervals (30 s to 5 min) at 20 °C, 50 µl of the reaction mixture was removed and filtered through nitrocellulose with the aid of a vacuum manifold. The filters were washed four times with 5 ml of 10 mM Tris-HCl, pH 7.5, containing 100 mM NaCl, 5 mM MgCl₂, and 2 mM DTT, dried, and analyzed for ³⁵S radioactivity by scintillation counting.

NMR Spectroscopy of ¹⁵N-ChiT and ¹⁵N-ChiT-reconstituted Heterotrimer—One-dimensional ¹⁵N-filtered and ¹⁵N HSQC water flip-back, water gate experiments (41) were recorded at 30 °C on a Bruker AVANCE 600-MHz spectrometer (Bruker Instruments, Billerica, MA) equipped with a triple resonance ¹H,¹³C,¹⁵N z axis gradient cryoprobe and linear amplifiers on all three channels. Spectra were collected on ¹⁵N-ChiT samples (150-300 µM) in Buffer B. The nitrogen frequency was centered at 118 ppm, and the proton frequency was centered on H₂O (7.5 ppm). One-dimensional spectra were collected using a sweep width of 7,200 Hz and 2,048 complex points, whereas two-dimensional data were collected using sweep widths of 7,200 Hz in ₂ and 2,000 Hz in ₁, 2,048 by 128 complex data points in t₂ and t₁, respectively, (t_1max = 293 ms and t_2max = 64 ms) and 128 scans per increment. All spectra were processed and analyzed on SGI UNIX or PC LINUX workstations using NMRPipe (42).

Trp Indole and Phe-350 Resonance Assignments—Trp indole ¹HN and ¹⁵N resonances were assigned using ChiT mutants as previously described (30). Note that G_t contains two Trp residues at positions 127 and 207, whereas ChiT contains an additional Trp residue at position 254 that is derived from the G_i1 sequence. The Phe-350 amide resonances were assigned through mutation of this residue to alanine (F350A) or tryptophan (F350W). By comparison of wild-type and mutant HSQC spectra, assignment could be made via the observation of a shift in a single resolved ¹HN,¹⁵N cross-peak belonging to residue 350.

Other Methods—Protein samples were analyzed by SDS-PAGE (43) with a 5% stacking gel and a 12% resolving gel and visualized by Coomassie Blue staining. Protein determinations were done using the method of Peterson (44) with bovine serum albumin as the standard.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Expression, Purification, and Heterotrimer Reconstitution of ¹⁵N-ChiT—Attempts to generate functional G_t through inducible bacterial expression have not been successful due to insolubility of the expressed protein. This has led to the design and construction of numerous G chimeras composed of various sequences from G_t and G_i1, some of which are more amenable to soluble expression (29, 45). The Chi6 chimera (29), which is composed of amino acids 1-215 and 295-350 from G_t and an intervening sequence (amino acids 216-294) from G_i1, contains the switch I, switch II, and receptor-interacting amino- and carboxyl-terminal regions of G_t and the switch III region of G_i1. Importantly, Chi6 exhibits basal guanine nucleotide exchange rates comparable with those of G_t (29, 35). We previously showed that this G chimera can be expressed to high levels in minimal medium formulations in a soluble form as a subtilisin BPN' prodomain fusion, and milligram quantities could be obtained in a single chromatographic step using an immobilized subtilisin mutant (30). The prodomain-released, full-length ChiT, has biochemical properties similar to those of G_t isolated from bovine retina. More importantly, ¹⁵N-labeled preparations of ChiT are amenable to study under NMR experimental conditions and undergo -dependent conformational changes (30), providing an opportunity to study the solution structures of G in various states. As shown in Fig. 2A, isolated G_t (lane 1) and purified ChiT (lane 2) comigrate and exhibit the same apparent molecular mass of 40 kDa by SDS-PAGE. Moreover, reconstitution experiments with isolated G_t subunits (Fig. 2A, lane 3), followed by analysis of the resulting protein complexes by SDS-PAGE, demonstrate that both G_t and ChiT form the corresponding heterotrimers (Fig. 2A, lanes 4 and 5, respectively). The heterotrimers are stable and show the expected 1:1 - to -subunit stoichiometry, thereby allowing us to investigate changes in the structure of ChiT accompanying heterotrimer formation and during R^*/G-protein interactions by high resolution NMR.

View larger version (22K): [in this window] [in a new window]   FIGURE 2. Reconstitution of the G-protein heterotrimer from ChiT and Gt. A, SDS-PAGE of purified and Gt-reconstituted ChiT and Gt. Isolated Gt (lane 1) and purified ChiT (lane 2) comigrate and can be reconstituted with isolated Gt subunits (lane 3) to form the corresponding Gt- and ChiT-reconstituted heterotrimers (lanes 4 and 5, respectively). The positions of molecular mass standards are shown on the left. Note that the 9-kDa Gt subunit does not stain intensely with Coomassie Blue and is therefore not seen at this level of protein loading (1-2 µg). B and C, fluorescence changes accompanying treatment of the GDP/Mg2+-bound form of ChiT in isolation and in the reconstituted heterotrimer. B, emission spectra for ChiT (150 nm) before (open circles) and after (solid circles) the addition of (final concentrations of 3 µM AlCl3 and 0.3 mM NaF) determined as described under "Experimental Procedures." Inset, emission spectra for Gt (150 nM) before (open circles) and after (solid circles) the addition of using the same experimental parameters. C, emission spectra for the ChiT-reconstituted heterotrimer (150 nM) before (open circles) and after (solid circles) the addition of determined as described under "Experimental Procedures." Inset, emission spectra for Gt (150 nM) before (open circles) and after (solid circles) the addition of using the same experimental parameters. In both B and C, representative traces from two or three independent determinations are shown.

-dependent Tryptophan Fluorescence Changes in ChiT and the ChiT-reconstituted Heterotrimer

t

t

t

t

t

R^*-catalyzed Guanine Nucleotide Exchange by the Reconstituted Heterotrimer—To test the guanine nucleotide exchange activity of the ChiT-reconstituted heterotrimer, light-activated R^*-catalyzed GTPS uptake was examined using a filter binding assay (Fig. 3). The ChiT-reconstituted heterotrimer shows nearly the same kinetics and levels of R^*-stimulated guanine nucleotide exchange activity as G_t and reconstituted heterotrimer formed from isolated G_t and G_t. These findings clearly show that the G chimera produced through bacterial expression as a subtilisin prodomain fusion not only retains the capacity to associate with G-protein -subunits but also exhibits comparable R^*-stimulated guanine nucleotide exchange.

NMR Analysis of ¹⁵ N-ChiT-reconstituted Heterotrimer—Having previously shown that ¹⁵N-ChiT can be prepared and studied by high resolution NMR (30), we set out to determine whether the ChiT-reconstituted heterotrimer, an 85-kDa complex, is also amenable for NMR study. For this purpose, ¹⁵N-ChiT in complex with GDP/Mg²⁺ was titrated with purified, unlabeled G_t to form a selectively ChiT-labeled heterotrimer. One-dimensional ¹⁵N-filtered spectra were collected to track the titration and showed that reconstitution of ¹⁵N-ChiT to form the heterotrimer resulted in both shifting and broadening of ChiT amide resonances (Fig. 4), which are attributed to the formation of the larger molecular weight complex. For example, changes in the chemical shifts for the assigned ¹HN tryptophan indole resonances are indicated on the traces and show that Trp-207 and Trp-254 shift and broaden as a result of heterotrimer formation. In contrast, the Trp-127 indole resonance remains essentially unchanged between these different states.

View larger version (17K): [in this window] [in a new window]   FIGURE 3. Time course of R*-catalyzed GTPS binding to the ChiT-reconstituted heterotrimer. Purified ChiT (circles) or Gt (triangles) was preincubated with Gt prior to the addition of purified ROS rhodopsin. Rhodopsin was also added to a similarly treated sample native Gt (squares). The samples were allowed to remain in darkness or illuminated prior to the initiation of guanine nucleotide exchange by the addition of [35S]GTPS. The amount of bound [35S]GTPS was determined using a filter binding assay as described under "Experimental Procedures." Representative data from two independent experiments are shown.

View larger version (23K): [in this window] [in a new window]   FIGURE 4. Titration of the GDP/Mg2+-bound form of 15N-ChiT with Gt subunits to form an isotope-labeled ChiT-reconstituted heterotrimer. Expansions of the amide region (10.8 to 5.9 ppm) of one-dimensional 15N-filtered proton spectra are shown for 15N-ChiT (150 µM), 15N-ChiT in the presence of 0.3 mol eq of unlabeled Gt subunits, and 1.2 mol eq of unlabeled Gt subunits. The titration of ChiT with the Gt subunits results in both a general broadening in the amide resonances of ChiT and some changes in the chemical shifts of these resonances. As an example, changes in the chemical shifts observed for the assigned Trp residue indole resonances are highlighted using green (Trp-127; W127), blue (Trp-254; W254), and red (Trp-207; W207) lines. Molar ratios of ChiT to Gt are indicated with each trace. Spectra were acquired at pH 7.5 and 30 °C using a Bruker 600-MHz NMR Cryoprobe system as described under "Experimental Procedures."

Fig. 5C shows an overlay of the HSQC spectra of the ¹⁵N-ChiT-reconstituted heterotrimer and the -bound form of ¹⁵N-ChiT. Strikingly, comparison of these two spectra reveals that the perturbations in chemical shifts of amide and side chain indole/amino resonances observed upon heterotrimer formation are nearly identical to those observed that occur upon adduct formation with GDP/Mg²⁺-bound ChiT. For example, the two cross-peaks assigned to the Trp-207 and Trp-254 indoles (Box 1), which report on conformational changes associated with the switch II and ₃ regions, respectively, as well as the cross-peak assigned to the carboxyl-terminal Phe-350 residue (Box 2), shift to similar positions in these two states. In addition, the HSQC spectrum of GTPS/Mg²⁺-bound ¹⁵N-ChiT generated via interaction of the reconstituted heterotrimer with detergent-solubilized light-activated rhodopsin also shows many changes in ¹HN and ¹⁵N chemical shifts that are similar to those observed in the spectra of the ¹⁵N-ChiT-reconstituted heterotrimer and the -bound form of ¹⁵N-ChiT. For example, comparison between the cross-peak positions observed for the ¹⁵N-ChiT reconstituted heterotrimer and the -bound form of ChiT suggests a similar conformation for the switch II and carboxyl-terminal regions in the GTPS/Mg²⁺-bound state (Fig. S1). This similarity in the HSQC spectra for the - and GTPS/Mg²⁺-bound forms of ChiT is consistent with the crystal structures for these two states of G (14), which show very similar conformations (root mean square deviation = 0.4 Å over all C carbons) and only subtle but key structural variations at the GTPase active site.

Fig. 5D highlights the changes observed for the Trp indole and carboxyl-terminal Phe-350 amide resonances. It shows the expanded regions of the HSQC spectra containing the Trp indole correlations (box 1 in Fig. 5, A-C) and the Phe-350 amide correlation (box 2 in Fig. 5, A-C), with cross-peak assignments indicated. These regions of the spectra clearly show a shift in the G structure from a "ground" (cross-peaks in black contours) to a transition/activated state conformation (cross-peaks in red ( adduct) and blue (heterotrimer) contours), which is in slow exchange on the NMR time scale (microseconds to milliseconds). Note that the broadening of the Trp-207 indole cross-peak in the heterotrimer spectrum appears more severe relative to cross-peaks observed for the Trp-127 and Trp-254 indoles, as well as other amide correlations in the spectrum, suggesting that this residue, which is located in the flexible switch II, is more conformationally dynamic in the heterotrimer state. In addition, cross-peak intensity is observed in both the ground and transition/activated state positions for the amide correlation of residue Phe-350 in both heterotrimer and -bound ChiT, suggesting that the extreme carboxyl terminus is in a dynamic equilibrium between these two conformations in these G states.

View larger version (22K): [in this window] [in a new window]   FIGURE 5. Comparison of HSQC spectra acquired for 15N-ChiT in various states. A, a two-dimensional 15N HSQC of the GDP/Mg2+-bound ChiT-reconstituted heterotrimer shown in blue contours. The assigned 1HN,15N cross-peaks for the three Trp indoles (Trp-127, Trp-207, and Trp-254) are shown in box 1, and those for the carboxyl-terminal Phe-350 residue are shown in box 2. B, overlay of HSQC spectra of the GDP/Mg2+-bound forms of ChiT (red) and the ChiT-reconstituted heterotrimer (blue). Differences in the conformations of ChiT in these different states are manifested in a number of changes in the chemical shifts of the 1HN,15N cross-peaks. C, overlay of HSQC spectra of the -bound form of ChiT (red) and the GDP/Mg2+-bound ChiT-reconstituted heterotrimer (blue). The conformations of ChiT in these different states are manifested in a number of similarities in the chemical shifts of the 1HN,15N cross-peaks. D, expansion of the Trp indole (left panel) and Phe-350 (right panel) resonance regions of the HSQC spectra of the GDP/Mg2+-(black) and -bound (red) forms of ChiT and the GDP/Mg2+-bound form of ChiT-reconstituted heterotrimer (blue). Assignments for the 1HN,15N cross-peaks of the Trp indoles and Phe-350 are indicated, and the changes in chemical shifts between ground and transition/activated states are indicated by the arrows. Note that two unassigned cross-peaks in the expanded Trp region are labeled with asterisks and may represent a second minor conformation of the heterotrimer. All spectra were acquired at pH 7.5 and 30 °C using a Bruker 600-MHz NMR Cryoprobe system as described under "Experimental Procedures."

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Our understanding of the structural basis for function in heterotrimeric G-proteins has come predominantly from a number of high resolution crystallographic studies of isolated G and G subunits as well as the holoenzyme (10-18). Although these structures reveal conformational changes in the three switch regions and thus provide a structural basis for understanding the specificity of interactions in the GDP/Mg²⁺- and GTP(GTPS)/Mg²⁺-bound states of G and the G-protein heterotrimer, they provide little information regarding the structures of the functionally significant amino- and carboxyl-terminal regions and, more generally, how activated GPCRs catalyze GDP/GTP exchange. Similarly, the availability of a crystal structure for the quiescent state of bovine rhodopsin has had a significant impact on our understanding of the functional organization for this light-sensing GPCR. What remains unclear, however, is how a heterotrimeric G-protein recognizes and interacts with an activated GPCR, R^*, leading to guanine nucleotide exchange. Our knowledge of the steps involved in these dynamic processes has largely been the realm of biochemical, mutational, and biophysical studies, which, in combination with the crystallographic data, have provided meaningful insights into the mechanism of signal transfer from R^* to the G-protein (reviewed in Refs. 1 and 47). High resolution NMR studies have also contributed to our understanding of R^*/G-protein interactions. In particular, transferred nuclear Overhauser enhancement spectroscopy methods have provided clues into the structures adopted by carboxyl-terminal G_t and G_t peptides upon binding to light-activated rhodopsin or a soluble mimic of R^* (25-27, 48, 49). Although such information is useful for understanding structural transitions at specific, localized regions of the G-protein, it is clearly evident that a comprehensive understanding of how R^*-catalyzed global changes in the structure of the G-protein or its constituent subunits are coupled to the process of guanine nucleotide exchange will absolutely require structural studies on the functionally intact proteins or protein assemblies.

We have previously shown that milligram quantities of full-length and functional isotope-labeled ChiT can be prepared using a coupled expression/purification approach based on the interaction between the prodomain region of subtilisin BPN' and a mutant form of the mature enzyme (30). NMR analysis of the GDP/Mg²⁺-bound form of the purified protein revealed a relatively well dispersed ¹⁵N HSQC spectrum with ¹HN,¹⁵N cross-peak line widths as would be expected for a 40-kDa / protein. Further, ¹⁵N-ChiT undergoes -dependent conformational changes that are characterized by a number of chemical shift perturbations, including those for two of the three Trp residues (Trp-207 and Trp-254). Collectively, these findings showed that backbone and selected side chain resonances can be used to map at atomic resolution changes in the structure and dynamics of G and raised the prospect that the application of high resolution NMR methods could be used to study the process of heterotrimer formation and R^*-stimulated guanine nucleotide exchange.

ChiT Can Be Reconstituted to Form a Functional Heterotrimer—The ability of ChiT to form a functional heterotrimer with G_t, as evidenced by differences in the level of -dependent enhancement of intrinsic tryptophan fluorescence relative to ChiT (Fig. 2, B and C), and the ability of the reconstituted heterotrimer to undergo R^*-catalyzed guanine nucleotide exchange (Fig. 3) demonstrate that our preparations of ChiT exhibit functional properties similar to those of isolated G_t. With regard to this latter point, it should be noted that bacterially expressed ChiT lacks N-myristoylation, a posttranslational modification found on G_t (50, 51) that appears to impart both structuring and stability to the amino terminus as well as facilitate R^* interactions (24, 33). The finding that the ChiT-reconstituted heterotrimer shows nearly the same kinetics and levels of R^*-stimulated guanine nucleotide exchange activity as G_t is consistent with previous observations that indicate that farnesylated G_t, which is present in our reconstituted heterotrimers, is sufficient for effective heterotrimer binding to light-activated rhodopsin in ROS membranes (33).

The observed decrease in sensitivity of the ChiT-reconstituted heterotrimer to -dependent changes in intrinsic tryptophan fluorescence, relative to the isolated subunit, is quite striking. is a well known "activator" of heterotrimeric G-proteins, and in the presence of GDP and Mg²⁺, it mimics the -phosphate of G·GTP and promotes changes in protein conformation that can lead to activation of downstream effectors (29, 34, 53, 54). Based on this property and the available crystal structures, is considered to be a representation of the transition/activated state conformation. The change in fluorescence is largely attributed to Trp-207 in switch II (46). The lower observed fluorescence increase for both the ChiT- and G_t-reconstituted heterotrimers can be simply attributed to the summation of the contributions to the fluorescence from changes in the tryptophan residues in G (20% quenching) and Trp-207 in G_t (30% increase) observed for the heterotrimer upon formation of the adduct. Although other possibilities may exist to explain this observation, preliminary data obtained using a W207F mutant of ChiT support this idea.⁴ In any case, it is clear that monitoring -dependent changes in Trp fluorescence provides a useful, albeit qualitative, measure of heterotrimer formation.

Association with G "Preactivates" G for Guanine Nucleotide Exchange—Titration of ¹⁵N-ChiT with unlabeled G_t resulted in both shifting and broadening of amide resonances (Fig. 4), which can be largely attributed to the formation of the 85-kDa heterotrimer. Changes in the chemical shifts for two of the three assigned NH tryptophan indole resonances show that Trp-207 and Trp-254 clearly shift and broaden as a result of heterotrimer formation. Broadening of Trp-207, however, is more severe than is observed on average for the majority of the indole and amide resonances, suggesting that this residue, which, as mentioned above, is located in flexible switch II region, becomes conformationally dynamic in the heterotrimer state.

Analysis of the HSQC spectrum of the ¹⁵N-ChiT reconstituted heterotrimer (Fig. 5A) demonstrates that heteronuclear NMR methods can be used to exclusively monitor changes in G associated with heterotrimer formation. By comparison of this spectrum with that acquired for the GDP/Mg²⁺-bound form of ¹⁵N-ChiT, it is clear from the numerous shifts in the ¹H,¹⁵N cross-peaks that ChiT undergoes significant conformational changes upon interaction with G, as might be expected based on the heterotrimer crystal structures (17, 18). Quite unexpectedly, however, chemical shift perturbations in the ¹⁵N HSQC spectra of the ¹⁵N-ChiT-reconstituted heterotrimer are quite similar to those observed in the - and GTPS/Mg²⁺-bound forms of ¹⁵N-ChiT (Figs. 5C and S1). For example, changes in the chemical shifts for the three tryptophan indole ¹HN,¹⁵N resonances show that both Trp-207 and Trp-254, which report on conformational changes in the switch region, are clearly shifted in the heterotrimer when compared with the GDP/Mg²⁺-bound form of ¹⁵N-ChiT yet are superimposable with those chemical shifts obtained upon adduct formation. In addition, heterotrimer formation also appears to alter the structure of the carboxyl terminus of ChiT, and this conformational change can also be induced by formation of the adduct in the GDP/Mg²⁺-bound form of ChiT. In particular, the carboxyl-terminal Phe-350 amide resonances are found to exhibit a similar chemical shift perturbation in the heterotrimer and - and GTPS/Mg²⁺-bound states, when compared with the ground state GDP/Mg²⁺-bound state. This latter finding is in agreement with the results of Herrmann et al. (33), which showed that the carboxyl-terminal region of G undergoes a conformational change upon heterotrimer formation. Taken together, our findings suggest a high degree of structural similarity between the G-associated state of G and the - and GTPS/Mg²⁺-bound states. Moreover, the conformational changes in the carboxyl terminus observed by NMR upon G association could preorganize this region for R^* interaction. Furthermore, the NMR data suggest that conformational changes in the switch II region of G upon G association result in a transition/activated structure with a guanine nucleotide binding site that is poised to make favorable contacts with the -phosphate of GTP. Through these structural changes in the guanine nucleotide binding site and the carboxyl-terminal region, G association may therefore function in stabilizing a G(GDP) conformation that is primed to facilitate R^*-catalyzed guanine nucleotide exchange.

Comparison of the NMR Results with the Conformations of the Switch II and ₃ Regions of G from Crystal Structures—Comparisons among the conformationally dynamic switch I, switch II, and switch III regions, as well as the ₃-₅ region, from the crystal structures of the GDP/Mg²⁺- and -bound forms of G_t and the chimeric G (Chi6) GDP/Mg²⁺-bound reconstituted heterotrimer highlight both similarities and differences in the conformations adopted by these regions in the various states of G (Fig. 6). For example, a comparison of the GDP/Mg²⁺-bound Chi6-reconstituted heterotrimer with the GDP/Mg²⁺-bound form of G_t (Fig. 6A) reveals differences in all three switch regions and particularly in switch II. Similarly, a comparison of these regions from the GDP/Mg²⁺-bound Chi6-reconstituted heterotrimer with the -bound form of G_t (Fig. 6B) again shows distinct differences in the conformations of the switch II region and, to a lesser extent, switch I, both of which are dissimilar from those observed from a direct comparison of the GDP/Mg²⁺- and the -bound forms of the G_t (Fig. 6C). In contrast, comparison of the -bound forms of G_t with the GTPS/Mg²⁺-bound form of G_t (Fig. 6D) reveals very similar conformations for all three switch regions, with the most significant differences between these states found in switch I.

View larger version (70K): [in this window] [in a new window]   FIGURE 6. Comparison of the conformations found for the switch and 3 regions of G from the crystal structures. A, a superposition of the polypeptide backbone of the switch I, II, and III and 3 regions in the GDP/Mg2+-bound forms of Gt in isolation (Protein Data Bank code 1TAG; yellow) and the Chi6-reconstituted heterotrimer (Protein Data Bank code 1GOT [PDB] ; blue). B, a superposition of the polypeptide backbone of the switch I, II, and III and 3 regions in the -bound form of Gt (Protein Data Bank code 1TAD [PDB] ; magenta) and the Chi6-reconstituted heterotrimer (Protein Data Bank code 1GOT [PDB] ; blue). C, a superposition of the polypeptide backbone of the switch I, II, and III and 3 regions in the GDP/Mg2+-bound forms of Gt in isolation (Protein Data Bank code 1TAG; yellow) and the -bound form of Gt (Protein Data Bank code 1TAD [PDB] ; magenta). D, a superposition of the polypeptide backbone of the switch I, II, and III and 3 regions in the -bound form of Gt (Protein Data Bank code 1TAD [PDB] ; magenta) and the GTPS/Mg2+-bound form of Gt (Protein Data Bank code 1TND [PDB] ; green). The side chains of Trp-207 in the switch II region and Trp-254 in the loop between 3 and 5 are shown in CPK in the GDP/Mg2+-bound form of ChiT (yellow), the heterotrimer state (blue), the transition/activated state of GDP/Mg2+-bound ChiT (magenta), and the GTPS/Mg2+-bound ChiT (green). In all cases, the backbones are shown in stick representation, and GDP, ( in white), and GTPS are shown in CPK. The crystal structures shown in A-C display different conformations for the switch II region in each of the different states of G, whereas the NMR data suggest that switch II as well as other regions of the G structure adopt similar conformations in the heterotrimer and the - and GTPS/Mg2+-bound states.

t

i1

t