Research ArticleBIOCHEMISTRY

Decreased conformational stability in the oncogenic N92I mutant of Ras-related C3 botulinum toxin substrate 1

See allHide authors and affiliations

Science Advances  07 Aug 2019:
Vol. 5, no. 8, eaax1595
DOI: 10.1126/sciadv.aax1595

Abstract

Ras-related C3 botulinum toxin substrate 1 (Rac1) functions as a molecular switch by cycling between an inactive guanosine diphosphate (GDP)–bound state and an active guanosine triphosphate (GTP)–bound state. An oncogenic mutant of Rac1, an N92I mutant, strongly promotes cell proliferation and subsequent oncogenic activities by facilitating the intrinsic GDP dissociation in the inactive GDP-bound state. Here, we used solution nuclear magnetic resonance spectroscopy to investigate the activation mechanism of the N92I mutant. We found that the static structure of the GDP binding site is not markedly perturbed by the mutation, but the overall conformational stability decreases in the N92I mutant, which then facilitates GDP dissociation by lowering the activation energy for the dissociation reaction. On the basis of these results, we proposed the activation mechanism of the N92I mutant, in which the decreased conformational stability plays important roles in its activation process.

INTRODUCTION

Ras-related C3 botulinum toxin substrate 1 (Rac1) is a member of the Rho family of small guanine nucleotide-binding proteins (G proteins), which plays critical roles in the maintenance of cell morphology and cell migration (1, 2). Rac1 functions as a binary molecular switch by cycling between two distinct functional states depending on the bound nucleotide, an inactive guanosine diphosphate (GDP)–bound state and an active guanosine triphosphate (GTP)–bound state. The active GTP-bound Rac1 binds to different classes of proteins, including kinases, scaffold proteins, and enzymes, thus leading to the activation of downstream signaling pathways. The aberrant functions of Rac1 are closely related to diseases, especially cancer generation and progression. Recently, extensive sequencing analyses have revealed that gain-of-function mutations of Rac1 are frequently found in human cancers and that these mutations strongly promote cell proliferation and subsequent oncogenic activities. The oncogenic mutation, Pro29 to Ser, was first identified from exome sequencing analyses of sun-exposed melanomas (3, 4), and subsequently, another mutation, Asn92 to Ile, was found from deep sequencing analyses of the HT1080 human sarcoma cell line (5). Biochemical studies of these oncogenic mutant proteins have revealed that the intrinsic GDP dissociation reaction is accelerated in these mutants, and hence, they are maintained in the active GTP-bound state (5, 6).

Recently, we have investigated the activation mechanism of the oncogenic P29S mutant by using solution nuclear magnetic resonance (NMR) techniques (7). Our NMR results revealed that, in the P29S mutant, the preexisting conformational equilibrium in the switch 1 region (residues 30 to 38) is shifted toward the minor conformation that exhibits low affinity for Mg2+, a cofactor that stabilizes the GDP binding, and that the increase in the Mg2+-unbound state greatly facilitates the overall GDP dissociation in the P29S mutant. Whereas the activation mechanism of the P29S mutant has been well characterized, less is known about the activation mechanism of another oncogenic mutant, the N92I mutant. Because the Asn92 residue is distant from the switch 1 region, the activation mechanism of the N92I mutant is expected to be distinct from that of the P29S mutant. In the crystal structures of the wild-type GDP-bound Rac2 in the Rho–guanine nucleotide dissociation inhibitor (GDI)–bound state [Protein Data Bank (PDB) ID: 1DS6] and Rac1 in the uncomplexed state (PDB ID: 5N6O), the side chain of Asn92 forms a hydrogen bond with the side chain of Asp11 located on the phosphate-binding loop (P-loop; residues 10 to 17), which forms direct interactions with the bound GDP (fig. S1) (8, 9). From the modeled structure of the N92I mutant, predicted on the basis of the wild-type structure, an original proposal was that the N92I mutation results in a structural difference in the P-loop region, which then causes a decrease in the binding affinity for GDP and leads to the accelerated GDP dissociation (5). However, because of the lack of high-resolution structural information for the N92I mutant, the detailed mechanism has remained unclear.

Here, we investigated the activation mechanism of the Rac1 N92I oncogenic mutant by using solution NMR spectroscopy. We demonstrated that the P-loop structure is actually not markedly perturbed by the introduction of the N92I mutation. Instead, we found that the structure of the α3 helix is locally perturbed and that the overall conformational stability decreases in the mutant, which facilitates the GDP dissociation by lowering the activation energy for the dissociation reaction. We further revealed that the hydrogen bond interaction, formed between the Asp11 and Trp97 side chains in the wild type, determines the overall conformational stability and the GDP binding affinity and that this interaction is sterically blocked by the introduction of the N92I mutation. On the basis of these results, we proposed the activation mechanism of the N92I mutant, in which the decreased conformational stability plays important roles in its activation process.

RESULTS

The N92I mutant has a reduced affinity for GDP

First, we characterized the GDP dissociation rates in the presence of various concentrations of Mg2+, which is known to be critical for maintaining stable GDP binding (10). The GDP dissociation rates were measured by monitoring the reduction in the fluorescence intensity caused by the dissociation of preloaded 2′-(or-3′)-O-(N-methylanthraniloyl)–GDP (mant-GDP), in the presence of an excess amount of GTP. In the N92I mutant, the dissociation rates of mant-GDP were significantly faster than those of the wild type, in both the presence and absence of Mg2+, and up to 30-fold faster than those of the wild type in the physiological range of Mg2+ concentration (0.2 to 1.2 mM) (Fig. 1A and table S1) (11). Whereas the apparent dissociation constant (Kd) of Mg2+ in the N92I mutant (26.6 ± 3.5 μM) was only threefold weaker than that that of the wild type (8.9 ± 0.7 μM), the mant-GDP dissociation rate in the absence of Mg2+ was about 11-fold faster in the N92I mutant (Fig. 1B). These results indicate that the accelerated GDP dissociation in the N92I mutant is largely caused by the reduced binding affinity for GDP itself rather than the reduced binding affinity for Mg2+. We note that this hyperactivation mechanism is in contrast to that of another oncogenic mutant, the P29S mutant. Recently, we revealed that the wild-type Rac1 exists in a conformational equilibrium between the major substate (A state) with high affinity for Mg2+ and the minor substate with low affinity for Mg2+ (B state), exchanging on the microsecond time scale (7). We showed that, in the P29S mutant, the equilibrium is shifted toward the B state and thus exhibits lower affinity for Mg2+, which is responsible for the accelerated GDP dissociation under physiological conditions. In the case of the N92I mutant, this conformational equilibrium is not markedly different from that in the wild type, as supported by the results that the 13C and 1H relaxation dispersion (RD) profiles were not largely different from those observed in the wild type (fig. S2). These results support the hypothesis that the activation mechanism of the N92I mutant is distinct from that of the P29S mutant.

Fig. 1 Mant-GDP dissociation rates of the wild type and the N92I mutant Rac1.

(A) Mant-GDP dissociation rates of the wild type (left) and the N92I mutant (right). The data depicted as EDTA were measured in the presence of 1 mM EDTA. (B) Apparent Kd of Mg2+ of the wild type (left) and the N92I mutant (right). Kd values were calculated by using the equation, koff,GDP([Mg2+]) = A × Kd/([Mg2+] + Kd) + B, where A and B are constants. Each point reflects means ± SE of three independent experiments. The data for the wild type were originally reported in (7).

Structural differences in the α3 helix of the N92I mutant

To investigate the structural mechanisms underlying the reduced affinity for GDP in the N92I mutant, we prepared u-[2H, 15N]–labeled Rac1 samples and compared the 1H-15N transverse relaxation optimized spectroscopy (TROSY) spectra between the wild type and the N92I mutant (12). In all of the NMR experiments, we introduced the R66E mutation to suppress the self-association. We previously demonstrated that the introduction of the R66E mutation did not largely affect the apparent Kd for Mg2+ and that the chemical shift differences between the wild type and the R66E mutant were observed only for the residues that are in close vicinity to Arg66 (7). When we compared the NMR spectra between the wild type and the N92I mutant, significant chemical shift differences greater than the SD [0.15 parts per million (ppm)] were observed for the residues from the entire α3 helix, on which Asn92 is located, indicating that notable changes in the chemical environments occur on the α3 helix in the N92I mutant (Fig. 2). We also compared the secondary structure populations between the wild type and the N92I mutant, using the backbone chemical shifts with the δ2D software (13). We found that differences in the helix population greater than 2 SDs (0.054) were observed mainly in the residues forming the α3 helix (residues 91 to 96) and that relatively small differences were observed for the residues around the α4 helix as well (residues 145, 149, 150, and 151) (fig. S3). These results indicate that the structural differences between the wild type and the N92I mutant are highly localized to the region around the α3 helix and that these structural differences allosterically perturb the structure of the adjacent α4 helix.

Fig. 2 Chemical shift differences between the wild type and the N92I mutant Rac1.

(A) Plot of the normalized chemical shift differences between the wild type and the N92I mutant, Δδ. Δδ is calculated by the equation, Δδ = {(Δδ1H)2 + (Δδ15N /5)2}0.5. The residues with Δδ values larger than the SD (0.15 ppm) are colored red. (B) Overlay of the 1H-15N TROSY spectra of the wild type (black) and the N92I mutant (red), measured at 11.7 T (1H frequency, 500 MHz) and 25°C. The signals with Δδ values larger than 0.4 ppm are labeled. (C) Mapping of the residues with Δδ values larger than 0.15 ppm on the structure of Rac1 (PDB ID: 1DS6) (8). The residues with no data are colored gray.

Notably, the residues from the P-loop region showed smaller chemical shift differences up to 0.2 ppm than those observed in the α3 helix (Fig. 2), suggesting that the P-loop structure is not markedly perturbed by the introduction of the N92I mutation. To further characterize the differences in the P-loop structure, we measured nuclear Overhauser spectroscopy (NOESY) spectra and compared the results between the wild type and the N92I mutant. The cross peaks were observed between the same atomic pairs in the wild type and the N92I mutant, and the observed NOE patterns were compatible with the crystal structures of the Rac2-RhoGDI complex (PDB ID: 1DS6) and of the uncomplexed Rac1 (PDB ID: 5N6O; figs. S1 and S4, A and B) (8, 9). We further compared the backbone amide 1H-15N residual dipolar couplings (RDCs) between the wild type and the N92I mutant. The observed RDC patterns were not very different between them, and their values were compatible with the calculated RDCs obtained from the crystal structure (fig. S4C). We also found that the 31P chemical shifts of the bound GDP were almost the same between the wild type and the N92I mutant, reflecting that the interactions formed between the bound GDP and the P-loop residues are not largely affected (fig. S4D). Collectively, these results indicate that neither the structure of the P-loop nor the interactions formed between the P-loop and the bound-GDP are markedly different between the wild type and the N92I mutant and that the local structural differences in the α3 helix are responsible for the increased GDP dissociation rate in the N92I mutant.

Decreased conformational stability in the N92I mutant

To characterize how the structural differences in the α3 helix affect the thermal stability and the GDP binding property, we first compared the thermal denaturation temperatures (Tm) between the wild type and the N92I mutant. The Tm values of the wild type and the N92I mutant were 64.8° and 53.6°C, respectively, showing that the conformational stability greatly decreases in the N92I mutant (Fig. 3A). We also measured the Tm value of another oncogenic mutant, the P29S mutant, and found that the Tm value of the P29S mutant (65.0°C) was almost the same as that of the wild type, indicating that the decrease in thermal stability is the characteristic feature of the N92I mutant. We then quantitatively evaluated the difference in the activation free energy of unfolding by measuring the unfolding rate constants in the presence of various concentrations of a denaturant, guanidine hydrochloride (14). By linear extrapolation, we found that the unfolding rate constant of the N92I mutant is about 10.5-fold faster than that of the wild type in the absence of guanidine hydrochloride, which corresponds to a difference in the activation free energy of unfolding, ΔΔGunfold, of 1.39 kcal/mol (Fig. 3B). Notably, this free energy difference was consistent with the difference in the activation free energy of GDP dissociation, ΔΔGGDP,off, of 1.22 kcal/mol. This ΔΔGGDP,off value was calculated from the differences in the mant-GDP dissociation rates between the wild type and the N92I mutant in the absence of Mg2+ to estimate the energetic contributions to the GDP binding free energy and not to the Mg2+ binding free energy. These results strongly suggest that the Rac1 unfolding and the GDP dissociation events share part of their energetic landscapes and that the decreased conformational stability lowers the activation energy barrier for the GDP dissociation reaction. In these energetic landscapes, note that the conformational substates with different affinities for Mg2+ (states A and B), which are exchanging on the microsecond time scale, are not markedly altered by the introduction of the N92I mutation (Fig. 3C) (7). We also found that the mant-GDP dissociation rate greatly increased in the presence of low concentrations of guanidine hydrochloride, which facilitates the unfolding of wild-type Rac1 (fig. S5).

Fig. 3 Conformational stability of Rac1.

(A) Thermal denaturation curves of the wild type (black), the P29S mutant (green), and the N92I mutant (red), based on circular dichroism measurements observing the molar ellipticity at 222 nm. (B) Plots of the unfolding rates of the wild type (black) and the N92I mutant (red), measured in the presence of various concentrations of guanidine hydrochloride. (C) Schematic representations of the free energy landscape of the GDP binding reaction (left) and the folding reaction (right). The free energy landscapes of the wild type and the N92I mutant are colored black and red, respectively. The A and B states are the previously identified substates in the GDP-bound state, which differ in the affinity for Mg2+ (7).

The α3 helix–P-loop interaction determines the conformational stability

To identify the key structural elements that play critical roles in determining the overall conformational stability and the GDP binding affinity, we analyzed the temperature-dependent changes in the amide NH and Trp ε-NH chemical shifts, which not only define the intramolecular hydrogen bonding but also serve as sensitive reporters for both the local and global structural stabilities of proteins (1518). We measured 1H-15N TROSY spectra at 20° to 35°C and calculated the 1H chemical shift changes per degree of temperature, hereafter referred to as the temperature coefficient Δδ/ΔT. We then compared the Δδ/ΔT values of the wild type with those of the N92I mutant (Fig. 4A and fig. S6). The Δδ/ΔT values of the wild type and the mutants were very similar for most of the residues, indicating that the structural reorganizations in the N92I mutant are restricted to local regions. The notable decrease in Δδ/ΔT, up to −5.3 parts per billion/°C, was observed in the side chain of Trp97 in the N92I mutant, indicating that the hydrogen bond interaction involving the side chain of Trp97 is lost in the mutant. Small but significant differences in Δδ/ΔT were also observed in the main chains of Ala95, Lys96, and Trp97, reflecting the stretching or shortening of hydrogen bonds caused by the local structural differences of the α3 helix. In the crystal structures (PDB IDs: 1DS6 and 5N6O), the side-chain ε-NH of Trp97, located on the α3 helix, forms a hydrogen bond with the carboxyl oxygen of Asp11, located on the P-loop region (Fig. 4B and fig. S1). To test whether the loss of the hydrogen bond formed between the Asp11 and Trp97 side chains is responsible for the accelerated GDP dissociation, we prepared a W97A mutant protein. We confirmed that the chemical shift differences between the wild type and the W97A mutant were mainly observed for the residues in the close vicinity of Trp97 and that the overall structure is minimally affected by the introduction of the W97A mutation (fig. S7, A to C). Notably, when we measured its mant-GDP dissociation rate, the mant-GDP dissociation rate markedly increased in the W97A mutant, as compared to the mutants of other residues on the α3 helix (Fig. 4, B and C, and table S2). In addition, the Tm value of the W97A mutant was 50.0°C, as compared to 64.8°C in the wild type, showing that the conformational stability greatly decreased in the W97A mutant (fig. S7D). These results indicate that the side chain of Trp97 is a critical factor that determines the stability of GDP binding. As stated above, the side chain of Trp97 participates in the hydrogen bond with the side chain of Asp11. Thus, these results strongly suggest that the loss of this hydrogen bond in the N92I mutant is responsible for the decreased overall conformational stability in the mutant.

Fig. 4 Experimental identification of the key hydrogen bond interaction that determines the conformational stability.

(A) Plots of the amide and Trp ε-NH 1H temperature coefficients, Δδ/ΔT, of the wild type (top) and the N92I mutant (middle). The plot of the differences in Δδ/ΔT between the wild type and the N92I mutant is also shown (bottom). ppb, parts per billion. (B) Close-up view of the interactions formed between the α3 helix and the switch 2 region (PDB ID: 1DS6) (8). (C) Mant-GDP dissociation rates of the mutants of the residues that participate in the interaction formed between the α3 helix and the switch 2 region. (D) Mant-GDP dissociation rates of the Asn92 mutants. (E) Correlation plots of the mant-GDP dissociation rates of the Asn92 mutants versus the side-chain volumes (40).

We also investigated the structural mechanism for the loss of the Asp11-Trp97 hydrogen bond in the N92I mutant by replacing Asn92 with other amino acids with different properties (Fig. 4D and table S3). When we mutated Asn92 to charged amino acids, the mant-GDP dissociation rates changed slightly. In contrast, when we replaced Asn92 with hydrophobic amino acids with different steric bulkiness, the mant-GDP dissociation rate increased as the side-chain volume increased, indicating that the formation of the Asp11-Trp97 hydrogen bond is hampered by the steric hindrance with the side chain of the residue at position 92 (Fig. 4E). This proposed mechanism is also supported from the molecular dynamics simulations of Rac1 (Fig. 5). During the 50-ns simulation using the wild-type structure, the hydrogen bond formed between the Trp97 ε-NH and the carboxyl oxygen of Asp11 remained stable. In contrast, when we conducted the simulations using the N92I mutant structure, the distance between the Trp97 ε-NH and the carboxyl oxygen of Asp11 was mainly distributed around 4 Å, indicating that the hydrogen bond is not stably formed between these atoms. The trajectory of the N92I mutant revealed that the Asp11 side-chain conformer exclusively adopts the trans conformation because of steric hindrance with the γ2 methyl group of Ile92. In the trans conformation, the carboxylic group of Asp11 is further away from the side chain of Trp97, and thereby, the hydrogen bond cannot be formed between them.

Fig. 5 Molecular dynamics simulations of Rac1.

(A) Close-up view of the wild-type Rac1 structure obtained after the 50-ns simulation (left). Plots of the distances between Asn92Hδ-Asp11Oδ1, Trp97Hε-Asp11Oδ1, and Trp97Hε-Asp11Oδ2 (center). Histogram of the distribution of Asp11 side-chain χ1 angles (right). (B) Close-up view of the N92I mutant Rac1 structure obtained after the 50-ns simulation (left). Plots of the distances between Trp97Hε-Asp11Oδ1 and Trp97Hε-Asp11Oδ2 (center). Histogram of the distribution of Asp11 side-chain χ1 angles (right).

DISCUSSION

In this study, we investigated the activation mechanism of the oncogenic N92I mutant of Rac1. We found that, whereas the structure of the P-loop is not markedly perturbed, the structure of the α3 helix is locally perturbed, and the overall conformational stability decreases in the N92I mutant. We also revealed that the hydrogen bond interaction formed between the Asp11 and Trp97 side chains determines the overall conformational stability and that the replacement of Asn92 with Ile sterically blocks this interaction and thus substantially decreases the overall conformational stability. Considering that both the decrease in the conformational stability and the loss of the hydrogen bond interaction in the N92I mutant were observed in the GDP-bound state, we propose that the decreased conformational stability increases the ground-state free energy of the GDP-bound state, and hence, lowers the activation energy along the GDP dissociation reaction coordinate, leading to faster GDP dissociation in the mutant (Fig. 6). This mechanism is consistent with the observation that, in the structure of the GDP-bound Rac1, the bound GDP forms an extensive hydrogen bond network that stabilizes the ternary complex and the dissociation of GDP is expected to involve the relaxation of the Rac1 structure (8). Several lines of evidence have shown that the nucleotide-free states of G proteins represent partially unfolded structures and that the GDP dissociation occurs via partially unfolded intermediates (1921). We note that our conclusion is in contrast to the originally proposed mechanism, in which GDP binding is affected in the N92I mutant through an effect on the P-loop structure by the loss of the hydrogen bond interaction formed between the side chains of Asp11 and Asn92 in the wild type (5). We presume that the interaction formed between Asp11 and Asn92 does not greatly contribute to maintaining the P-loop structure in the wild type because the simple loss of this interaction in the N92A mutant hardly affected the mant-GDP dissociation rate (Fig. 4D) and that the P-loop structure is highly preserved in the structural model of the N92I mutant despite the absence of this interaction (Fig. 5). The decrease in the conformational stability would also affect the GTP dissociation rate; however, the difference in the GTP dissociation rate is expected to be smaller than that observed in the GDP dissociation rate. This is because the GTP dissociation rate is about 25-fold slower than the GDP dissociation rate (22) and because the activation energy barrier is higher for the GTP dissociation reaction than that for the GDP dissociation. This notion is fully supported by the previous observation that the GTP dissociation rates were not markedly different between the wild type and the N92I mutant (5).

Fig. 6 Schematic illustration of the decreased conformational stability in the N92I mutant and its activation mechanism.

In the wild type (black), a hydrogen bond interaction (blue line) is formed between Asp11 and Trp97. In the N92I mutant (red), the side chain of Ile92 sterically blocks the hydrogen bond interaction (magenta circle). The loss of this hydrogen bond in the N92I mutant increases the GDP-bound ground-state energy and hence accelerates the GDP dissociation by lowering the activation energy along the GDP dissociation reaction coordinate.

Our results have established that the local structural differences around the α3 helix, which is distant from the GDP binding pocket, markedly alter the binding affinity for GDP by affecting its overall conformational stability. This mechanism indicates that, in the N92I mutant, the acceleration of the GDP dissociation and the subsequent stabilization of the GTP-bound state are achieved in an allosteric manner without perturbing the structures and dynamics of the P-loop and the switch regions. Because the interactions with downstream effector proteins occur exclusively on the switch regions, and not on the α3 helix (2326), the interactions with the effector proteins are expected to be highly preserved in the N92I mutant. This property would explain the strong oncogenic activity of the N92I mutant, as previously observed in cellular experiments (5).

Recently, a growing amount of evidence has shown that the formation of unfolded states plays important roles in a wide variety of biological processes, such as ligand binding and enzymatic reactions (27, 28). Our findings have established that the decrease in the conformational stability can be induced by disease-related mutations and that the aberrant functions of proteins caused by the decreased stability actually play critical roles in pathogenic processes. Therefore, the present study may open the possibility of designing drugs that can modulate the activities of proteins by regulating their conformational stability.

METHODS

Protein expression and purification

The Rac1 protein (human Rac1, residues 1 to 184) was expressed in Escherichia coli cells as a glutathione-S-transferase (GST) fusion protein, as described previously (5, 7). To suppress the self-association into a dimer at concentrations more than 102 μM, we used an R66E mutant in all of the NMR experiments (7). The Rac1 mutants were constructed with the QuikChange II Site-Directed Mutagenesis Kit (Agilent Technologies). E. coli BL21-CodonPlus(DE3)-RP cells (Agilent Technologies), transformed with the pGEX-6P-1 plasmid encoding Rac1, were grown at 37°C in LB media for preparing nonlabeled samples or in deuterated M9 media for preparing isotopically labeled NMR samples. When 15N labeling was required, 15NH4Cl was used as the sole nitrogen source. For the selective 13CH3 labeling of methyl groups, [3-13C, 2-2H]-l-alanine (200 mg/liter, for Alaβ), [methyl-13C, 3,3-2H2]-α-ketobutyric acid [50 mg/liter, for Ileδ1; Cambridge Isotope Laboratories (CIL)], [3-methyl-13C, 3,4,4,4-2H4]-α-ketoisovaleric acid (80 mg/liter, for Leuδ1, Leuδ2, Valγ1, and Valγ2; CIL), [methyl-13C]-l-methionine (100 mg/liter, for Metε; CIL), and [2,2,3,3-2H4]-succinic acid (2.5 g/liter; CIL) were added 1 hour before the induction. The production of the Rac1 protein was induced with 0.1 mM isopropyl β-d-1-thiogalactopyranoside at 20°C for 16 hours. The protein was purified to homogeneity by chromatography on Glutathione Sepharose 4B resin (GE). After cleavage of the GST-tag with PreScission protease (GE), the protein was further purified by size exclusion chromatography using the HiLoad 26/60 Superdex 75 pg (GE) in buffer containing 20 mM tris (pH 7.5), 150 mM NaCl, 5 mM MgCl2, and 1 mM dithiothreitol (DTT). To observe the amide signals, the purified protein was incubated at 37°C for 36 hours in MgCl2-free buffer to exchange the amide hydrogen atom from 2H to 1H. NMR samples were prepared by changing the buffer to 20 mM Hepes-NaOH (pH 7.0), 0.5 mM GDP, 5 mM MgCl2, and 5 mM DTT by sequential dilution and concentration with an Amicon Ultra Centrifugal Filter Unit, NMWL (nominal molecular weight limit) 10 kDa (Merck Millipore).

Mant-GDP dissociation rate measurements

The dissociation rate of mant-GDP was measured by monitoring the reduction in the fluorescence intensity due to the dissociation of mant-GDP preloaded in Rac1 (10). The purified Rac1 protein (20 μM) was incubated at 30°C for 30 min in buffer containing 50 mM tris-HCl (pH 7.5), 100 mM NaCl, 0.1 mM MgCl2, 1 mM DTT, and 50 μM mant-GDP (Invitrogen). The dissociation reaction was initiated by a 100-fold dilution in buffer containing 200 μM GTP and various concentrations of MgCl2. The excitation and emission wavelengths were set to 355 and 448 nm, respectively, with slit widths of 5 nm. The measurements were performed at 30°C. The difference in the activation free energy was calculated according to ΔΔGGDP,off = –RTln(koff,WT/koff,N92I), in which R denotes the gas constant and T represents the absolute temperature.

Circular dichroism measurements

The Tm values were obtained from the temperature-dependent changes in the molar ellipticity. The purified Rac1 protein (10 μM) was prepared in buffer containing 50 mM tris-HCl (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 10 μM GDP, and 1 mM DTT, and its molar ellipticity at 222 nm was monitored. The temperature was changed from 30° to 90°C in 1.5°C increments, with 30 s of equilibration time. The Tm values were obtained by fitting the temperature-dependent changes in molar ellipticity by using the Gibbs-Helmholtz equation (29).

Measurements of unfolding rates

The unfolding rate of Rac1 was measured by monitoring the reduction in the tryptophan fluorescence caused by protein unfolding (14). The purified Rac1 protein (1 μM) was prepared in buffer containing 50 mM tris-HCl (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 5 μM GDP, and 1 mM DTT. Then, the protein solution was rapidly mixed with an equivalent volume of buffer containing various concentrations of guanidine hydrochloride, and the time-dependent changes in the fluorescence were monitored. The unfolding rate under native conditions was determined by linear extrapolation to 0 M guanidine hydrochloride. The excitation and emission wavelengths were set to 281 and 340 nm, respectively, with slit widths of 5 nm. The measurements were performed at 25°C. The difference in the activation free energy was calculated according to ΔΔGunfold = –RTln(kunfold,WT/ kunfold,N92I), in which R denotes the gas constant and T represents the absolute temperature.

NMR analyses

All 1H-detected experiments were performed on Bruker AVANCE 500, 600, or 800 spectrometers equipped with cryogenic probes. 31P-detected experiments were performed on a Bruker AVANCE 400 spectrometer equipped with a BBFO probe. All spectra were processed by the Bruker TopSpin 2.1 or 3.1 software, and the data were analyzed using Sparky (T. D. Goddard and D. G. Kneller, Sparky 3, University of California, San Francisco, CA). Backbone resonance and side-chain methyl resonance assignments were obtained by combining out-and-back type triple-resonance experiments (30), NOE analyses based on the crystal structure, and mutagenesis approaches.

[1H-1H] NOESY-[15N-1H]-TROSY spectra were obtained using u-[2H, 15N] Rac1 samples at 25°C with a Bruker AVANCE 600 spectrometer. The mixing time was set to 200 ms.

The 1H-15N RDCs were obtained from the difference in the 15N-1H J-coupling constants in aligned (pentaethylene glycol monododecyl ether/hexanol) and isotropic media. The J-coupling constants were measured from the differences between the TROSY and heteronuclear single-quantum coherence peak positions. The measurements were performed at 20°C with a Bruker AVANCE 600 spectrometer, using u-[2H, 15N] Rac1 samples. The alignment tensors and the calculated RDCs were obtained by singular value decomposition with the dipolar coupling computation server developed by the Ad Bax group (National Institutes of Health) (https://spin.niddk.nih.gov/bax/nmrserver/dc/svd.html) by using the coordinates of the Rac1 structure modeled from the crystal structure of the Rac2-RhoGDI complex (PDB ID: 1DS6). In these analyses, the region with structural changes upon the binding of RhoGDI (residues 34 to 42) was not used.

To obtain the backbone 1H (amide NH), 13C (Cα, Cβ, and CO), and 15N (amide N) chemical shifts for δ2D analyses (13), we prepared u-[13C, 15N] Rac1 samples and conducted triple-resonance experiments at 35°C (for the N92I mutant) or 40°C (for the wild type) with a Bruker AVANCE 500 spectrometer.

The 13C single-quantum (SQ) and 1H triple-quantum (3Q) Carr-Purcell-Meiboom-Gill (CPMG) RD analyses were recorded at 25°C with Bruker AVANCE 600 and 800 spectrometers, using {u-[2H, 15N]; Alaβ, Ileδ1, Metε-[13CH3], Leu, and Val-[13CH3, 12C2H3]} Rac1 samples (31, 32). The constant-time CPMG relaxation period T was set to 20 ms for the 13C SQ experiments and 5 ms for the 1H 3Q experiments. The νCPMG values were varied between 50 and 1500 Hz for 13C SQ and between 200 and 3000 Hz for 1H 3Q RD experiments. The values of the effective relaxation rates measured in the presence of a νCPMG Hz CPMG pulse train, R2,effCPMG), were calculated using Eq. 1, where ICPMG) and I (0) represent the peak intensities with and without the relaxation period T, respectively.R2,eff(νCPMG)=1Tln{I(νCPMG)I(0)}(1)

For residue-specific fitting, the 13C SQ and 1H 3Q RD curves obtained with two static magnetic fields were simultaneously fitted to the Luz-Meiboom equation (Eq. 2) (33), where ΦX (X = 13C or 1H) denotes the dispersion amplitude parameter, R2,0 denotes the intrinsic transverse relaxation rate, kex denotes the exchange rate, ΔωX (X = 13C or 1H) denotes the chemical shift difference, and pB denotes the B state population (7)R2,eff=R2,0+ΦXkex(14νCPMGkextanh(kex4νCPMG))ΦC=pB(1pB)ΔωC2ΦH=pB(1pB)(3ΔωH)2(2)

Backbone amide NH and Trp ε-NH 1H temperature coefficients were calculated from the 1H-15N TROSY spectra obtained at 293, 298, 303, and 308 K with a Bruker AVANCE 500 spectrometer, using u-[2H, 15N] Rac1 samples.

Molecular dynamics simulations

The all-atom model of Rac1 was constructed with CHARMM-GUI (34), using the crystal structure of the Rac2-RhoGDI complex (PDB ID: 1DS6), on the basis of our previous result that this structure well represents the ground-state structure of GDP-bound Rac1 in the uncomplexed state (7, 8). The Rac2 sequence was converted to that of Rac1 with or without the N92I mutation, and the structure including the bound GDP and Mg2+ was solvated in a periodic water box with a size of 74 Å by 74 Å by 74 Å, using TIP3P water molecules. The system was then neutralized with about 150 mM KCl. Molecular dynamics simulations were performed using the NAMD2.12 software (35) and the CHARMM36 parameter set (36). A cutoff distance of 12 Å was used for the van der Waals and short-range electrostatic interactions, and the long-range electrostatic interactions were computed with the particle mesh Ewald method (37). Bonds containing hydrogen atoms were restrained with the SHAKE algorithm (38). Bonded and short-range electrostatic interactions were computed every 2 fs, and long-range electrostatic interactions were computed every 4 fs. The system was first energy-minimized for 500 steps and then equilibrated under isothermal-isobaric (NPT) conditions for 5 ns at 1 atm and 310 K. The production simulations were performed for 50 ns at 1 atm and 310 K. The trajectories were analyzed using the Visual Molecular Dynamics 1.9.3 software (39).

SUPPLEMENTARY MATERIALS

Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/8/eaax1595/DC1

Table S1. MANT-GDP dissociation rates of the wild type and the N92I mutant in the presence of various concentrations of Mg2+.

Table S2. MANT-GDP dissociation rates of the mutants of the residues from the α3 helix and the switch 2 region.

Table S3. MANT-GDP dissociation rates of the mutants of Asn92.

Fig. S1. Comparisons of the structures of Rac between the RhoGDI-bound state and the uncomplexed state.

Fig. S2. Microsecond-order conformational exchange processes in the wild type and the N92I mutant.

Fig. S3. Secondary structure populations of the wild type and the N92I mutant.

Fig. S4. Structural characterizations of the P-loop.

Fig. S5. Mant-GDP dissociation rates in the presence of guanidine hydrochloride.

Fig. S6. Temperature-dependent changes in the 1H-15N TROSY spectra.

Fig. S7. Characterizations of the structure and conformational stability of the W97A mutant.

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

REFERENCES AND NOTES

Acknowledgments: Funding: This work was supported in part by the Japan Agency for Medical Research and Development (AMED) under grant number JP18ae0101046 and the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/Japan Society for the Promotion of Science KAKENHI under grant number JP17H06097 (both to I.S.). Author contributions: Y.T. performed the experiments. Y.T., K.K., T.K., and I.S. designed the study, analyzed the data, and wrote the paper. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors. The backbone and methyl resonance assignments for Rac1 have been deposited in the Biological Magnetic Resonance Bank with the accession number 27577.
View Abstract

Stay Connected to Science Advances

Navigate This Article