Identification of preferred substrate sequences of microbial transglutaminase from Streptomyces mobaraensis using a phage-displayed peptide library

Microbial transglutaminase (TGase) from Streptomyces mobaraensis (MTG) has been used in many industrial applications because it effectively catalyzes the formation of covalent cross-linking between glutamine residues in various substrate proteins and lysine residues or primary amines. To better understand the sequence preference around the reactive glutamine residue by this enzymatic reaction, we screened preferred peptide sequences using a phage-displayed random peptide library. Most of the peptides identified contained a consensus sequence, which was different from those previously found for mammalian TGases. Of these, most sequences had a specific reactivity toward MTG when produced as a fusion protein with glutathione-S-transferase. Furthermore, the representative sequence was found to be reactive even in the peptide form. The amino acid residues in the sequence critical for the reactivity were further analyzed, and the possible interaction with the enzyme has been discussed in this paper.     

Illuminating structure and acyl donor sites of a physiological transglutaminase substrate from Streptomyces mobaraensis

Transglutaminase from Streptomyces mobaraensis (MTG) has become a powerful tool to covalently and highly specifically link functional amines to glutamine donor sites of therapeutic proteins. However, details regarding the mechanism of substrate recognition and interaction of the enzyme with proteinaceous substrates still remain mostly elusive. We have determined the crystal structure of the Streptomyces papain inhibitory protein (SPI_p), a substrate of MTG, to study the influence of various substrate amino acids on positioning glutamine to the active site of MTG. SPI_p exhibits a rigid, thermo‐resistant double‐psi‐beta‐barrel fold that is stabilized by two cysteine bridges. Incorporation of biotin cadaverine identified Gln‐6 as the only amine acceptor site on SPI_p accessible for MTG. Substitution of Lys‐7 demonstrated that small and hydrophobic residues in close proximity to Gln‐6 favor MTG‐mediated modification and are likely to facilitate introduction of the substrate into the front vestibule of MTG. Moreover, exchange of various surface residues of SPI_p for arginine and glutamate/aspartate outside the glutamine donor region influences the efficiency of modification by MTG. These results suggest the occurrence of charged contact areas between MTG and the acyl donor substrates beyond the front vestibule, and pave the way for protein engineering approaches to improve the properties of artificial MTG‐substrates used in biomedical applications.     

The glutamine residues reactive in transglutaminase-catalyzed cross-linking of involucrin

The protein involucrin, synthesized by human keratinocytes, contains 585 amino acids, largely in the form of 10 amino acid repeats, each containing glutamines in 3 conserved positions. Involucrin is a substrate for the keratinocyte transglutaminase and is labeled by the cosubstrate amine, glycine ethyl ester. Study of tryptic peptides of involucrin shows that a single glutamine (residue 496), located 89 residues from the C-terminal end, is preferentially labeled by the enzyme. Additional glutamine residues become reactive when the molecule is fragmented. The C-terminal end, isolated as a cyanogen bromide fragment of 275 residues, is labeled equally at 2 glutamine residues. The polypeptide containing residues 148 to 280 accepts practically no amine while in intact involucrin but as a free fragment is labeled at multiple glutamine residues. It is concluded that the C-terminal and N-terminal ends of the protein are directive influences in that they suppress the reactivity of a number of glutamine residues in the intact molecule, leaving one glutamine highly preferred by the transglutaminase.     

Microbial transglutaminase displays broad acyl-acceptor substrate specificity

The great importance of amide bonds in industrial synthesis has encouraged the search for efficient catalysts of amide bond formation. Microbial transglutaminase (MTG) is heavily utilized in crosslinking proteins in the food and textile industries, where the side chain of a glutamine reacts with the side chain of a lysine, forming a secondary amide bond. Long alkylamines carrying diverse chemical entities can substitute for lysine as acyl-acceptor substrates, to link molecules of interest onto peptides or proteins. Here, we explore short and chemically varied acyl-acceptor substrates, to better understand the nature of nonnatural substrates that are tolerated by MTG, with the aim of diversifying biocatalytic applications of MTG. We show, for the first time, that very short-chain alkyl-based amino acids such as glycine can serve as acceptor substrates. The esterified α-amino acids Thr, Ser, Cys, and Trp—but not Ile—also showed reactivity. Extending the search to nonnatural compounds, a ring near the amine group—particularly if aromatic—was beneficial for reactivity, although ring substituents reduced reactivity. Overall, amines attached to a less hindered carbon increased reactivity. Importantly, very small amines carrying either the electron-rich azide or the alkyne groups required for click chemistry were highly reactive as acyl-acceptor substrates, providing a robust route to minimally modified, “clickable” peptides. These results demonstrate that MTG is tolerant to a variety of chemically varied natural and nonnatural acyl-acceptor substrates, which broadens the scope for modification of Gln-containing peptides and proteins.     

Identification of Transglutaminase Reactive Residues in Human Osteopontin and Their Role in Polymerization

Osteopontin (OPN) is a highly posttranslationally modified protein present in several tissues where it is implicated in numerous physiological processes. OPN primarily exerts its functions through interaction with integrins via the Arg-Gly-Asp and Ser-Val-Val-Tyr-Gly-Leu-Arg sequences located in the N-terminal part of the protein. OPN can be polymerized by the cross-linking enzyme transglutaminase 2 (TG2), and polymerization has been shown to enhance the biological activity of OPN. However, little is known about the reactivity and location of the glutamine and lysine residues involved in the TG2-mediated modification of OPN. Here we show that TG2 catalyses the incorporation of 5-(Biotinamido)pentylamine at glutamines in both the N- and C-terminal parts of OPN, whereas TG2 primarily incorporated the glutamine-donor peptide biotinyl-TVQQEL-OH into the C-terminal part of OPN. By mass spectrometric analyses we identified Gln34, Gln42, Gln193 and Gln248 as the major TG2 reactive glutamines in OPN. The distribution of reactive Gln and Lys residues in OPN proved to be important, as the full-length protein but not the physiologically highly active integrin-binding N-terminal part of OPN were able to polymerize in a TG2-mediated reaction. Collectively, these data provide important new molecular knowledge about the mechanism of OPN polymerization.     

Transglutaminase-mediated internal protein labeling with a designed peptide loop

Post-translational internal protein labeling was explored through the insertion of a 13-mer peptidyl loop specifically recognized by microbial transglutaminase (MTG). The peptidyl loop included one lysine residue (abbreviated as the K-loop), and was designed and inserted into two different regions of the protein bacterial alkaline phosphatase (BAP). MTG-mediated selective labeling of a lysine residue in the K-loop was achieved with a functional Gln-donor substrate. Internal protein labeling in the vicinity of the active site of BAP (residues 91–93) markedly decreased the activity of the enzyme. Conversely, insertion of the K-loop at a site distal from the active site (residues 219–221) afforded site-specific and covalent internal protein labeling without impairing the activity of the enzyme.     

α2-Macroglobulin-like protein 1 can conjugate and inhibit proteases through their hydroxyl groups,because of an enhanced reactivity of its thiol ester

Proteins in the α-macroglobulin (αM) superfamily use thiol esters to form covalent conjugation products upon their proteolytic activation. αM protease inhibitors use theirs to conjugate proteases and preferentially react with primary amines (e.g. on lysine side chains), whereas those of αM complement components C3 and C4B have an increased hydroxyl reactivity that is conveyed by a conserved histidine residue and allows conjugation to cell surface glycans. Human α₂-macroglobulin–like protein 1 (A2ML1) is a monomeric protease inhibitor but has the hydroxyl reactivity–conveying histidine residue. Here, we have investigated the role of hydroxyl reactivity in a protease inhibitor by comparing recombinant WT A2ML1 and the A2ML1 H1084N mutant in which this histidine is removed. Both of A2ML1s'' thiol esters were reactive toward the amine substrate glycine, but only WT A2ML1 reacted with the hydroxyl substrate glycerol, demonstrating that His-1084 increases the hydroxyl reactivity of A2ML1''s thiol ester. Although both A2ML1s conjugated and inhibited thermolysin, His-1084 was required for the conjugation and inhibition of acetylated thermolysin, which lacks primary amines. Using MS, we identified an ester bond formed between a thermolysin serine residue and the A2ML1 thiol ester. These results demonstrate that a histidine-enhanced hydroxyl reactivity can contribute to protease inhibition by an αM protein. His-1084 did not improve A2ML1''s protease inhibition at pH 5, indicating that A2ML1''s hydroxyl reactivity is not an adaption to its acidic epidermal environment.     

Analysis of amino acid substitutions in AraC variants that respond to triacetic acid lactone

The Escherichia coli regulatory protein AraC regulates expression of ara genes in response to l ‐arabinose. In efforts to develop genetically encoded molecular reporters, we previously engineered an AraC variant that responds to the compound triacetic acid lactone (TAL). This variant (named “AraC‐TAL1”) was isolated by screening a library of AraC variants, in which five amino acid positions in the ligand‐binding pocket were simultaneously randomized. Screening was carried out through multiple rounds of alternating positive and negative fluorescence‐activated cell sorting. Here we show that changing the screening protocol results in the identification of different TAL‐responsive variants (nine new variants). Individual substituted residues within these variants were found to primarily act cooperatively toward the gene expression response. Finally, X‐ray diffraction was used to solve the crystal structure of the apo AraC‐TAL1 ligand‐binding domain. The resolved crystal structure confirms that this variant takes on a structure nearly identical to the apo wild‐type AraC ligand‐binding domain (root‐mean‐square deviation 0.93 Å), suggesting that AraC‐TAL1 behaves similar to wild‐type with regard to ligand recognition and gene regulation. Our results provide amino acid sequence–function data sets for training and validating AraC modeling studies, and contribute to our understanding of how to design new biosensors based on AraC.     

A method for site‐specific labeling of multiple protein thiols

We present a generic method for the site‐specific and differential labeling of multiple cysteine residues in one protein. Phenyl arsenic oxide has been employed as a protecting group of two closely spaced thiols, allowing first labeling of a single thiol. Subsequently, the protecting group is removed, making available a reactive dithiol site for labeling with a second probe. For proof‐of‐principle, single and triple Cys mutants of the sulphate binding protein of an ABC transporter were constructed. The closely spaced thiols were engineered on the basis of the crystal structure of the protein and placed in different types of secondary structure elements and at different spacing. We show that phenyl arsenic oxide is a good protecting group for thiols spaced 6.3–7.3 Å. Proteins were labeled with two different fluorescent labels and the labeling ratios were determined with UV‐Vis spectroscopy and MALDI‐Tof mass spectrometry. The average labeling efficiency was ～80% for the single thiol and 65–90% for the dithiol site.     

Substrate specificity analysis of microbial transglutaminase using proteinaceous protease inhibitors as natural model substrates

The substrate specificity of microbial transglutaminase (MTG) from Streptomyces mobaraensis (formerly categorized Streptoverticillium) was studied using a Streptomyces proteinaceous protease inhibitor, STI2, as a model amine-donor substrate. Chemical modification and mutational analysis to address the substrate requirements for MTG were carried out around the putative reactive site region of STI2 on the basis of the highly refined tertiary structure and the solvent accessibility index of Streptomyces subtilisin inhibitor, SSI, a homolog of STI2. The results suggest that the P1 reactive center site (position 70 of STI2) for protease subtilisin BPN' or trypsin may be the prime Lys residue that can be recognized by MTG, when succinylated beta-casein was used as a partner Gln-substrate. It is characteristic in that the same primary enzyme contact region of STI2 is shared by both enzymes, MTG and proteases. For quantitative analysis of the TG reaction, we established an ELISA-based monitoring assay system using an anti-SSI polyclonal antibody highly cross-reactive with STI2. Site-specific STI2 mutants were prepared by an Escherichia coli expression-secretion vector system and subjected to the assay system. We reached several conclusions concerning the nature of the flanking amino acid residues affecting the MTG reactivity of the substrate Lys residue: (i) site-specific mutations from Asn to Lys or Arg at position 69 preceding the amine-donor 70Lys, led to enhanced substrate reactivity; (ii) amino acid replacement at 67Ile with Ser led to higher substrate reactivity, (iii) additive effects were obtained by a combination of the positive mutations at positions 67 and 69 as described above, and (iv) Gly at position 65 might be essential for MTG reaction. Moreover, the substrate specificity of guinea pig liver tissue transglutaminase (GTG) was compared with that of MTG using STI2 and its mutants. In contrast to MTG, replacement of Gly by Asp at position 65 was the most favorable for substrate reactivity. Also, 70Lys appeared not to be a prime amine-donor site for GTG-mediated cross-linking, suggesting a difference in substrate recognition between MTG and GTG.     

Association of bovine β‐casein protein variant I with milk production and milk protein composition

The aim of this study was to detect new polymorphisms in the bovine β‐casein (β‐CN) gene and to evaluate association of (new) β‐CN protein variants with milk production traits and milk protein composition. Screening of the β‐CN gene in genomic DNA from 72 Holstein Friesian (HF) bulls resulted in detection of 19 polymorphisms and revealed the presence of β‐CN protein variant I in the Dutch HF population. Studies of association of β‐CN protein variants with milk composition usually do not discriminate protein variant I from variant A2. Association of β‐CN protein variants with milk composition was studied in 1857 first‐lactation HF cows and showed that associations of protein variants A2 and I were quite different for several traits. β‐CN protein variant I was significantly associated with protein percentage and protein yield, and with α_s1‐casein (α_s1‐CN), α_s2‐casein (α_s2‐CN), κ‐casein (κ‐CN), α‐lactalbumin (α‐LA), β‐lactoglobulin (β‐LG), casein index and casein yield. Inferring β‐κ‐CN haplotypes showed that β‐CN protein variant I occurred only with κ‐CN variant B. Consequently, associations of β‐κ‐CN haplotype IB with protein percentage, κ‐CN, α‐LA, β‐LG and casein index are likely resulting from associations of κ‐CN protein variant B, while associations of β‐κ‐CN haplotype IB with α_s1‐CN and α_s2‐CN seem to be resulting from associations of β‐CN variant I.     

Re‐engineering a β‐lactamase using prototype peptides from a library of local structural motifs

B. licheniformis exo‐small β‐lactamase (ESBL) has a complex architecture with twelve α helices and a five‐stranded beta sheet. We replaced, separately or simultaneously, three of the ESBL α helices with prototype amphiphatic helices from a catalog of secondary structure elements. Although the substitutes bear no sequence similarity to the originals and pertain to unrelated protein families, all the engineered ESBL variants were found able to fold in native like structures with in vitro and in vivo enzymic activity. The triple substituted variant resembles a primitive protein, with folding defects such as a strong tendency to oligomerization and very low stability; however it mimics a non homologous recombinant abandoning the family sequence space while preserving fold. The results test protein folding and evolution theories.     

An engineered viral protease exhibiting substrate specificity for a polyglutamine stretch prevents polyglutamine-induced neuronal cell death

Background

Polyglutamine (polyQ)-induced protein aggregation is the hallmark of a group of neurodegenerative diseases, including Huntington''s disease. We hypothesized that a protease that could cleave polyQ stretches would intervene in the initial events leading to pathogenesis in these diseases. To prove this concept, we aimed to generate a protease possessing substrate specificity for polyQ stretches.

Methodology/Principal Findings

Hepatitis A virus (HAV) 3C protease (3CP) was subjected to engineering using a yeast-based method known as the Genetic Assay for Site-specific Proteolysis (GASP). Analysis of the substrate specificity revealed that 3CP can cleave substrates containing glutamine at positions P5, P4, P3, P1, P2′, or P3′, but not substrates containing glutamine at the P2 or P1′ positions. To accommodate glutamine at P2 and P1′, key residues comprising the active sites of the S2 or S1′ pockets were separately randomized and screened. The resulting sets of variants were combined by shuffling and further subjected to two rounds of randomization and screening using a substrate containing glutamines from positions P5 through P3′. One of the selected variants (Var26) reduced the expression level and aggregation of a huntingtin exon1-GFP fusion protein containing a pathogenic polyQ stretch (HttEx1(97Q)-GFP) in the neuroblastoma cell line SH-SY5Y. Var26 also prevented cell death and caspase 3 activation induced by HttEx1(97Q)-GFP. These protective effects of Var26 were proteolytic activity-dependent.

Conclusions/Significance

These data provide a proof-of-concept that proteolytic cleavage of polyQ stretches could be an effective modality for the treatment of polyQ diseases.     

An enzymatic strategy for site-specific immobilization of functional proteins using microbial transglutaminase

A novel strategy for site-specific immobilization of recombinant proteins was investigated using microbial transglutaminase (MTG). Alkaline phosphatase (AP) was selected as a model protein and tagged with a short peptide (MKHKGS) at the N-terminus to provide a reactive Lys residue for MTG. On the other hand, casein, a well-known substrate for MTG, was chemically attached onto a polyacrylic resin to provide reactive Gln residues for the enzymatic immobilization of the recombinant AP. As a result, we succeeded in MTG-mediated functional immobilization of the recombinant AP onto casein-coated polyacrylic resin. It was found that the immobilized AP prepared using MTG exhibited much higher specific activity than that prepared by chemical modification. Moreover, enzymatic immobilization gave an immobilized formulation with higher stability upon repeated use than that obtained by physical adsorption. Use of this ability of MTG in posttranslational protein modification will provide us with a benign, site-specific immobilization method for functional proteins.     

A chemically modified glass surface that facilitates transglutaminase-mediated protein immobilization

An amino-modified glass surface for enzymatic protein immobilization by microbial transglutaminase (MTG) was developed. Diamine substrates with secondary amino groups in the linker moiety, like triethylenetetramine (TETA), exhibited at most a 2-fold higher reactivity in the MTG-catalyzed reaction compared to those with the alkyl linker. A 96-well glass plate was subsequently modified with selected diamine substrates. Validation of the modified surface by enzymatic immobilization of enhanced green fluorescent protein tagged with a glutamine donor-substrate peptide (LLQG) of MTG revealed that the protein loading onto the TETA-modified glass surface was approximately 15-fold higher than that on the unmodified one.     

A photostable monomeric superfolder green fluorescent protein

The green fluorescent protein (GFP) from Aequorea victoria has been engineered extensively in the past to generate variants suitable for protein tagging. Early efforts produced the enhanced variant EGFP and its monomeric derivative mEGFP, which have useful photophysical properties, as well as superfolder GFP, which folds efficiently under adverse conditions. We previously generated msGFP, a monomeric superfolder derivative of EGFP. Unfortunately, compared to EGFP, msGFP and other superfolder GFP variants show faster photobleaching. We now describe msGFP2, which retains monomeric superfolder properties while being as photostable as EGFP. msGFP2 contains modified N‐ and C‐terminal peptides that are expected to reduce nonspecific interactions. Compared to EGFP and mEGFP, msGFP2 is less prone to disturbing the functions of certain partner proteins. For general‐purpose protein tagging, msGFP2 may be the best available derivative of A. victoria GFP.     

Synthesis of pathological and nonpathological human exon 1 huntingtin

Huntington's disease (HD) is a neurodegenerative disorder that affects approximately 1 in 10 000 individuals. The underlying gene mutation was identified as a CAG‐triplet repeat expansion in the gene huntingtin. The CAG sequence codes for glutamine, and in HD, an expansion of the polyglutamine (poly‐Q) stretch above 35 glutamine residues results in pathogenicity. It has been demonstrated in various animal models that only the expression of exon 1 huntingtin, a 67‐amino acid‐long polypeptide plus a variable poly‐Q stretch, is sufficient to cause full HD‐like pathology. Therefore, a deeper understanding of exon 1 huntingtin, its structure, aggregation mechanism and interaction with other proteins is crucial for a better understanding of the disease. Here, we describe the synthesis of a 109‐amino acid‐long exon 1 huntingtin peptide including a poly‐Q stretch of 42 glutamines. This microwave‐assisted solid phase peptide synthesis resulted in milligram amounts of peptide with high purity. We also synthesized a nonpathogenic version of exon 1 huntingtin (90‐amino acid long including a poly‐Q stretch of 23 glutamine residues) using the same strategy. In circular dichroism spectroscopy, both polypeptides showed weak alpha‐helical properties with the longer peptide showing a higher helical degree. These model peptides have great potential for further biomedical analyses, e.g. for large‐scale pre‐screenings for aggregation inhibitors, further structural analyses as well as protein–protein interaction studies. Copyright © 2010 European Peptide Society and John Wiley & Sons, Ltd.     

In vitro characterization of the intramelanosomal domain of human recombinant TYRP1 and its oculocutaneous albinism type 3-related mutant variants

Tyrosinase related protein 1 (TYRP1) is the most abundant melanosomal protein of the melanocyte, where plays an important role in the synthesis of eumelanin, possibly catalyzing the oxidation of 5,6-dihydroxyindole-2-carboxylic acid to 5,6-quinone-2-carboxylic acid. Mutations to the TYRP1 gene can result in oculocutaneous albinism type 3 (OCA3), a rare disease characterized by reduced synthesis of melanin in skin, hair, and eyes. To investigate the effect of genetic mutations on the TYRP1 structure, function, and stability, we engineered the intramelanosomal domain of TYRP1 and its mutant variants mimicking either OCA3-related changes, C30R, H215Y, D308N, and R326H or R87G mutant variant, analogous to OCA1-related pathogenic effect in tyrosinase. Proteins were produced in Trichoplusia Ni larvae, then purified, and analyzed by biochemical methods. Data shows that D308N and R326H mutants keep the native conformations and demonstrate no change in their stability and enzymatic activity. In contrast, mutations C30R and R87G localized in the Cys-rich domain show the variants misfolding during the purification process. The H215Y variant disrupts the binding of Zn²⁺ in the active site and thus reduces the strength of the enzyme/substrate interactions. Our results, consistent with the clinical and in silico studies, show that mutations at the protein surface are expected to have a negligible phenotype change compared to that of TYRP1. For the mutations with severe phenotype changes, which were localized in the Cys-rich domain or the active site, we confirmed a complete or partial protein misfolding as the possible mechanism of protein malfunction caused by OCA3 inherited mutations.     

Design of a specific peptide tag that affords covalent and site-specific enzyme immobilization catalyzed by microbial transglutaminase

Transglutaminase-mediated site-specific and covalent immobilization of an enzyme to chemically modified agarose was explored. Using Escherichia coli alkaline phosphatase (AP) as a model, two designed specific peptide tags containing a reactive lysine (Lys) residue with different length Gly-Ser linkers for microbial transglutaminase (MTG) were genetically attached to N- or C-termini. For solid support, agarose gel beads were chemically modified with beta-casein to display reactive glutamine (Gln) residues on the support surface. Recombinant APs were enzymatically and covalently immobilized to casein-grafted agarose beads. Immobilization by MTG markedly depended on either the position or the length of the peptide tags incorporated to AP, suggesting steric constraint upon enzymatic immobilization. Enzymatically immobilized AP showed comparable catalytic turnover (k(cat)) to the soluble counterpart and comparable operational stability with chemically immobilized AP. These results indicate that attachment of a suitable specific peptide tag to the right position of a target protein is crucial for MTG-mediated formulation of highly active immobilized proteins.     

Labeling of specific lysine residues at the active site of glutamine synthetase

Glutamine synthetase (Escherichia coli) was incubated with three different reagents that react with lysine residues, viz. pyridoxal phosphate, 5'-p-fluorosulfonylbenzoyladenosine, and thiourea dioxide. The latter reagent reacts with the epsilon-nitrogen of lysine to produce homoarginine as shown by amino acid analysis, nmr, and mass spectral analysis of the products. A variety of differential labeling experiments were conducted with the above three reagents to label specific lysine residues. Thus pyridoxal phosphate was found to modify 2 lysine residues leading to an alteration of catalytic activity. At least 1 lysine residue has been reported previously to be modified by pyridoxal phosphate at the active site of glutamine synthetase (Whitley, E. J., and Ginsburg, A. (1978) J. Biol. Chem. 253, 7017-7025). By varying the pH and buffer, one or both residues could be modified. One of these lysine residues was associated with approximately 81% loss in activity after modification while modification of the second lysine residue led to complete inactivation of the enzyme. This second lysine was found to be the residue which reacted specifically with the ATP affinity label 5'-p-fluorosulfonylbenzoyladenosine. Lys-47 has been previously identified as the residue that reacts with this reagent (Pinkofsky, H. B., Ginsburg, A., Reardon, I., Heinrikson, R. L. (1984) J. Biol. Chem. 259, 9616-9622; Foster, W. B., Griffith, M. J., and Kingdon, H. S. (1981) J. Biol. Chem. 256, 882-886). Thiourea dioxide inactivated glutamine synthetase with total loss of activity and concomitant modification of a single lysine residue. The modified amino acid was identified as homoarginine by amino acid analysis. The lysine residue modified by thiourea dioxide was established by differential labeling experiments to be the same residue associated with the 81% partial loss of activity upon pyridoxal phosphate inactivation. Inactivation with either thiourea dioxide or pyridoxal phosphate did not affect ATP binding but glutamate binding was weakened. The glutamate site was implicated as the site of thiourea dioxide modification based on protection against inactivation by saturating levels of glutamate. Glutamate also protected against pyridoxal phosphate labeling of the lysine consistent with this residue being the common site of reaction with thiourea dioxide and pyridoxal phosphate.