A modified infrared spectrometer with high time resolution and its application for investigating fast conformational changes of the GTPase Ras.

Time-resolved infrared spectroscopy is a valuable tool for the investigation of proteins and protein interactions. The investigation of many biological processes is possible by means of caged compounds, which set free biologically active substances upon light activation. Some caged compounds could provide sub-nanosecond time resolution, e.g., para-hydroxyphenacyl-guanosine 5'-triphosphate (GTP) forms GTP in picoseconds. However, the time resolution in single shot experiments with rapid-scan Fourier transform infrared (FT-IR) spectrometers is limited to about 10 ms. Here we use an infrared diode laser instead of the conventional globar and achieve a time resolution of 100 ns. This allows for the time-resolved measurement of the fast Ras(off) to Ras(on) conformational change at room temperature. We quantified the activation parameters for this reaction and found that the free energy of activation for this reaction is mainly enthalpic. Investigation of the same reaction in the presence of the Ras binding domain of the effector Raf (RafRBD) reveals a four orders of magnitude faster reaction, indicating that Ras·RafRBD complex formation directly induces the conformational change. Recent developments of broadly tunable quantum cascade lasers will further improve time resolution and usability of the setup. The reported 100 ns time resolution is the best achieved for a non-repetitive experiment so far.

The function of conserved amino acid residues adjacent to the effector domain in elongation factor G.

Bacterial elongation factor G (EF-G) physically associates with translocation-competent ribosomes and facilitates transition to the subsequent codon through the coordinate binding and hydrolysis of GTP. In order to investigate the amino acid positions necessary for EF-G functions, a series of mutations were constructed in the EF-G structural gene (fusA) of Escherichia coli, specifically at positions flanking the effector domain. A mutated allele was isolated in which the wild-type sequence from codons 29 to 47 ("EFG2947") was replaced with a sequence encoding 28 amino acids from ribosomal protein S7. This mutated gene was unable to complement a fusAts strain when supplied in trans at the nonpermissive temperature. In vitro biochemical analysis demonstrated that nucleotide crosslinking was unaffected in EFG2947, while ribosome binding appeared to be completely abolished. A series of point mutations created within this region, encoding L30A, Y32A, H37A, and K38A were shown to give rise to fully functional proteins, suggesting that side chains of these individual residues are not essential for EF-G function. A sixth mutant, E41A, was found to inefficiently rescue growth in a fusAts background, and was also unable to bind ribosomes normally in vitro. In contrast E41Q could restore growth at the nonpermissive temperature. These results can be explained within the context of a three-dimensional model for the effector region of EF-G. This model indicates that the effector domain contains a negative potential field that may be important for ribosome binding.

Fourier transform infrared photolysis studies of caged compounds.

Time-resolved FTIR difference spectroscopy is a powerful tool for investigating molecular reaction mechanisms of proteins. In order to detect, beyond the large background absorbance of the protein and the water, absorbance bands of protein groups that undergo reactions, difference spectra have to be performed between a ground state and an activated state of the sample. Because the absorbance changes are small, the reaction has to be started in situ, in the apparatus, and in thin protein films. The use of caged compounds offers an elegant approach to initiate protein reactions with a nanosecond UV laser flash. Here, time-resolved FTIR and FT-Raman photolysis studies of the commonly used caged compounds, caged Pi, caged ATP, caged GTP, and caged calcium are presented. The use of specific isotopic labels allows us to assign the IR bands to specific groups. Because metal ions play an important role in many biological systems, their influence on FTIR spectra of caged compounds is discussed. The results presented should provide a good basis for further FTIR studies on molecular reaction mechanisms of energy or signal transducing proteins. As an example of such investigations, the time-resolved FTIR studies on the GTPase reaction of H-ras p21 using caged GTP is presented.

When an ATPase is not an ATPase: at low temperatures the C-terminal domain of the ABC transporter CvaB is a GTPase.

The ATP-binding cassette (ABC) transporters belong to a large superfamily of proteins which share a common function and a common nucleotide-binding domain. The CvaB protein from Escherichia coli is a member of the bacterial ABC exporter subfamily and is essential for the export of the peptide antibiotic colicin V. Here we report that, surprisingly, the CvaB carboxyl-terminal nucleotide-binding domain (BCTD) can be preferentially cross-linked to GTP but not to ATP at low temperatures. The cross-linking is Mg2+ and Mn2+ dependent. However, BCTD possesses similar GTPase and ATPase activities at 37 degrees C, with the same kinetic parameters and with similar responses to inhibitors. Moreover, a point mutation (D654H) in CvaB that completely abolishes colicin V secretion severely impairs both GTPase and ATPase activities in the corresponding BCTD, indicating that the two activities are from the same enzyme. Interestingly, hydrolysis activity of ATP is much more cold sensitive than that of GTP: BCTD possesses mainly GTP hydrolysis activity at 10 degrees C, consistent with the cross-linking results. These findings suggest a novel mechanism for an ABC protein-mediated transport with specificity for GTP hydrolysis.

The SpoIIAA protein of Bacillus subtilis has GTP-binding properties.

SpoIIAA is the first protein of the spoIIA operon. Here we show that SpoIIAA can bind and hydrolyze GTP. The protein also accepts ATP, but with lower affinity. GDP competes poorly for binding of GTP. The GTPase activity of SpoIIAA is within the range found for other GTP-binding proteins.

Isoproterenol and GTP gamma S inhibit L-type calcium channels of differentiating rat skeletal muscle cells.

In adult skeletal muscle, G-proteins have been shown to modulate the calcium channels both directly and through a cAMP-dependent phosphorylating mechanism. We have investigated the action of G-proteins on the L-type calcium current in cultured rat muscle cells (myoballs) under voltage clamp in whole cell or perforated patch modes. Intracellular photolytic release of 200 microM GTP gamma S inhibited the L-type calcium current. Inclusion of 500 microM uncaged GTP gamma S in the patch pipette in the whole cell configuration reduced the calcium current by a similar amount. Under perforated patch conditions external application of 10 microM of the beta-adrenergic agonist isoproterenol also reduced the calcium current. Pretreatment of the cells with pertussis toxin reversed the effect of GTP gamma S and removed that of isoproterenol. We conclude that rat myoballs contain beta-adrenergic receptors that inhibit the L-type calcium current, and that this inhibition is mediated by a pertussis toxin-sensitive G-protein.

Activation of L-type Ca2+ currents in cardiac myocytes by photoreleased GTP.

L-type calcium currents (ICa) were recorded from isolated ventricular myocytes by using standard patch-clamp methods. In the absence of agonist, photorelease of GTP by flash photolysis of intracellularly applied caged-GTP rapidly increased the amplitude of ICa over a wide range of membrane potentials. Control experiments clearly demonstrated that this effect was not due to either the release of photolytic by-products or to the light flash itself. The timecourse for activation of ICa by photolysis of caged-GTP was markedly altered by intracellular application of either GDP beta S or GTP gamma S. Upon maximal stimulation of ICa by intracellular dialysis with cAMP, photoreleased GTP induced a small, rapid increase in ICa followed by a gradual inhibition. The presence of Rp-cAMPS intracellularly reduced both the magnitude of the response to photoreleased GTP and its time to peak. Similar effects were observed when protein kinase inhibitor dialysed the cell interior, suggesting that both cAMP-dependent and independent processes were involved in this effect. We conclude that rapid release of GTP within ventricular myocytes, in the absence of agonist, causes rapid activation of L-type Ca2+ current. Mechanisms underlying this effect include stimulation of adenylate cyclase, together with other, as yet uncharacterized, GTP-dependent pathways for increasing ICa in the heart.

Flash photolysis of intracellular caged GTP gamma S increases L-type Ca2+ currents in cardiac myocytes.

L-type calcium currents were recorded from isolated ventricular myocytes using standard patch-clamp methods. Rapid release of photolabile guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S), tetralithium salt, by flash photolysis of intracellular caged-GTP gamma S increased the magnitude of the L-type calcium current: an effect independent of the ultraviolet flash per se or the production of photolytic by-products. This increase in calcium current was markedly reduced by intracellularly applied guanosine 5-O-(thiodiphosphate), trilithium salt or by an excess of GTP gamma S. It is therefore likely that rapid release of GTP gamma S intracellularly in the absence of an agonist can, via a G protein-mediated mechanism, cause L-type calcium current activation. In the presence of Rp-adenosine 3,5-mono-thionophosphate (Rp-cAMP), photoreleased GTP gamma S results in a transient and much reduced increase in the amplitude of the L-type calcium current. We conclude that activation of Gs coupled to adenylate cyclase and ultimately cAMP-dependent phosphorylation may be primarily responsible for L-type channel activation, although a fast membrane-delimited (direct) pathway, not involving cytoplasmic second messengers, may also contribute to this effect.

Microtubule assembly and oscillations induced by flash photolysis of caged-GTP.

Microtubule assembly and oscillations have been induced using the rapid liberation of GTP by UV flash photolysis of caged-GTP and monitored by time-resolved X-ray scattering. The flash photolysis method of achieving assembly conditions is much faster than the temperature jump method used earlier (msec vs. s range). However, the structural transitions and their rates are similar to those described previously. This means that the rates of the transitions in microtubule assembly observed before are determined by the protein itself, and not by the rate at which assembly conditions are induced. The advantages and limitations of using the photolysis of caged-GTP in microtubule assembly studies are compared with temperature jump methods. Caged-GTP itself reduces the rate of microtubule assembly and oscillations at mM concentrations, consistent with a weak interaction between the nucleotide analogue and the protein. X-rays are capable of slowly liberating GTP and other breakdown products from caged-GTP, even in the absence of UV flash photolysis, thus causing an apparent "X-ray-induced" microtubule assembly. This effect depends on the X-ray dose but is independent of the caged-GTP concentrations used here (mM range), suggesting that the breakdown of caged-GTP is caused not by the direct absorption of X-rays by the compound but by another intermediate reaction such as the generation of radicals by the X-rays.

Use of 8-azidoguanosine 5'-[gamma-32P]triphosphate as a probe of the guanosine 5'-triphosphate binding protein subunits in bovine rod outer segments.

In an in vitro incubation, 8-azidoguanosine 5'-[gamma-32P]triphosphate ( [gamma-32P]-8-azido-GTP) labeled bleached rhodopsin independent of ultraviolet light. Characterization of this labeling indicated that rhodopsin was phosphorylated with [gamma-32P]-8-azido-GTP as a phosphate donor. At low concentrations, ATP increased this labeling activity 5-fold. In the same incubation, [gamma-32P]-8-azido-GTP also labeled G alpha (Mr 40 000). This labeling was ultraviolet light dependent. G beta (Mr 35 000) was also labeled dependent for the most part upon ultraviolet light, but a smaller component of labeling appeared to result from phosphorylation. Differential labeling of G alpha and G beta was found to vary intricately with experimental conditions, especially prebleaching of rhodopsin, tonicity of the medium, and the presence or absence of 2-mercaptoethanol. Affinity labeling of G alpha and G beta by [gamma-32P]-8-azido-GTP in competition with ATP or GTP was kinetically complex, consistent with possible multiple binding sites for GTP on both subunits. Independent evidence for two or more binding sites on G alpha has been offered by other laboratories, and recently, at least one binding site on G beta and its analogues among the N proteins of adenylate cyclases has been identified.

The alpha and gamma subunits of initiation factor eIF-2 can be cross-linked to 18S ribosomal RNA within the quaternary initiation complex, eIF-2.Met-tRNAf.GDPCP.small ribosomal subunit.

Initiation factor eIF-2 from rat liver was reacted with the hetero-bifunctional cross-linking reagents ABAI or APTPI without diminishing its ability to form the quaternary initiation complex with Met-tRNAf, GDPCP and the small ribosomal subunit. Upon irradiation with UV light, subunits alpha and gamma of eIF-2 became covalently linked to 18S ribosomal RNA. The subunits were identified electrophoretically after isolation of the covalent protein-rRNA complexes and subsequent degradation of the rRNA by nuclease and alkali treatments. The close proximity of the two factor subunits to sequences of ribosomal RNA within the quaternary complex could be confirmed in a second set of experiments using unmodified, 125I-labeled factor and diepoxybutane as cross-linking reagent.

Interactions of a photoaffinity analog of GTP with the proteins of microtubules.

Tubulin dimers isolated from brain contain two GTP binding sites, a nonexchangeable site and an exchangeable site. To localize the exchangeable site, we used a photoaffinity analog of GTP, 8-azidoguanosine triphosphate (8-N3GTP), which supports tubulin polymerization in the absence of activating light. Photolysis of tubulin polymerized in the presence of 0.01 to 0.1 mM [beta, gamma-32P]8-N3GTP resulted in covalent incorporation of radioactivity only onto the beta monomer. Photolysis with 8-N3GTP also prevented any further repolymerization of the tubulin whereas like treatment in the presence of GTP had no effect. Preincubation of tubulin with GTP prevented photo-incorporation of [beta, gamma-32P]8-N3GTP whereas preincubation with ATP did not.