The Escherichia coli GTPase CgtA_E Cofractionates with the 50S Ribosomal Subunit and Interacts with SpoT, a ppGpp Synthetase/Hydrolase

Abstract

CgtA_E/Obg_E/YhbZ is an Escherichia coli guanine nucleotide binding protein of the Obg/GTP1 subfamily whose members have been implicated in a number of cellular functions including GTP-GDP sensing, sporulation initiation, and translation. Here we describe a kinetic analysis of CgtA_E with guanine nucleotides and show that its properties are similar to those of the Caulobacter crescentus homolog CgtA_C. CgtA_E binds both GTP and GDP with moderate affinity, shows high guanine nucleotide exchange rate constants for both nucleotides, and has a relatively low GTP hydrolysis rate. We show that CgtA_E is associated predominantly with the 50S ribosomal subunit. Interestingly, CgtA_E copurifies with SpoT, a ribosome-associated ppGpp hydrolase/synthetase involved in the stress response. The interaction between CgtA_E and SpoT was confirmed by reciprocal coprecipitation experiments and by two-hybrid assays. These studies raise the possibility that the ribosome-associated CgtA_E is involved in the SpoT-mediated stress response.

The Obg/GTP1 subfamily is a distinct group of monomeric GTP-binding proteins originally characterized in bacteria but found in the genomes of all organisms sequenced thus far. Bacterial proteins of the Obg subfamily are typically essential for cell growth (26, 34, 47, 61), although their precise function is unknown. Based on the evolutionary relationship between these and other GTP-binding proteins, it has been proposed that Obg GTPases act as translation factors (28). This is certainly the case for Nog1p, a eukaryotic Obg protein that is associated with pre-60S ribosomal complexes (4, 12, 14, 18, 22, 40, 53) and is critical for 60S assembly (25, 54). Although a direct role in translation has not been assigned to the bacterial Obg proteins, accumulating evidence suggests that these proteins are ribosome associated and play roles in stress response and in stationary-phase survival.

The initiation of sporulation in Bacillus subtilis and Streptomyces spp. is correlated with changes in cellular GTP-GDP pools (33, 42, 44, 45). We, and others, have proposed that the Obg proteins act as sensors of these pools (29, 48). Overproduction of Obg, which leads to high levels of Obg-GTP, prevents spore development in Streptomyces spp., suggesting that levels of Obg-GDP are critical for this process (47, 48). Increasing the levels of Obg-GDP in Streptomyces coelicolor by decreasing the GTP/GDP ratio restores spore development (48). Similarly, the S. coelicolor obgP168V mutant, which is predicted to be predominantly in the GTP-bound state, inhibits sporulation (48). In B. subtilis, depletion of Obg also results in a sporulation defect (63). The Caulobacter crescentus Obg protein CgtA_C binds guanine nucleotides with modest affinity and displays rapid guanine nucleotide exchange but relatively slow hydrolysis (29). Together with the sporulation data, these biochemical features are consistent with a model whereby the nucleotide-bound state of the Obg protein is determined by GTP-GDP pools and, in turn, affects stationary-phase survival. This model is complicated, however, by the discovery that the B. subtilis Obg protein cocrystallized with the alarmone ppGpp (6). Interestingly, the ppGpp pools increase as cells enter stationary phase (43, 45).

The bacterial Obg proteins have also been implicated in ribosome function. The C. crescentus CgtA_C protein cofractionates with the 50S ribosomal subunit (32). The B. subtilis Obg protein also cofractionates with ribosomal proteins and interacts specifically with the 50S ribosomal subunit protein L13 (56). Evidence for a role of the Escherichia coli Obg protein CgtA_E (also called YhbZ or Obg_E) in ribosome function comes from genetic interactions of CgtA_E with RrmJ (also called FtsJ), a methyltransferase that modifies uridine 2552 of the 23S rRNA (7). The ΔrrmJ mutant grows slowly, has a significantly altered polysome profile, displays an increase in unassembled 30S and 50S ribosomal subunits, and shows a decrease in 70S ribosomes and polyribosomes (5, 8). Overexpression of CgtA_E suppresses both the growth defect and the polysome profile defect of the ΔrrmJ mutant, suggesting that CgtA_E plays an active role in ribosome assembly or stability (60). Interestingly, CgtA_E was not associated with ribosomes separated on standard sucrose gradients (26).

In this study, we characterized the guanine nucleotide binding, exchange, and GTP hydrolysis kinetics of the E. coli CgtA_E protein and show that they are similar to those of the C. crescentus CgtA_C protein. Further, CgtA_E was found to be in a large RNA-containing complex that was associated predominantly with the 50S ribosomal subunit. Interestingly, CgtA_E copurified with SpoT, a ppGpp synthase/hydrolase important in the stress response. The interaction with SpoT was confirmed by reciprocal coprecipitation and yeast two-hybrid studies, raising the possibility that these proteins are involved in the same cellular function.

MATERIALS AND METHODS

Cell growth and plasmid construction.

E. coli cells were grown at 37°C (unless otherwise indicated) in Luria-Bertani broth (LB; 10 g of tryptone, 5 g of yeast extract, 10 g of NaCl/liter) or LB agar containing 100 μg of ampicillin/ml or 30 μg of kanamycin/ml, as required. Saccharomyces cerevisiae was grown at 30°C in synthetic dropout (SD) medium (57) lacking Trp, Leu, or both, as indicated.

E. coli genes were amplified using colony PCR with E. coli W3110 cells (Table 1) as the source of template DNA and Advantage cDNA polymerase (Clontech) on a PTC-100 programmable thermal controller (MJ Research, Inc.) with the oligonucleotides listed in Table 2. The two-hybrid clones were generated by PCR amplification (with the primers listed in Table 2), and the relevant DNA was cloned into the TOPO shuttle vector and subcloned into pAS2-1 or pACT2, as indicated (Table 1). The identities of the clones were confirmed by restriction mapping and verified by dideoxy sequencing.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Genotype	Reference or source
Strains
E. coli
W3110	Wild type	ATCC 27325
BL21(DE3)	F−ompT hsdSB (rB− mB−) gal dcm (DE3)	Novagen
DH5α	λ− φ80d/lacΔM15 Δ(lacZYA-argF) U169 recA1 endA1 hsdR17 (rk− mk+) supE44 thi-1 gyrA relA1	Gibco-BRL
S. cerevisiae
Y187	MATα ura3-52 his3-200 ade2-101 trp1-901 leu2-3,112 gal4Δ met−gal80Δ URA3::GAL4UAS-GAL1TATA-lacZ	19
Y190	MATaura3-52 his3-200 lys2-801 ade2-101 trp1-901 leu2-3, -112 gal4Δ gal80Δ cyh-2 LYS::GALIUAS-HIS3TATA-HIS3 URA3::GAL1UAS- GAL1TATA-lacZ	19
Plasmids
pET28a(+)	Expression vector for N-terminal His-tagged fusions; Kan	Novagen
2.1-TOPO	Cloning vector for PCR products; Amp Kan	Invitrogen
pAS2-1	GAL4-binding domain; TRP1 Amp	Clontech
pACT2	GAL4-activating domain; LEU2 Amp	Clontech
pMAL-c	MBP expression vector	New England Biolabs
pMB24	DsbC-His expression vector	17
pMBPSpoT	MBP-SpoT in pMAL-c	J. Hernandez
pJM1138	cgtAE PCR product from YhbZ3 and YhbZ5 in pET28a(+)	This study
pKP10	cgtAE(FL) PCR product from YhbZ-2 and YhbZ-3 in TOPO	This study
pKP11	NdeI-BamHI cgtAE from pKP10 in pAS2	This study
pKP12	cgtAE(FL) PCR product from YhbZ-6a and YhbZ-2 in TOPO	This study
pKP13	SmaI-BamHI cgtAE from pKP12 in pACT2	This study
pKP22	spoT(FL) PCR product from SpoT-1 and SpoT-2 in TOPO	This study
pKP23	NcoI-BamHI spoT(FL) from pKP22 in pACT2	This study
pKP24	NcoI-BamHI spoT(FL) from pKP23 in pAS2	This study
pKP25	spoT(N) PCR product from SpoT-1 and SpoT-3 in TOPO	This study
pKP26	NcoI-BamHI spoT(N) from pKP25 in pACT2	This study
pKP27	NcoI-BamHI spoT(N) from pKP25 in pAS2	This study
pKP31	spoT(C) PCR product from SpoT-6 and SpoT-2 in TOPO	This study
pKP32	NcoI-BamHI spoT(C) from pKP31 in pACT2	This study
pKP33	NcoI-BamHI spoT(C) from pKP33 in pAS2	This study

TABLE 2.

Oligonucleotides used in this study

Name	Relevant gene	Sequence
YhbZ-5	cgtAE	5′AGAATCATATGAAGTTTGTTGATGAAGCA3′
YhbZ-3	cgtAE	5′ACAAACTCGAGCAAATTCAACATATTGCA3′
YhbZ-2	cgtAE	5′TCATCAGGGATCCACGCTTGTAAAT3′
YhbZ-3	cgtAE	5′GAGAATCATATGAAGTTTGTTGATGAA3′
YhbZ-6a	cgtAE	5′GGACCCGGGAATGAAGTTTGTTGATGAA3′
SpoT-1	spoT	5′GCGCCATGGCCTTGTATCTGTTTGAAAGCC3′
SpoT-2	spoT	5′TTCGGATCCCATTAATTTCGGTTTCGGGTG3′
SpoT-3	spoT	5′CTGGATCCATTCAAACGAACTACC3′
SpoT-6	spoT	5′ATAACCATGGTTACCGCTCCGGGCGCT3′

Purification of E. coli CgtA_E protein.

His-tagged CgtA_E protein was overexpressed and purified from E. coli BL21(DE3) cells (Table 1) containing pJM1138 as previously described (58), with the exception that cultures were grown at 30 instead of 37°C and that 1 mg of lysozyme/ml was added to the cell suspension and incubated for 30 min on ice prior to lysis with a French press. In addition, the lysis buffer contained 20 mM, not 10 mM, imidazole, and protein was eluted by a linear gradient of 20 to 300 mM imidazole. The mass of the purified protein was determined by matrix-assisted laser desorption ionization (MALDI)-time of flight mass spectrometry (Protein and Carbohydrate Structure Facility, University of Michigan).

Guanine nucleotide binding studies.

The guanine nucleotide binding ability of CgtA_E was confirmed, and the guanine nucleotide dissociation rate constants and single turnover hydrolysis rate constants of nucleotides were determined using the fluorescent GDP and GTP analogs 2′ or 3′ mant-GDP and -GTP (mGDP and mGTP, respectively) as previously described (29, 58). The equilibrium binding constant (K_D) was determined using [α-³²P]GTP as described previously (29). The binding buffer for all assays was as follows: 10% glycerol, 50 mM Tris-Cl (pH 8.0), 50 mM KCl, 2 mM dithiothreitol, and 10 μM ATP. MgCl₂ was added to a final concentration of 5 to 12 mM unless otherwise indicated; samples without exogenous MgCl₂ also contained 1 mM EDTA. Data were curve fitted with GraphPad Prism version 3.00 for Windows (GraphPad Software, San Diego, Calif.) or Kaleidagraph 3.09 (Synergy Software). Equilibrium binding constants were determined by fitting the data to a hyperbolic binding function. For dissociation rate constant determinations, the curves were fitted to a single exponential decay equation. To obtain the rate constant of hydrolysis, the decrease in fluorescence was fitted to a single exponential decay equation of the form F = A₀ + A_0e^−kt; the half-life of hydrolysis, T_1/2, is calculated as ln2/k.

Gel filtration.

His-tagged CgtA_E was purified by Ni-nitrilotriacetic acid (NTA) chromatography (as described above), and 0.8 mg was loaded directly onto a 100-ml (1.5- by 70-cm) Toyopearl HW 55S (TosoHaas) column with a flow rate of 0.4 ml/min. When indicated, the sample was treated with 1 mg of RNase A/ml for 2 h at room temperature prior to loading on the column. Fractions of 1.5 ml beginning at 20 ml of elution were collected and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on an SDS-10% polyacrylamide gel, followed by silver staining.

Copurification.

The BL21(DE3) E. coli strain was used as the host strain for all protein fusion constructs, which include pJM1138 (His-CgtA_E), pMB24 (DsbC-His), pMBPSpoT (maltose binding protein [MBP]-SpoT), and pMAL-c (MBP). Cells were grown in LB containing 0.2% glucose and the appropriate antibiotic. One-liter cultures were grown to an optical density at 600 nm (OD₆₀₀) of 0.4 to 0.6, and protein expression was induced with 0.25 mM isopropyl-β-d-thiogalactopyranoside (IPTG). Cells expressing MBP fusions were induced at 23°C for 3 h. Cells expressing histidine-tagged fusions were induced for 1 h. Preceding inductions, cells were collected at 2,400 × g for 20 min and then resuspended in 10 ml of binding buffer (100 mM NaCl, 50 mM NaPO₄ [pH 8.0], 1 mM phenylmethylsulfonyl fluoride) with the addition of 10 mM imidazole when preparing for Ni-NTA column purification. Cells were lysed by two passages through a French pressure cell, and cleared lysates were obtained by centrifugation at 20,000 × g for 45 min.

Lysates from cells expressing either His-CgtA_E or DsbC-His (2 ml) were incubated with an equal volume of lysate from cells expressing SpoT-MBP for 1 h at 4°C with constant mixing, after which 250 μl of prewashed Ni-NTA resin was added and the lysates were incubated with mixing for another hour. The column was then placed upright, and unbound protein was allowed to flow through. The column was then washed sequentially with 1 ml of binding buffer, 1 ml of wash buffer 1 (binding buffer with 20 mM imidazole), and 1 ml of wash buffer 2 (binding buffer with 40 mM imidazole). Finally the protein was eluted with 1.5 ml of elution buffer (binding buffer with 150 mM imidazole).

Lysates from cells expressing either MBP-SpoT or MBP alone were incubated with an equal volume of lysate from cells expressing CgtA_E-His as described above with the exception that 250 μl of prewashed amylose resin was used, the resin was washed with 3 ml of binding buffer, and protein was eluted with 1 ml of binding buffer containing 10 mM maltose.

Proteins were separated by SDS-PAGE and stained with Coomassie blue or electroblotted onto nitrocellulose membranes for analysis by immunoblotting. Blots were blocked with 10% skim milk in phosphate-buffered saline-Tween overnight at 4°C. Primary antibodies were incubated with membranes in the presence of 5% skim milk in phosphate-buffered saline-Tween for 1 h at room temperature at dilutions of 1:1,000 for anti-MBP (U.S. Biologicals) and 1:2,000 for anti-CgtA_E and anti-His (Sigma). Detection of primary antibody was conducted by addition of horseradish peroxidase-conjugated anti-rabbit antibody (Pierce, 1:20,000 dilution) for the CgtA_E antibody or rabbit anti-mouse antibodies (Sigma, 1:10,000 dilution) for both the His and the MBP antibodies. Antibodies were detected by fluorography with ECL (Amersham Pharmacia Biotech) as recommended by the manufacturer.

Preparation of ribosomal particles.

E. coli ribosomal particles were prepared as described previously (13) with the following alterations. E. coli BL21(DE3) cells containing pJM1138 were grown to an OD₆₀₀ of 0.5 in the absence of antibiotic. The cell pellet was resuspended in lysis buffer consisting of 10 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 1.0 mM EDTA, 6 mM 2-mercaptoethanol, 30 mM NH₄Cl, 100 μg of chloramphenicol/ml, and 1 mg of lysozyme/ml. After two rounds of freeze-thaw, deoxycholate, DNase I, and dithiothreitol were added, as described previously (13). Thirteen OD₂₆₀ units of cleared lysate was loaded onto 15 to 45% sucrose gradients (10 ml; 10 mM Tris-HCl [pH 7.5], 10 mM MgCl₂, and 100 mM NH₄Cl), and the gradients were centrifuged in a Beckman SW41 Ti rotor at 41,000 rpm for 3 h. The gradients were fractionated as described previously (32) and analyzed at 254-nm UV absorbance. Each fraction (500 μl) was precipitated in the presence of 15% trichloroacetic acid and 0.03% deoxycholate, resuspended in SDS loading buffer, and analyzed by SDS-PAGE and immunoblotting with CgtA_E antibodies.

Sample preparation for peptide mass fingerprinting.

Bands were excised from Coomassie blue-stained gels and prepared for peptide mass fingerprinting as described previously (49). Briefly, samples were washed in 50% (vol/vol) acetonitrile and 100 mM ammonium bicarbonate. Protein digestions were carried out using 150 to 500 ng of modified porcine trypsin (Promega) in 10 μl of ammonium bicarbonate (100 mM) at 37°C overnight. The peptides were extracted in 60% (vol/vol) acetonitrile-1% (vol/vol) trifluoroacetic acid (TFA), dried, resuspended in 8 μl of 3% (vol/vol) TFA, and sonicated in a Branson 1200 bath for 10 min. Samples were loaded onto a MALDI plate with an equal volume of 10-mg/ml (wt/vol) α-cyano-4-hydroxycinnamic acid-50% (vol/vol) acetonitrile-1% (vol/vol) TFA. MALDI-time-of-flight mass spectrometry was performed on a Voyager-DE STR instrument (PerSeptive Biosystems, Framingham, Mass.) run in delayed extraction reflector mode at the University of Michigan Proteome Consortium. The resulting peptide mass fingerprints were searched using a local copy of the program MS-Fit (part of the Protein Prospector package by P. Baker and K. Clauser [http://prospector.ucsf.edu/]).

Yeast two-hybrid analysis.

Full-length CgtA_E, SpoT, and SpoT fragments were cloned into pAS2-1 (GAL4-BD) and pACT2 (GAL4-AD) and coexpressed in Y190 and a diploid from the mating of Y187 and Y190. Primers used to PCR amplify the relevant genomic DNA are listed in Table 2. Protein expression of each fusion protein in yeast was monitored by immunoblotting with antibodies to the hemagglutinin tag.

β-Galactosidase assays.

For the filter lift assays, yeast two-hybrid strains were patched or spotted on SD-Trp-Leu plates and grown overnight at 30°C. Nitrocellulose filters (BA-S 85; Schleicher and Schuell) were overlaid on the cells, removed, and placed in liquid nitrogen (5 to 10 s) to permeabilize the cells. The filters were placed (cell side up) in petri dishes containing 3MM chromatography paper soaked with Z buffer (60 mM Na₂HPO₄, 35 mM NaH₂PO₄, 10 mM KCl, 1 mM l MgSO₄, pH 7.0) containing 18 mM β-mercaptoethanol and 1 mg of X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside)/ml and incubated at 30°C.

For the o-nitrophenyl-β-d-galactopyranoside assays, yeast two-hybrid strains were grown overnight in 5 ml of liquid SD-Trp-Leu medium. Triplicate independent cultures were diluted to an OD₆₀₀ of 0.2 to 0.3 and incubated for 3 to 5 h with shaking (230 to 250 rpm) until the cells doubled at least twice. Cells from 1.5 ml of each culture were harvested by centrifugation (16,000 × g, 30 s), washed once with Z buffer, and resuspended in 200 or 300 μl of Z buffer. One-hundred-microliter samples were lysed by three freeze-thaw cycles in liquid nitrogen, diluted with 700 μl of Z buffer with 18 mM β-mercaptoethanol, and incubated with 160 μl of 4-mg/ml o-nitrophenyl β-d-galactopyranoside (Sigma; catalog no. N-1127) in Z buffer at 30°C. After color development, reactions were stopped by addition of 400 μl of 1 M Na₂CO₃, cell debris was removed by centrifugation (16,000 × g, 10 min), and the OD₄₂₀ of the supernatants was recorded. β-Galactosidase units are as in the Clontech manual.

RESULTS

The E. coli CgtA_E protein is ribosome associated.

We were intrigued by the observation that the E. coli CgtA_E protein did not appear to be associated with ribosomes (26), whereas the B. subtilis and C. crescentus proteins do associate with ribosomes (32, 56). Since overexpression of CgtA_E suppresses the ribosome assembly-stability defect of an rrmJ deletion mutant (60), we felt that it was possible that, in vivo, CgtA_E is ribosome associated. We speculated that perhaps the association of CgtA_E was not stable under the conditions used during the published sucrose density centrifugation (26). To reexamine the relationship between CgtA_E and the ribosome, His-CgtA_E was purified on a Ni-NTA column followed by fractionation by gel filtration chromatography. CgtA_E eluted over an extensive range of sizes from a very large complex to smaller states (Fig. 1A) as had been previously noted for the B. subtilis Obg protein (56). One possibility is that the CgtA_E protein eluting in the intermediate ranges resulted from disassociation of CgtA_E from the large complex during either purification step.

FIG. 1.

Purified CgtA_E fractionates with an RNase-sensitive complex. His-tagged CgtA_E purified from a Ni-NTA column was separated by gel filtration chromatography and not treated (A) or treated (B) with RNase prior to separation by gel filtration chromatography. Fractions were subjected to SDS-PAGE, and the position of the proteins was detected by silver staining. The column was calibrated with blue dextran (2,000 kDa, peak at fraction 9), catalase (232 kDa, peak at fraction 19), alkaline dehydrogenase (141 kDa, peak at fraction 22), and albumin (44 kDa, peak at fraction 33).

We predicted that the large CgtA_E-associated complex was the ribosome. To provide evidence for this model, we treated the cell extracts with RNase A prior to separation by gel filtration (Fig. 1B) under conditions that should digest the vast majority of the rRNA. This method was previously used to show a ribosome association of the E. coli GTPase Era (36). As seen with Era (36), the majority of the higher complex CgtA_E was diminished and the vast majority of CgtA_E eluted from the column in later fractions. Thus, in vivo, a significant amount of CgtA_E is associated with an RNase A-sensitive complex, and we propose that this complex is the large ribosomal subunit. Interestingly, either with or without RNase A, CgtA_E did not elute at the predicted monomer size, perhaps due to its elongated structure (6) or because it exists as a stable multimer.

Next, we directly addressed whether CgtA_E cofractionated with the ribosomes after ultracentrifugation through sucrose gradients. We previously showed that the ribosome association of the C. crescentus CgtA_C protein detected after sucrose density centrifugation is affected by the divalent cations used in the buffer (32). Therefore, here we used buffer conditions optimized for retaining the ribosome association of CgtA_C. Cell lysates were separated by ultracentrifugation through sucrose gradients and the polyribosome profiles. The majority of the cellular proteins were in the top of the gradient followed by the peaks for the 30S and 50S subunits, the 70S monosome, and the polyribosomes, as indicated in Fig. 2. Immunoblot analysis with anti-CgtA_E antibodies revealed that the majority of CgtA_E cosedimented with free 50S ribosomal subunits with a reduced, but significant, association with the 70S monosome peak (Fig. 2). Longer exposures also reveal CgtA_E throughout the polysome fractions (data not shown). In this study, we used cells that were also expressing His-CgtA_E. The profile of this fusion protein was identical to that of the native protein, demonstrating that the tagged variant also associates with ribosomes and behaves similarly to the wild-type protein in respect to ribosome association (Fig. 2).

FIG. 2.

CgtA_E cofractionates primarily with the large ribosomal subunit. E. coli cell lysate from BL21(DE3) cells containing pJM1138 was sedimented through a 15 to 45% sucrose gradient at 207,000 × g for 3 h. During fractionation the samples were monitored with UV at 254 nm. The subsequent fractions were analyzed by immunoblotting with anti-CgtA_E antibodies. Both the His-tagged and wild-type protein sediment primarily in the 50S fractions. L, load.

The E. coli CgtA_E protein has kinetic properties similar to those of the C. crescentus CgtA_C protein.

An N-terminally polyhistidine-tagged CgtA_E protein was stably overexpressed and purified from E. coli BL21(DE3) cells. His-CgtA_E migrated on SDS-PAGE as a 50-kDa protein, 5 kDa larger than the predicted value of 45.4 kDa. The mass of the protein as determined by electrospray mass spectrometry, however, was within 0.05% of the expected value (45,470 Da; data not shown). The soluble protein was purified by affinity chromatography through Ni-NTA resin and eluted as a single peak.

The C. crescentus CgtA_C protein binds guanine nucleotides with modest affinity (1.2 and 0.5 μM for GTP and GDP, respectively [29]). The K_m for GTP for CgtA_E was reported to be 18 μM (60), whereas that of B. subtilis Obg was 5.4 μM (64). We determined the equilibrium binding constant (K_D) of CgtA_E for [α-³²P]GTP using an equilibrium centrifugal ultrafiltration assay (29). At 30°C, CgtA_E displayed a K_D of 7.9 ± 1.5 μM, a number similar to that of the other characterized Obg proteins. It appears, therefore, that a hallmark feature of the Obg proteins is their modest affinity for guanine nucleotides.

In order to investigate the kinetic properties of CgtA_E in detail, we took advantage of the fluorescent properties of N-methyl-3′-O-anthranoyl (mant) guanine nucleotide analogs. The intensity of mant fluorescence reflects the hydrophobicity of the environment of the nucleotide analog and, therefore, can be used as an indicator of the interaction between protein and ligand (23, 39, 41, 51). Binding of CgtA_E to mGTP and mGDP nucleotides led to a substantial increase in mantnucleotide fluorescence (Fig. 3A, compare black lines to gray line), confirming that CgtA_E is a guanine nucleotide binding protein. Peak fluorescence emission for free and bound nucleotides occurs at an excitation wavelength of 361 nm. CgtA_E-mGDP and CgtA_E-mGTP displayed a 1.3- and 2.1-fold fluorescence increase over unbound nucleotide, respectively, when excited at 361 nm (Fig. 3A).

FIG. 3.

The E. coli CgtA_E protein binds mGDP and mGTP with nucleotide-specific Mg²⁺ dependence. (A) Excitation spectra from 310 to 410 nm were recorded at an emission wavelength of 446 nm in the presence of 5 mM MgCl₂. The fluorescence intensity of the mant moiety in the absence of protein is identical whether mant is coupled to GDP or to GTP (represented by superimposed gray line). Upon addition of CgtA_E, the fluorescence intensity of both mGDP and mGTP increases. The fluorescence signal from protein-mGDP is shown by the black dotted line, and the fluorescence signal from protein-mGTP is shown by the black solid line. (B) Binding of mGDP (circles) and mGTP (squares) to CgtA_E was assayed (excitation, 361 nm; emission, 446 nm) in the presence of various concentrations of MgCl₂. Binding of CgtA_E to mGDP was unaffected by the presence or absence of Mg²⁺, whereas CgtA_E bound mGTP in a Mg²⁺-dependent manner. Data shown reflect the averages of three trials.

The C. crescentus CgtA_C protein binds mGTP and, to a lesser degree, mGDP in an Mg²⁺-dependent manner (29). To investigate whether this is also true of CgtA_E, we examined the change in fluorescent signal upon incubation with mGTP and mGDP in the presence of various amounts of Mg²⁺. CgtA_E binds mGDP equally well in the presence and absence of Mg²⁺, whereas the binding to mGTP is enhanced in the presence of Mg²⁺ (Fig. 3B).

We used the difference in fluorescence between bound and free mGDP and mGTP to determine the guanine nucleotide exchange rates k_d of CgtA_E for GDP and GTP (Table 3; Fig. 4) as we have done previously (29-31, 58). CgtA_E was prebound to mGDP or mGTP until apparent saturation was achieved. Excess unlabeled nucleotide was then added, and the rate of fluorescence decrease was measured over time. In the absence of Mg²⁺, the release of mGDP and mGTP at 37°C occurred with average rate constants of 1.1 and 0.57 s⁻¹, respectively. In the presence of 5 mM Mg²⁺, the rates of nucleotide exchange at 37°C were modestly lower (0.65 and 0.21 s⁻¹, respectively). Similar rates were observed when the assay was performed at 30°C (Table 3). There were no significant differences in the observed dissociation rate constants whether GDP or GTP was used as the competing nucleotide (data not shown). These rates are comparable to those of the C. crescentus CgtA_C protein (29), indicating that rapid exchange of guanine nucleotides may be a common feature of the Obg family of GTPases.

TABLE 3.

Nucleotide exchange rate constants, k_d (second⁻¹), of E. coli CgtA_E

Assay temp (°C)	kd (s−1) for nucleotide at Mg2+ concn (mM)
	mGDP		mGTP
	0	5	0	5
37	1.1 ± 0.1	0.65 ± 0.01	0.57 ± 0.01	0.21 ± 0.01
30	0.67 ± 0.01	0.61 ± 0.01	0.55 ± 0.01	0.14 ± 0.01

FIG. 4.

mGTP hydrolysis by CgtA_E. Hydrolysis of mGTP by CgtA_E was monitored by saturating CgtA_E with mGTP and observing the decrease in fluorescence over time as CgtA_E-mGTP was converted to CgtA_E-mGDP. Data (black circles) were fitted to a single exponential decay equation (gray line).

We also determined the hydrolysis rate of GTP by CgtA_E by monitoring the reduction in fluorescence that accompanies the single turnover conversion of bound mGTP to bound mGDP (29). At 37°C, we obtained a T_1/2 of 42 ± 5 min corresponding to a single turnover rate constant of 0.017 ± 0.002 min⁻¹ (Fig. 2B), a rate similar to that previously reported for CgtA_E (0.02 min⁻¹ [60]), B. subtilis (0.006 min⁻¹ [64]), and C. crescentus (0.03 min⁻¹ [29]).

CgtA_E and SpoT are interacting proteins.

We observed that several additional faint bands coeluted with His-CgtA_E during purification with a Ni-NTA column (Fig. 5). Many of these bands appeared to be coeluting with the E. coli CgtA_E protein specifically, as we did not observe them during the purification of other E. coli His-tagged GTPases such as Era (58), HflX, YchF, and YfgK (data not shown). To determine the identity of some of these coeluting proteins, we purified His-CgtA_E on a Ni-NTA column, excised bands from SDS-polyacrylamide gels, subjected them to trypsin digestion, determined the masses of the resulting peptides, and identified the corresponding proteins by mass fingerprinting as indicated (Fig. 5). Several of these proteins were CgtA_E. In addition, two additional proteins copurified with CgtA_E, CsdA/DeaD and SpoT. CsdA is a DEAD-box helicase important for assembly of the 30S ribosomal subunit (38) and is a major ribosome-associated protein in cells grown at 15°C (24). SpoT is a bifunctional enzyme that has both (p)ppGpp synthetase and hydrolase activity (21). Bacteria other than the γ- and β-proteobacteria possess only one RelA/SpoT protein, called Rel (37). In these bacteria, the ribosome-associated Rel protein is responsible for both (p)ppGpp synthetase and (p)ppGpp hydrolase activity (2, 35). It should be noted that, due to the large number of CgtA_E degradation products, our analysis was restricted to bands that migrated more slowly than CgtA_E; therefore, we would not have identified any of the smaller ribosomal proteins, if present.

FIG. 5.

CgtA_E copurifies with SpoT and CsdA. E. coli extracts expressing His-tagged CgtA_E were subjected to Ni-NTA chromatography, and the His-CgtA_E eluting fractions were pooled and separated by SDS-PAGE and visualized by Coomassie blue staining. Several protein bands were subjected to mass fingerprinting and identified, as indicated. The size of the molecular weight (MW) markers is indicated on the left (in thousands).

To determine whether SpoT and CgtA_E interact directly, we expressed in E. coli, individually, His-CgtA_E and MBP-SpoT; mixed the resulting cell lysates; and subjected them to purification by Ni-NTA or amylose chromatography. Purification of His-CgtA_E on Ni-NTA resin results in a major His-CgtA_E band, a band migrating at the size of MBP-SpoT, and several additional bands (Fig. 6A, lane 2). Immunoblot analysis with antibodies to MBP confirmed that a significant amount of MBP-SpoT was immobilized on the Ni-NTA column with His-CgtA_E (Fig. 6B, lane 2). As a negative control, we also mixed cell lysates from DsbC-His (DsbC is an E. coli periplasmic disulfide-bond isomerase [3]) and MBP-SpoT and subjected them to Ni-NTA chromatography. As predicted, DsbC-His was found in the eluate whereas MBP-SpoT was not, indicating that MBP-SpoT did not associate nonspecifically with the Ni-NTA column or with DsbC-His (Fig. 6).

FIG. 6.

Copurification of MBP-SpoT and His-CgtA_E by Ni-NTA and amylose columns. (A) Coomassie blue-stained SDS-polyacrylamide gel. (B) Analysis of the samples from panel A by immunoblotting with anti-MBP (MBP), anti-histidine (His), and anti-CgtA_E antibodies. Lanes 1 and 3 are lysates expressing either His-CgtA_E (∼50 kDa) or DsbC-His (∼26 kDa) incubated with MBP-SpoT (∼115 kDa)-expressing lysate, respectively, and the resultant eluates after Ni-NTA affinity purification are shown in lanes 2 and 4, respectively. Lanes 5 and 7 are lysates expressing His-CgtA_E or MBP (∼50 kDa) incubated with MBP-SpoT-expressing lysates, respectively, and the resultant eluates after amylose column purification are shown in lanes 6 and 8, respectively. The protein molecular weight standards are labeled “M.” The asterisks indicate breakdown products of DsbC-His and MBP in lanes 3 and 4 and lanes 7 and 8, respectively.

Next, we asked whether His-CgtA_E copurified with MBP-SpoT on an amylose column. Mixed-cell lysates of strains expressing His-CgtA_E and MBP-SpoT were mixed and applied to an amylose column. In addition to the major MBP-SpoT band, many other bands including one migrating at the size of His-CgtA_E were observed (Fig. 6A, lane 6). Immunoblot analysis revealed that a significant amount of His-CgtA_E also eluted with MBP-SpoT (Fig. 6B, lane 6). As a negative control, we performed the same experiment with His-CgtA_E and MBP. The Coomassie blue-stained gel of the eluting fraction was uninformative, as His-CgtA_E and MBP migrate at similar positions (Fig. 6A, lane 8). Immunoblot analysis revealed, however, that very little His-CgtA_E was in the eluate (Fig. 6B, lane 8) and represented a contaminating association of His-CgtA_E that we also observed with passage of His-CgtA_E lysates over amylose resin alone (data not shown).

To confirm that the interaction between CgtA_E and SpoT was direct, we assayed their interactions in vivo using the yeast two-hybrid method. Full-length (FL) CgtA_E was expressed as a GAL4-binding domain fusion, GAL4-BD, and full-length SpoT was expressed as a GAL4-activating domain fusion, GAL4-AD. The presence of the GAL4-BD-CgtA_E(FL) alone resulted in modest expression of the reporter genes (Fig. 7). Coexpression of the CgtA_E(FL) and SpoT(FL) fusion constructs resulted in enhanced activation of the lacZ reporter gene (10-fold; Fig. 7), indicating that these two fusion proteins interact as analyzed by the yeast two-hybrid system. CgtA_E(FL) also showed an interaction with the N-terminal ppGpp synthetase/hydrolase domain of SpoT (amino acids 1 to 374) as well as the putative regulatory domain at the C terminus (the ACT domain; amino acids 625 to 702) (Fig. 7). Thus, CgtA_E appears to make extensive contacts with SpoT that include the catalytic domains as well as the potential regulatory domain. Reciprocal fusion proteins were also examined, as yeast two-hybrid analysis often shows reporter gene expression in only one direction. In this case, interactions between CgtA_E and SpoT(FL) or SpoT(N) were detected but an interaction with SpoT(C) was not (data not shown).

FIG. 7.

Domain structure of SpoT and interaction with CgtA_E by the yeast two-hybrid assay. The overlapping ppGpp degradation and synthesis domains (16), the TGS domain (65), and the ACT domain (1) are indicated for the full-length SpoT [SpoT(FL)] protein. Also shown are cartoons of the SpoT(N) and SpoT(C) subclones used in this study. For each GAL4-AD fusion, the amino acids of SpoT expressed are indicated. The levels of β-galactosidase observed (± standard deviations) from cells coexpressing the GAL4-BD-CgtA_E(FL) and the indicated GAL4-AD-SpoT fusion constructs or empty pACT2 vector (“None”) are given.

DISCUSSION

In this report we show that CgtA_E is associated predominantly with the 50S ribosomal subunit. Prior to this report, the ribosome association of E. coli CgtA_E was of some controversy. CgtA_E has been previously implicated in 50S ribosomal subunit assembly and/or stability based on the ability of high-copy-number CgtA_E to suppress the ribosome defect of an rrmJ mutant (60). CgtA_E was reported, however, to be associated with the cell membrane and not with ribosomes (26), although the conditions used to pellet the membrane were also sufficient to pellet the ribosomes. Moreover, CgtA_E did not cofractionate with ribosomes separated by sucrose density centrifugation (26). We have noted that the ribosome association of the C. crescentus CgtA_C protein following sucrose density centrifugation is sensitive to the divalent anions used in the buffer (32). Here we show that, under the conditions that revealed cosedimentation of CgtA_C with the 50S subunit, the majority of CgtA_E also associated with the 50S subunit. In addition, a small but reproducible amount of CgtA_E cofractionated with the 70S particle and polyribosomes.

CgtA_E also interacts with the stress response protein SpoT. The universal cellular response to stress is through the synthesis of the alarmone (p)ppGpp. In E. coli, during amino acid starvation, (p)ppGpp is synthesized by RelA, a ribosome-associated enzyme (50) activated by uncharged tRNAs (20) and controlled, in part, by the 50S subunit protein L11 (66). SpoT is a bifunctional enzyme that has both (p)ppGpp synthetase and hydrolase activity and is 50% similar in sequence to RelA (21). The intracellular location of SpoT has historically been controversial. SpoT purifies with crude ribosomal fractions (21, 52, 59) but is not found associated with ribosomes separated by sucrose density centrifugation (15). Given that SpoT interacts with CgtA_E, a ribosome-associated protein, we predicted that SpoT is also ribosome associated but not detected on sucrose gradients under the conditions used previously (15). Recent work from our laboratory reveals that SpoT is, in fact, associated with ribosomes (P. Wout, M. J. Jiang, and J. Maddock, unpublished data).

Interestingly, a connection among ribosomes, Obg, and stress response has been recently suggested for B. subtilis. In addition to the nutritional stress response pathway mediated by RelA, a limited number of gram-positive bacteria, such as B. subtilis, respond to physical stress through RsbW-mediated activation of the transcription factor σ^B. RsbW is an anti-σ^B protein controlled by a regulatory cascade that involves the RsbT, RsbS, RsbR, and RsbV proteins (see reference 56). Interestingly, Obg interacts with several of these Rsb proteins and is necessary for activation of the stress response (55). Obg also copurifies with ribosomes and is specifically associated with the ribosomal protein L13 (56). It is not clear whether Obg communicates a signal to the ribosomes or whether the ribosomal state acts as an Obg effector. Recently, it was shown that Rel_Bsu also plays a role in the σ^B activation that does not appear to involve ppGpp (67), raising the possibility that Obg and Rel_Bsu are functionally related in B. subtilis.

Several lines of evidence have implicated ribosomes in direct or indirect roles in the stress response in E. coli. First, amino acid starvation induces ribosome-mediated activation of RelA (50). Second, addition of ribosome-specific antibiotics increases the expression of heat and cold shock proteins in a manner similar to that resulting from stress conditions (62). Finally, a variety of heat shock proteins, such as Hsp15, Hsp70, RrmJ, DnaK/J, and GrolEL/S, have been found either to bind to ribosomes (27, 46) or to be involved in rRNA modification during ribosome biogenesis (5, 7, 11).

It is becoming widely appreciated that bacteria contain a number of essential bacterial GTP-binding proteins, including the Obg proteins, that are likely to be involved in ribosome function (9, 28, 37). An interesting feature of the Obg proteins is their relatively modest affinity for guanine nucleotides (29, 60, 64), rapid exchange kinetics for both GTP and GDP (29), and relatively low hydrolysis rates (29, 60, 64). These biochemical features are also seen with the 30S ribosome-associated GTPase, Era (58). It has previously been proposed that these proteins are controlled, in part, by the relative levels of GTP-GDP in the cell (29, 48). An additional possibility is that the association with the ribosome or other interacting partners effects guanine nucleotide binding and/or GTP hydrolysis, as has been shown for the 30S binding GTPase YjeQ (10).

We report here that CgtA_E is associated with the 50S ribosomal subunit and the SpoT protein. Since CgtA_E interacts with SpoT, it is possible that the downstream effector of CgtA_E is SpoT. Interestingly, based on our two-hybrid studies, CgtA_E interacts with both the catalytic and putative regulatory (ACT) domains of SpoT. Thus, CgtA_E is well poised to be involved in the control of SpoT function. An alternate possibility is that SpoT could affect the activity of CgtA_E. SpoT is a binding partner of CgtA_E but also catalyzes the synthesis of ppGpp. Moreover, the B. subtilis Obg protein cocrystallized with ppGpp (6), and this alarmone, therefore, should be considered a bona fide substrate of CgtA_E in vivo. Through its interaction with SpoT, CgtA_E is well positioned to receive the ppGpp nucleotide, which could affect the activity of CgtA_E. In E. coli, and perhaps other bacteria, it is likely that SpoT, CgtA_E, the 50S ribosomal subunit, and ppGpp are functionally intertwined. Unraveling their relationships is critical to completely understanding their respective functions.