Mechanism of inactivation of the beta 2 subunit of Escherichia coli tryptophan synthase by monoclonal antibodies.

Monoclonal antibodies directed against the native form of the beta 2 subunit of Escherichia coli tryptophan synthase strongly inhibit both its tryptophan synthase and its serine deaminase activities. The mechanism of this inactivation is studied here, by monitoring quantitatively the absorption and fluorescence properties of different well-characterized successive intermediates in the catalytic cycle of tryptophan synthase. It is shown that the antibodies interfere specifically with the formation of one or the other of these intermediates. It is concluded that the antibodies either modify or block the molecular flexibility of the protein, thus preventing conformational changes that the protein has to undergo during the catalysis. At least two different stages of the catalytic process, each one sensitive to a different class of antibodies, are shown to involve molecular movements of the polypeptide chain. Indications are given on the regions of the molecule involved in these movements.     

Conformational effects of ligand binding on the beta 2 subunit of Escherichia coli tryptophan synthase analyzed with monoclonal antibodies

Five monoclonal antibodies recognizing five different epitopes of the native beta 2 subunit of Escherichia coli tryptophan synthase (EC 4.1.2.20) were used to analyze the conformational changes occurring upon ligand binding or chemical modifications of the enzyme. For this purpose, the affinities of each antibody for the different forms of the enzyme were determined by using an enzyme-linked immunosorbent assay which allows measurement of the dissociation constant of antigen-antibody equilibrium in solution. The fixation of the coenzyme pyridoxal 5'-phosphate and the substrate L-serine modifies the affinity constants of most of the antibodies for the enzyme, thus showing the existence of extended conformational rearrangements of the protein. The association of the alpha subunit with the beta 2 subunit, which brings about an increase of the tryptophan synthase activity and abolishes the serine deaminase activity of beta 2, is accompanied by an important conformational change of the N-terminal domain of beta 2 (F1) since none of the anti-F1 monoclonal antibodies can bind to alpha 2 beta 2. Similarly, chemical modifications of beta 2 which are known to produce significant effects on the enzymatic activities of beta 2 result in changes of the affinities of the monoclonal antibodies which can be interpreted as the acquisition of different conformational states of the enzyme.     

Conformational changes induced by domain assembly within the beta 2 subunit of Escherichia coli tryptophan synthase analysed with monoclonal antibodies

The effects of domain assembly on the conformation of the F1 (N-terminal) and F2 (C-terminal) domains of the beta 2 subunit of Escherichia coli tryptophan synthase (EC 4.2.1.20) were analysed using six monoclonal antibodies which recognize six different epitopes of the native beta 2 subunit (five carried by the F1 domain and one carried by the F2 domain). For this purpose, the affinity constant of each monoclonal antibody for the isolated domains F1 or F2, the associated domains in the trypsin-nicked apo-beta 2 and in the native apo-beta 2 subunits were determined, both with the intact immunoglobulin and the Fab fragment. It was found that the association of the F1 and F2 domains within beta 2 is accompanied by structural changes of the two domains, as detected by variations of their affinity constants for the monoclonal antibodies.     

L-serine binds to arginine-148 of the beta 2 subunit of Escherichia coli tryptophan synthase

Inactivation of the beta 2 subunit and of the alpha 2 beta 2 complex of tryptophan synthase of Escherichia coli by the arginine-specific dicarbonyl reagent phenylglyoxal results from modification of one arginyl residue per beta monomer. The substrate L-serine protects the holo beta 2 subunit and the holo alpha 2 beta 2 complex from both inactivation and arginine modification but has no effect on the inactivation or modification of the apo forms of the enzyme. This result and the finding that phenylglyoxal competes with L-serine in reactions catalyzed by both the holo beta 2 subunit and the holo alpha 2 beta 2 complex indicate that L-serine and phenylglyoxal both bind to the same essential arginyl residue in the holo beta 2 subunit. The apo beta 2 subunit is protected from phenylglyoxal inactivation much more effectively by phosphopyridoxyl-L-serine than by either pyridoxal phosphate or pyridoxine phosphate, both of which lack the L-serine moiety. The phenylglyoxal-modified apo beta 2 subunit binds pyridoxal phosphate and the alpha subunit but cannot bind L-serine or L-tryptophan. We conclude that the alpha-carboxyl group of L-serine and not the phosphate of pyridoxal phosphate binds to the essential arginyl residue in the beta 2 subunit. The specific arginyl residue in the beta 2 subunit which is protected by L-serine from modification by phenyl[2-14C]glyoxal has been identified as arginine-148 by isolating a labeled cyanogen bromide fragment (residues 135-149) and by digesting this fragment with pepsin to yield the labeled dipeptide arginine-methionine (residues 148-149). The primary sequence near arginine-148 contains three other basic residues (lysine-137, arginine-141, and arginine-150) which may facilitate anion binding and increase the reactivity of arginine-148. The conservation of the arginine residues 141, 148, and 150 in the sequences of tryptophan synthase from E. coli, Salmonella typhimurium, and yeast supports a functional role for these three residues in anion binding. The location and role of the active-site arginyl residues in the beta 2 subunit and in two other enzymes which contain pyridoxal phosphate, aspartate aminotransferase and glycogen phosphorylase, are compared.     

Fluorescence-quenching studies on a conformational transition within a domain of the beta 2 subunit of Escherichia coli tryptophan synthase

The fluorescence quenching by acrylamide of the single tryptophan residue in the beta 2 subunit of tryptophan synthase from Escherichia coli K12 is studied for different states of the protein: the native apo-enzyme and holo-enzyme, the nicked apo-protein and holo-protein and the isolated proteolytic fragment F1 corresponding to the N-terminal two thirds of beta 2. The quenching constants measured are used to estimate the accessibility of the tryptophan residue in these different forms. The results are discussed in terms of conformational transition within the F1 domain, occurring in the presence of the cofactor, pyridoxal 5'-phosphate, in the native enzyme. The proteolytic cleavage of the native enzyme is shown to render the nicked protein unable to undergo this conformational change.     

Affinity labeling of the pyridoxal phosphate binding site of the beta2 subunit of Escherichia coli tryptophan synthase.

We have synthesized bromoacetylpyridoxamine phosphate and bromoacetylpyridoxamine and have shown that they meet three criteria for affinity labels of the beta2 subunit of tryptophan synthase: (i) the kinetic data of inactivation indicate that a binary complex is formed prior to covalent attachment; (ii) inactivation is largely prevented by the presence of pyridoxal phosphate; and (iii) inactivation is stoichiometric with incorporation of 0.7 to 0.8 mol of chromophore/mol of beta monomer. Our conclusion that inactivation of the apo beta2 subunit by bromoacetylpyridoxamine phosphate is due to the modification of cysteine is based on the disappearance of 1 mol of -SH/beta monomer and on the finding that [14C]carboxymethyl derivative in the acid hydrolysate of the protein modified by bromo[14C]acetylpyridixamine phosphate. A 39-residue tryptic peptide containing this essential cysteine has been isolated and purified from the bromo[14C]acetylpyridoxamine phosphate-labeled beta2 subunit.     

Selection and analysis of non-interactive mutants in the Escherichia coli tryptophan synthase alpha subunit.

Summary The inherent infidelity of Taq DNA polymerase in the polymerase chain reaction was exploited to produce random mutations in thetrp A gene. Screening of the resulting clones allowed selection of non-interactive mutant subunits retaining their intrinsic catalytic activity. Two single changes responsible for this phenotype were identified by DNA sequencing as: 126 valine (GTG)glutamic acid (GAG) and 128 valine (GTT)aspartic acid (GAT). Three single changes giving a non-interactive phenotype with an impaired intrinsic catalytic activity were identified by DNA sequencing as a66 asparagine (AAC)aspartic acid (GAC); 109lysine (AAA) arginine (AGA); 118 cysteine (TGC)arginine (CGC). Where possible, we individually assessed the importance of these residues in interaction in light of structural information from X-ray crystallography and by intergeneric protein sequence comparison.     

Probing ribosome function and the location of Escherichia coli ribosomal protein L5 with a monoclonal antibody

A monoclonal antibody specific for Escherichia coli ribosomal protein L5 was isolated from a cell line obtained from Dr. David Schlessinger. Its unique specificity for L5 was confirmed by one- and two-dimensional electrophoresis and immunoblotting. The antibody recognized L5 both in 50 S subunits and 70 S ribosomes. Both antibody and Fab fragments had similar effects on the ribosome functions tested. Antibody bound to 50 S subunits inhibited their reassociation with 30 S subunits at 10 mM Mg2+ but not 15 mM, the concentration present for in vitro protein synthesis. The 70 S couples were not dissociated by the antibody. The antibody caused inhibition of polyphenylalanine synthesis at molar ratios to 50 S or 70 S particles of 4:1. The major inhibitory effect was on the peptidyltransferase reaction. There was no effect on either elongation factor binding or the associated GTPase activities. The site of antibody binding to 50 S was determined by electron microscopy. Antibody was seen to bind beside the central protuberance or head of the particle, on the side away from the L7/L12 stalk, and on or near the region at which the 50 S subunit interacts with the 30 S subunit. This site of antibody binding is fully consistent with its biochemical effects.     

Evolution of the tryptophan synthetase of fungi. Analysis of experimentally fused Escherichia coli tryptophan synthetase alpha and beta chains

During evolution of fungi, the separate tryptophan synthetase alpha and beta polypeptides of bacteria appear to have been fused in the order alpha-beta rather than the beta-alpha order that would be predicted from the order of the corresponding structural genes in all bacteria. We have fused the tryptophan synthetase polypeptides of Escherichia coli in both orders, alpha-beta and beta-alpha, with and without a short connecting (con) sequence, to explore possible explanations for the domain arrangement in fungi. We find that proteins composed of any of the four fused polypeptides, beta-alpha, beta-con-alpha, alpha-beta, and alpha-con-beta, are highly active enzymatically. However, only the alpha-beta and alpha-con-beta proteins are as active as the wild type enzyme. All four fusion proteins appear to be less soluble in vivo than the wild type enzyme; this abnormal characteristic is minimal for the alpha-con-beta enzyme. The alpha and beta domains of the four fusion polypeptides were not appreciably more heat labile than the wild type polypeptides. Competition experiments with mutant tryptophan synthetase alpha protein, and the fusion proteins suggest that in each fusion protein the joined alpha and beta domains have a functional tunnel connecting their alpha and beta active sites. Three tryptophan synthetase beta'-alpha fusion proteins were examined in which the carboxyl-terminal segment of the wild type beta polypeptide was deleted and replaced by a shorter, unnatural sequence. The resulting deletion fusion proteins were enzymatically inactive and were found predominantly in the cell debris. Evaluation of our findings in relation to the three-dimensional structure of the tryptophan synthetase enzyme complex of Salmonella typhimurium (5) and the results of mutational analyses with E. coli suggest that tryptophan synthetase may have evolved via an alpha-beta rather than a beta-alpha fusion because in beta-alpha fusions the amino-terminal helix of the alpha chain cannot assume the conformation required for optimal enzymatic activity.     

Kinetics of appearance of an early immunoreactive species during the refolding of acid-denatured Escherichia coli tryptophan synthase beta 2 subunit

A reversible acid-denaturation process of the beta 2 subunit of Escherichia coli tryptophan synthase has been set up. The acid-denatured state has been physically characterized: though not in a random-coiled conformation, it is extensively denatured. The renaturation of this denatured state of beta 2 has been observed in a stopped-flow system, in the presence of a monoclonal antibody directed against native beta 2. It is shown that the association occurs very early in the folding of beta 2. The association rate constants of the antibody with the immunoreactive folding intermediate and with native beta 2 are the same (3 X 10(5) M-1.s-1). But at high antibody concentrations the formation of the antigen/antibody complex is rate limited by a rapid (5.4 X 10(-2) s-1) isomerization of refolding beta chains. This isomerization appears to reflect the formation of at least part of the epitope recognized by the antibody during the folding of beta 2. Further conformational adjustments occurring later in the folding pathway would then allow the ultimate structuring of the epitope.     

Kinetics and importance of the dimerization step in the folding pathway of the beta 2 subunit of Escherichia coli tryptophan synthase

During its folding, the polypeptide chain of the beta 2 subunit of Escherichia coli tryptophan synthase (L-serine hydrolyase (adding indole) EC 4.2.1.20) undergoes dimerization. To decide whether this dimerization precedes or follows the formation of the native, functional, tertiary structure of the polypeptide chain, the kinetics of renaturation of beta 2 are studied by monitoring both the regain of native conformation and the dimerization. Dimer formation is followed by the increase of the fluorescence polarization, or by energy transfer between a fluorescence donor and a fluorescence acceptor, which occur upon association of adequately labelled beta chains. Renaturation is followed by the regain of functional properties of beta 2, i.e. its ability to bind pyridoxal-5'-phosphate or to form a fluorescent ternary complex with this coenzyme and L-serine. It is shown that for beta 2 the dimerization obeys first-order kinetics, presumably because it occurs rapidly after a rate-limiting isomerization of the monomer. The dimerization is followed by another isomerization, taking place within the dimer, which leads to the formation of the pyridoxal-5'-phosphate binding site. Still another, slow, isomerization reaction involving the F1 (N-terminal) domain completes the renaturation. With a modified form of beta 2 (trypsin-nicked beta 2) where this isomerization of F1 can be made to occur before the dimerization, the dimer is also shown to appear before the functional properties. It is concluded that a non-native dimer indeed exists as an obligatory intermediate on the folding pathway of nicked beta 2 and of beta 2, and that interdomain interactions are needed to force the polypeptide chains into their native conformations.     

The beta subunit of the Escherichia coli ATP synthase exhibits a tight membrane binding property

We have developed a chromatographic procedure to analyze the association of the subunits of the Escherichia coli F1Fo-ATP synthase with the cytoplasmic membrane. Minicells containing [35S]-labeled ATP synthase subunits are treated with lysozyme, solubilized, and chromatographed on a Sepharose CL-2B column in buffer containing urea and taurodeoxycholate. ATP synthase subunits are resolved into membrane intrinsic and membrane extrinsic subunits. Interestingly, a significant amount (36%) of the F1 subunit beta fractionates with the membrane intrinsic Fo subunits. About half of this amount (19%) of beta is non-specifically bound to the membrane. Interaction of beta with the membrane is not mediated by the amino terminal portion of beta.     

Immunochemical evidence for conformational flexibility and its modulation by specific ligands in the beta 2 subunit of Escherichia coli tryptophan synthase

The immunochemical reactivity of unfractionated antibodies elicited by denatured beta 2 subunits of Escherichia coli tryptophan synthase [L-serine hydro-lyase (adding indole) EC 4.2.1.20] with the homologous antigen and with the native enzyme is examined. These antibodies recognize the native apoenzyme nearly as well as the denatured protein. On the contrary, after binding of its cofactor, pyridoxal 5'-phosphate, the protein exhibits a much lower immunoreactivity toward these antibodies. This decrease of affinity becomes even more pronounced when the beta 2 protein interacts with the alpha subunit. Similarly, reduction of the Schiff base formed between the cofactor and the protein leads to a strong decrease of immunoreactivity. To account for these results, it is proposed that apo-beta 2 must be a dynamic flexible structure that easily exposes to the solvent regions of its polypeptide chain that normally are buried in its interior. The increase in rigidity of this structure upon binding of the cofactor, reduction of Schiff base, and formation of the alpha 2 beta 2 complex would then account for the decreased immunoreactivity of these various states of the native beta 2 protein.     

Energetic adaptation of ligand binding to subunit structure of tryptophan synthase from Escherichia coli

The binding of indole and L-serine to the isolated alpha and beta 2 subunits and the native alpha 2 beta 2 complex of tryptophan synthase from Escherichia coli was investigated by direct microcalorimetry to reveal the energetic adaptation of ligand binding to the subunit structure of a multienzyme complex. In contrast to the general finding that negative heat capacity changes are associated with ligand binding to proteins, complex formation of indole and the alpha subunit involves a small positive change in heat capacity. This unusual result was considered as being indicative of a loosening of the protein structure. Such an interpretation is in good agreement with results of chemical accessibility studies (Freedberg & Hardman, 1971). Whereas the thermodynamic parameters of indole binding are not influenced by the subunit interaction, the large negative change in heat capacity of -6.5 kJ/(K X mol of beta 2) measured for the binding of L-serine to the isolated beta 2 subunit disappears completely when serine interacts with the tetrameric complex. These data demonstrate that the energy transduction pattern and therefore the functional roles of the substrates indole and L-serine vary strongly with the subunit structure of tryptophan synthase.     

5-Azidoindole binding to the Escherichia coli tryptophan synthase alpha 2 beta 2 complex

The tryptophan synthase alpha 2 beta 2 complex catalyzes tryptophan (Trp) biosynthesis from serine plus either indole (IN) or indole-3-glycerol phosphate (InGP). The photoreactive 5-azido analog in IN (AzIN), itself a substrate in the dark, was utilized to examine the substrate binding sites on this enzyme. When irradiated with AzIN at concentrations approaching IN saturation for the IN----Trp activity (0.1 mM), in the absence of serine, the enzyme was increasingly inactivated (up to 70-80%) concomitant with the progressive binding of a net of 2 mol AzIN per alpha beta equivalent. Little or no cooperativity in the binding of the 2 mol AzIN was observed. In contrast, there was minimal effect on the IN----InGP activity. Under these conditions AzIN appeared to be incorporated equally into each subunit. No significant inactivation nor binding occurred in the presence of serine. A quantitatively similar inactivation of InGP----Trp activity was observed over the same AzIN concentration range, suggesting common IN sites for Trp biosynthesis from either indole substrate. At higher concentrations (0.1-0.7 mM), no further inactivation occurred, although there was extensive additional binding (up to 10 mol/alpha beta equivalent). These data are consistent, although more clear-cut quantitatively, with the high- and low-affinity sites proposed from equilibrium dialysis studies. AzIN binding studies utilizing the isolated beta 2 subunit confirmed earlier reports suggesting the existence of many nonspecific IN binding sites on this subunit.     

Selective proteolysis of the beta 2 subunit of Serratia marcescens tryptophan synthase.

Mild digestion of Serratia marcescens tryptophan synthase β₂ subunit produces a modified β₂ subunit (nicked β₂). The nicked β₂ subunit remains essentially intact and is immunochemically reactive with native β₂ subunit antiserum. Denaturation of the nicked β₂ subunit yields two principal peptide fragments whose minimum molecular weights are 29,500 and 13,400. Loss of enzyme activity is associated with the selective proteolysis. The enzyme cofactor pyridoxal phosphate binding site is on the larger fragment. Following separation of the fragments by urea-gel chromatography, the separated peptides retain immunological cross-reactivity with native β₂ subunit antiserum. These fragments apparently represent two domains that comprise the native Holo β₂ subunit. The immunochemical data suggest that these fragments, when isolated, can assume some tertiary structure and that they may exist as such prior to β monomer or β₂ dimer assembly. The folded fragments may represent intermediates in the biosynthesis of the β₂ subunit as has been suggested for the E. coli enzyme (A. Högberg-Raibaud and M. E. Goldberg, 1977, Proc. Nat. Acad. Sci. USA74, 442; Biochemistry16, 4014).     

Role of diffusion in the folding of the alpha subunit of tryptophan synthase from Escherichia coli

The rate-limiting step in the folding of the alpha subunit of tryptophan synthase has been proposed to be the association of two folding units. To probe the role of diffusion in this rate-limiting step, the urea-induced unfolding and refolding of the protein was examined in the presence of a number of viscosity-enhancing agents. The analysis was simplified by studying the effect of these agents on folding unit dissociation, the rate-limiting unfolding reaction, and the reverse of the rate-limiting step in refolding. In the presence of ethylene glycol, the relaxation times for unfolding to the same final conditions increased with increasing concentration of the cosolvent. When the effects of the cosolvent on protein stability were taken into account, the rates were found to show a unitary linear dependence on the viscosity of the solution. Similar results were obtained with glycerol and low concentrations of glucose, demonstrating that the effect is general and not specific to any viscogenic agent. These results clearly demonstrate that the rate-limiting folding unit association/dissociation reaction in the alpha subunit of tryptophan synthase involves a diffusional process.     

Homodimers of mutant tryptophan synthase alpha-subunits in Escherichia coli.

Tryptophan synthase alpha-subunit from Escherichia coli functionally exists as a heterotetramer of alpha(2)beta(2) with beta-subunit. While wild-type and mutant (F139W, T24M/F139W, and T24L/F139W) alpha-subunits were expressed as a monomer from recombinant plasmids in Escherichia coli, T24A/F139W, T24S/F139W, and T24K/F139W mutant alpha-subunits were abnormally expressed as soluble homodimers in addition to monomers. Monomers of dimer-forming mutant alpha-subunits retain high affinity to beta-subunit, high activity in stimulating catalytic activities of beta-subunit, and nearly intact content of secondary structure, indicating that the global structures of these monomers are identical to that of F139W alpha-subunit. However, fluorescence spectra of Trp139 and ANS binding indicate that significant perturbations occur in the mutant proteins. Interestingly, these defective properties of monomers caused by residue replacement were partially repaired by the dimer formation. As a result, it is suggested that dimers may be formed by domain or loop swapping, and that residue 24 may play important role in maintaining on-pathway of alpha-subunit folding.