Molecular recognition between Azotobacter vinelandii rhodanese and a sulfur acceptor protein

The occurrence of rhodanese-like proteins in the major evolutionary phyla, together with the observed abundance of these proteins also within the same genome, suggests that their function cannot be limited to cyanide scavenging. The aim of the present study was to investigate whether Azotobacter vinelandii RhdA, an enzyme possessing unique biochemical and structural features with respect to other members of rhodanese homology superfamily, could recognize a suitable protein as a potential acceptor of the sulfane sulfur held on its catalytic Cys residue. Both the potential sulfur-delivery RhdA-S and the sulfur-deprived RhdA were found to interact with either holo- or apo-adrenodoxin, the 'substrate' protein used in this work. Interaction of RhdA-S with apo-adrenodoxin led to mobilization of RhdA-S sulfane sulfur. Under appropriate conditions, the sulfur released from RhdA-S was productively used for 2Fe-2S cluster reconstitution to yield holo-adrenodoxin from apo-adrenodoxin in the absence of any other sulfur source. A comparison of the reactivity of RhdA-S with protein and non-protein thiols allowed also some insights into the accessibility of the sulfane sulfur carried by RhdA.     

Fast conformational exchange between the sulfur-free and persulfide-bound rhodanese domain of E. coli YgaP

Rhodanese domains are abundant structural modules that catalyze the transfer of a sulfur atom from thiolsulfates to cyanide via formation of a covalent persulfide intermediate that is bound to an essential conserved cysteine residue. In this study, the three-dimensional structure of the rhodanese domain of YgaP from Escherichia coli was determined using solution NMR. A typical rhodanese domain fold was observed, as expected from the high homology with the catalytic domain of other sulfur transferases. The initial sulfur-transfer step and formation of the rhodanese persulfide intermediate were monitored by addition of sodium thiosulfate using two-dimensional ¹H–¹⁵N correlation spectroscopy. Discrete sharp signals were observed upon substrate addition, indicting fast exchange between sulfur-free and persulfide-intermediate forms. Residues exhibiting pronounced chemical shift changes were mapped to the structure, and included both substrate binding and surrounding residues.     

Common themes and variations in the rhodanese superfamily

The rhodanese homology domain is a ubiquitous fold found in several phylogenetically related proteins encoded by eubacterial, archeal, and eukaryotic genomes. Although rhodanese-like proteins share evolutionary relationships, analysis of their sequences highlights that they are so heterogeneous to form the rhodanese superfamily. The variability occurs at different levels including sequence, active site loop length, presence of a critical catalytic Cys residue, and domain arrangement. Even within the same genome, multiple genes encode rhodanese-like proteins presenting with variably arranged rhodanese domain(s): as single or tandem domain(s), or combined with other protein domain(s). Given the highly variable organization of the rhodanese domain(s) and the context where it is found, here we review the structural organization and function of the rhodanese-like proteins. The overview of the most recent findings about rhodanese allow us to depict a superfamily of versatile proteins relying on persulfide chemistry to accomplish cellular functions spanning from resistance to environmental threats, such as cyanide, and key cellular reactions related to sulfur metabolism and progression of cell cycle.     

Identification of putative sulfurtransferase genes in the extremophilic Acidithiobacillus ferrooxidans ATCC 23270 genome: structural and functional characterization of the proteins

Eight nucleotide sequences containing a single rhodanese domain were found in the Acidithiobacillus ferrooxidans ATCC 23270 genome: p11, p14, p14.3, p15, p16, p16.2, p21, and p28. Amino acids sequence comparisons allowed us to identify the potentially catalytic Cys residues and other highly conserved rhodanese family features in all eight proteins. The genomic contexts of some of the rhodanese-like genes and the determination of their expression at the mRNA level by using macroarrays suggested their implication in sulfur oxidation and metabolism, formation of Fe-S clusters or detoxification mechanisms. Several of the putative rhodanese genes were successfully isolated, cloned and overexpressed in E. coli and their thiosulfate:cyanide sulfurtransferase (TST) and 3-mercaptopyruvate/cyanide sulfurtransferase (MST) activities were determined. Based on their sulfurtransferase activities and on structural comparisons of catalytic sites and electrostatic potentials between homology- modeled A. ferrooxidans rhodaneses and the reported crystal structures of E. coli GlpE (TST) and SseA (MST) proteins, two of the rhodanese-like proteins (P15 and P16.2) could clearly be defined as TSTs, and P14 and P16 could possibly correspond to MSTs. Nevertheless, several of the eight A. ferrooxidans rhodanese-like proteins may have some different functional activities yet to be discovered.     

The crystal structure of a sulfurtransferase from Azotobacter vinelandii highlights the evolutionary relationship between the rhodanese and phosphatase enzyme families

Rhodanese is an ubiquitous enzyme that in vitro catalyses the transfer of a sulfur atom from suitable donors to nucleophilic acceptors by way of a double displacement mechanism. During the catalytic process the enzyme cycles between a sulfur-free and a persulfide-containing form, via formation of a persulfide linkage to a catalytic Cys residue. In the nitrogen-fixing bacteria Azotobacter vinelandii the rhdA gene has been identified and the encoded protein functionally characterized as a rhodanese. The crystal structure of the A. vinelandii rhodanese has been determined and refined at 1.8 A resolution in the sulfur-free and persulfide-containing forms. Conservation of the overall three-dimensional fold of bovine rhodanese is observed, with substantial modifications of the protein structure in the proximity of the catalytic residue Cys230. Remarkably, the native enzyme is found as the Cys230-persulfide form; in the sulfur-free state the catalytic Cys residue adopts two alternate conformations, reflected by perturbation of the neighboring active-site residues, which is associated with a partly reversible loss of thiosulfate:cyanide sulfurtransferase activity. The catalytic mechanism of A. vinelandii rhodanese relies primarily on the main-chain conformation of the 230 to 235 active-site loop and on a surrounding strong positive electrostatic field. Substrate recognition is based on residues which are entirely different in the prokaryotic and eukaryotic enzymes. The active-site loop of A. vinelandii rhodanese displays striking structural similarity to the active-site loop of the similarly folded catalytic domain of dual specific phosphatase Cdc25, suggesting a common evolutionary origin of the two enzyme families.     

The "rhodanese" fold and catalytic mechanism of 3-mercaptopyruvate sulfurtransferases: crystal structure of SseA from Escherichia coli

3-Mercaptopyruvate sulfurtransferases (MSTs) catalyze, in vitro, the transfer of a sulfur atom from substrate to cyanide, yielding pyruvate and thiocyanate as products. They display clear structural homology with the protein fold observed in the rhodanese sulfurtransferase family, composed of two structurally related domains. The role of MSTs in vivo, as well as their detailed molecular mechanisms of action have been little investigated. Here, we report the crystal structure of SseA, a MST from Escherichia coli, which is the first MST three-dimensional structure disclosed to date. SseA displays specific structural differences relative to eukaryotic and prokaryotic rhodaneses. In particular, conformational variation of the rhodanese active site loop, hosting the family invariant catalytic Cys residue, may support a new sulfur transfer mechanism involving Cys237 as the nucleophilic species and His66, Arg102 and Asp262 as residues assisting catalysis.     

Surface changes and role of buried water molecules during the sulfane sulfur transfer in rhodanese from Azotobacter vinelandii: a fluorescence quenching and nuclear magnetic relaxation dispersion spectroscopic study

The Azotobacter vinelandii rhodanese is a sulfurtransferase enzyme that catalyzes the transfer of the outer sulfur atom from thiosulfate to cyanide. Recently, investigations by NMR relaxation on the (15)N-enriched protein reported that interdomain contacts are rigidly maintained upon the sulfane sulfur transfer from the enzyme to the substrate. The modality of the enzymatic mechanism is then confined to a surface interaction, including dynamics of water molecules buried in the tertiary structure. Thus, investigations have been carried out by fluorescence, circular dichroism, and nuclear magnetic relaxation dispersion measurements. The comparison of circular dichroism spectra of the persulfurated enzyme and the sulfur-free form indicated that small changes occur. Fluorescence quenching studies have been performed to evaluate the conformational changes during catalysis using the fluorescent probe 8-anilinonaphthalene-2-sulfonic acid, and acrylamide, iodide, and cesium ions as quenchers. Changes in exchange dynamics of water molecules buried in the structure with bulk water, observed by nuclear magnetic relaxation dispersion, are due to local conformational transitions, likely involving residues around the active site, and are consistent with the global correlation time found by (15)N relaxation. These results, taken together, provide important information for elucidating the conformational features of the mechanism of action of the enzyme either in the role of a selective donor of a sulfur atom to small-sized substrates (i.e., to cyanide, transforming it into thiocyanate) or in the role of sulfur insertase for the formation of the Fe(2)S(2) iron-sulfur cluster in sulfur-deprived ferredoxins.     

Structural rearrangements of the two domains of Azotobacter vinelandii rhodanese upon sulfane sulfur release: essential molecular dynamics, 15N NMR relaxation and deuterium exchange on the uniformly labeled protein

The Azotobacter vinelandii rhodanese is a 31 kDa sulfurtransferase protein that catalyzes the transfer of sulfur atom from thiosulfate to cyanide in the detoxification process from cyanide and is able to insert sulfur atom in the iron–sulfur cluster. A study of the uniformly ¹⁵N isotopic labeling by high resolution NMR, before obtaining the backbone sequential assignment, has been carried out. The sulfur loaded and the sulfur discharged forms of the enzyme show very similar HSQC spectra with a good spectral dispersion. Few resonances show changes in chemical shift between the two forms. Relaxation parameters T₁, T₂ and ¹H–¹⁵N NOE of all amide nitrogen atoms, as well as isotope exchange kinetics, show that the two forms exhibit the same global correlation time and hydrodynamic properties. In parallel, essential dynamics studies show that formation and discharging of catalytic cysteine persulfide group has no significant impact on the overall conformation of the protein. These results, taken together, give a clearcut answer to the question if the catalytic mechanism of the enzyme involves a change in the conformation and/or in the mutual orientation of the two domains. On the contrary these results clearly indicate that upon the catalytic mechanism the two domains of the protein behave as a unique fold.     

Site-directed mutagenesis of the active site loop of the rhodanese-like domain of the human molybdopterin synthase sulfurase MOCS3. Major differences in substrate specificity between eukaryotic and bacterial homologs

Sequence alignments of human molybdopterin synthase sulfurase, MOCS3, showed that the N-terminal domain is homologous to Escherichia coli MoeB, whereas the C-terminal domain is homologous to rhodanese-like proteins. Previous studies showed that the activity of the separately purified rhodanese-like domain of MOCS3 displayed 1000-fold lower activity in comparison to bovine rhodanese with thiosulfate as sulfur source. When the six amino acid active site loop of MOCS3 rhodanese-like domain was exchanged with the loop found in bovine rhodanese, thiosulfate:cyanide sulfurtransferase activity was increased 165-fold. Site-directed mutagenesis of each individual residue of the active site loop of the MOCS3 rhodanese-like domain showed that the charge of the last amino acid determines thiosulfate sulfurtransferase activity. Replacing Asp417 by threonine resulted in 90-fold increased activity, whereas replacing it by arginine increased the activity 470-fold. Using a fully defined in vitro system containing precursor Z, MOCS2A, E. coli MoaE, E. coli MoeB, Mg-ATP, MOCS3 rhodanese-like domain, and thiosulfate, it was shown that sulfur transfer to MOCS2A was also affected by the alterations, but not as drastically. Our studies revealed that in humans and most eukaryotes thiosulfate is not the physiologic sulfur donor for MOCS3, whereas in bacterial homologs, which have an arginine at the last position of the active site loop, thiosulfate can be used as a sulfur source for molybdenum cofactor biosynthesis. The phylogenetic analysis of MoeB homologs showed that eukaryotic homologs are of bacterial origin. Furthermore, it could be shown that an MoeB homolog named MoeZ, where the dual CXXC zinc-binding motif of the MoeB domain is not present, arose independently several times during evolution.     

Characterization of a rhodanese from the cyanogenic bacterium Pseudomonas aeruginosa

Pseudomonas aeruginosa, the rRNA group I type species of genus Pseudomonas, is a Gram-negative, aerobic bacterium responsible for serious infection in humans. P. aeruginosa pathogenicity has been associated with the production of several virulence factors, including cyanide. Here, the biochemical characterization of recombinant P. aeruginosa rhodanese (Pa RhdA), catalyzing the sulfur transfer from thiosulfate to a thiophilic acceptor, e.g., cyanide, is reported. Sequence homology analysis of Pa RhdA predicts the sulfur-transfer reaction to occur through persulfuration of the conserved catalytic Cys230 residue. Accordingly, the titration of active Pa RhdA with cyanide indicates the presence of one extra sulfur bound to the Cys230 Sgamma atom per active enzyme molecule. Values of K(m) for thiosulfate binding to Pa RhdA are 1.0 and 7.4mM at pH 7.3 and 8.6, respectively, and 25 degrees C. However, the value of K(m) for cyanide binding to Pa RhdA (=14 mM, at 25 degrees C) and the value of V(max) (=750 micromol min(-1)mg(-1), at 25 degrees C) for the Pa RhdA-catalyzed sulfur-transfer reaction are essentially pH- and substrate-independent. Therefore, the thiosulfate-dependent Pa RhdA persulfuration is favored at pH 7.3 (i.e., the cytosolic pH of the bacterial cell) rather than pH 8.6 (i.e., the standard pH for rhodanese activity assay). Within this pH range, conformational change(s) occur at the Pa RhdA active site during the catalytic cycle. As a whole, rhodanese may participate in multiple detoxification mechanisms protecting P. aeruginosa from endogenous and environmental cyanide.     

A persulfurated cysteine promotes active site reactivity in Azotobacter vinelandii Rhodanese

Active site reactivity and specificity of RhdA, a thiosulfate:cyanide sulfurtransferase (rhodanese) from Azotobacter vinelandii, have been investigated through ligand binding, site-directed mutagenesis, and X-ray crystallographic techniques, in a combined approach. In native RhdA the active site Cys230 is found persulfurated; fluorescence and sulfurtransferase activity measurements show that phosphate anions interact with Cys230 persulfide sulfur atom and modulate activity. Crystallographic analyses confirm that phosphate and hypophosphite anions react with native RhdA, removing the persulfide sulfur atom from the active site pocket. Considering that RhdA and the catalytic subunit of Cdc25 phosphatases share a common three-dimensional fold as well as active site Cys (catalytic) and Arg residues, two RhdA mutants carrying a single amino acid insertion at the active site loop were designed and their phosphatase activity tested. The crystallographic and functional results reported here show that specific sulfurtransferase or phosphatase activities are strictly related to precise tailoring of the catalytic loop structure in RhdA and Cdc25 phosphatase, respectively.     

15N NMR spectroscopy of hydrogen-bonding interactions in the active site of serine proteases: evidence for a moving histidine mechanism

Nitrogen-15 NMR spectroscopy has been used to study the hydrogen-bonding interactions involving the histidyl residue in the catalytic triad of alpha-lytic protease in the resting enzyme and in the transition-state or tetrahedral intermediate analogue complexes formed with phenylmethanesulfonyl fluoride and diisopropyl fluorophosphate. The 15N shifts indicate that a strong hydrogen bond links the active site histidine and serine residues in the resting enzyme in solution. This result is at odds with interpretations of the X-ray diffraction data of alpha-lytic protease and of other serine proteases, which indicate that the serine and histidine residues are too far apart and not properly aligned for the formation of a hydrogen bond. In addition, the nitrogen-15 shifts demonstrate that protonation of the histidine imidazole ring at low pH in the transition-state or tetrahedral intermediate analogue complexes formed with phenylmethanesulfonyl fluoride and diisopropyl fluorophosphate triggers the disruption of the aspartate-histidine hydrogen bond. These results suggest a catalytic mechanism involving directed movement of the imidazole ring of the active site histidyl residue.     

Solution NMR Structure and Functional Analysis of the Integral Membrane Protein YgaP from Escherichia coli

The solution NMR structure of the α-helical integral membrane protein YgaP from Escherichia coli in mixed 1,2-diheptanoyl-sn-glycerol-3-phosphocholine/1-myristoyl-2-hydroxy-sn-glycero-3-phospho-(1′-rac-glycerol) micelles is presented. In these micelles, YgaP forms a homodimer with the two transmembrane helices being the dimer interface, whereas the N-terminal cytoplasmic domain includes a rhodanese-fold in accordance to its sequence homology to the rhodanese family of sulfurtransferases. The enzymatic sulfur transfer activity of full-length YgaP as well as of the N-terminal rhodanese domain only was investigated performing a series of titrations with sodium thiosulfate and potassium cyanide monitored by NMR and EPR. The data indicate the thiosulfate concentration-dependent addition of several sulfur atoms to the catalytic Cys-63, which process can be reversed by the addition of potassium cyanide. The catalytic reaction induces thereby conformational changes within the rhodanese domain, as well as on the transmembrane α-helices of YgaP. These results provide insights into a potential mechanism of YgaP during the catalytic thiosulfate activity in vivo.     

The single cysteine residue of the Sud protein is required for its function as a polysulfide-sulfur transferase in Wolinella succinogenes.

The periplasmic Sud protein which is induced in Wolinella succinogenes growing by polysulfide respiration, has been previously proposed to serve as a polysulfide binding protein and to transfer polysulfide-sulfur to the active site of polysulfide reductase [Klimmek, O, Kreis, V., Klein, C., Simon, J., Wittershagen, A. & Kr?ger, A. (1998) Eur. J. Biochem. 253, 263-269.]. The results presented in this communication suggest that polysulfide-sulfur is covalently bound to the single cysteine residue (Cys109) of the Sud monomer, and that Cys109 is required for tight binding of polysulfide-sulfur and for sulfur transfer. A modified Sud protein [(C109S)Sud-His6] in which the cysteine residue was replaced by serine, did not catalyze sulfur transfer from polysulfide to cyanide and did not stimulate electron transport to polysulfide, in contrast to Sud-His6. The polysulfide-sulfur bound to (C109S)Sud-His6 was fully removed upon dialysis against sulfide. After this treatment, Sud-His6 retained one sulfur atom per monomer; thiocyanate was formed upon addition of cyanide to the preparation. After incubation of Sud-His6 with polysulfide, a proportion of the Sud-His6 monomers carried one or two sulfur atoms, as shown by matrix-assisted laser desorption ionization mass spectrometry. The sulfur atoms were absent from monomers derived from Sud-His6 treated with cyanide and from (C109S)Sud-His6 incubated with polysulfide.     

1H, 13C and 15N resonance assignments of rhodanese GlpE from Escherichia coli

Rhodanese catalyzes the sulfur-transfer reaction in which a sulfur atom is transferred from thiosulfate to cyanide by a double-displacement mechanism. During the reaction, a persulfide-intermediate form of rhodanese is generated by the reaction of a conserved active cysteine residue with thiosulfate. Escherichia coli GlpE is a prototype for the single-domain rhodanese superfamily. Though there are some studies on rhodaneses, the molecular mechanism of the catalytic activity of rhodaneses is still unclear. Herein, we report the resonance assignments of (1)H, (13)C and (15)N atoms of E. coli GlpE, which provides the basis for further structural, dynamic and functional studies of rhodaneses using NMR technique.     

The conserved Cys76 plays a crucial role for the conformation of reduced glutathione peroxidase-type tryparedoxin peroxidase

The crystal structure of reduced tryparedoxin peroxidase shows Cys47 close to Gln82 and Trp137 and helix formation of residues 87 to 97 whereas the NMR structure of the reduced C76S mutant adopts a different conformation similar to the oxidized protein. Circular dichroism (CD), fluorescence and NMR spectroscopy reveal that the fully active C76S mutant differs from the wildtype (WT) enzyme mainly in its reduced form both in secondary structure content and Trp137 environment. This implies that Cys76 plays a critical role for the reduced enzyme assuming different conformational states and that the catalytic triad may only be necessary as short-lived intermediate during catalysis.     

A physical characterization of sulfane sulfurtransferase

The bacterial enzyme sulfane sulfurtransferase has been studied using spectroscopic techniques. The enzyme was characterized in terms of its near-UV absorption spectrum, molar ellipticity, intrinsic fluorescence spectra and the effects of general and ionic quenching reagents upon its fluorescence. Fluorescence model studies are consistent with sulfane sulfurtransferase having only a single tryptophan residue, which accounts for its low UV absorption coefficient and suggested that this residue is at least partially exposed to solvent. Second derivative absorption spectroscopy studies revealed that most of the bacterial enzyme's tyrosine residues are exposed to solvent. Unlike the better known sulfurtransferase, bovine liver rhodanese, sulfane sulfurtransferase does not undergo a detectable increase in quantum yield when shifting from the sulfur-containing covalent enzyme intermediate to the free enzyme form (which lacks sulfur) during catalysis. CD studies suggest that sulfane sulfurtransferase has a significantly higher proportion of alpha-helix than rhodanese. The renaturation of sulfane sulfurtransferase denatured in 6 M guanidine was shown to be rapid and complete provided that the enzyme had not been oxidized while in the denatured state. Sulfane sulfurtransferase, like rhodanese, catalyzes the transfer of sulfur from thiosulfate to cyanide via a persulfide intermediate, and displays remarkably similar kinetics in this process (Aird, B.A., Heinrikson, R.L. and Westley, J. (1987) J. Biol. Chem 262, 17327-17335). In light of this, the results of the structural studies with sulfane sulfurtransferase are compared and contrasted to data from similar experiments with rhodanese in hopes that they would provide insight about which phenomena observed with rhodanese are intrinsic to the process of transferring sulfur atoms.     

Evidence for tryptophan in proximity to histidine and cysteine as essential to the active site of an alkaline protease

The presence, microenvironment, and proximity of an essential Trp with the essential His and Cys residues in the active site of an alkaline protease have been demonstrated for the first time using chemical modification, chemo-affinity labeling, and fluorescence spectroscopy. Kinetic analysis of the N-bromosuccinimide- (NBS) or p-hydroxymercuribenzoate- (PHMB) modified enzyme from Conidiobolus sp. revealed that a single Trp and Cys are essential for activity in addition to the Asp, His, and Ser residues of the catalytic triad. Full protection by casein against inactivation of the enzyme by NBS and quenching of Trp fluorescence upon binding of the enzyme with NBS, substrate (sAAPF-pNA), or inhibitor (SSI) confirmed participation of the Trp residue at the substrate/inhibitor binding site of the alkaline protease. Comparison of the K(sv) values for the charged quenchers CsCI (1.66) and KI (7.0) suggested that the overall Trp microenvironment in the protease is electropositive. The proximity of Trp with His was demonstrated by the sigmoidal shape of the pH-dependent fluorometric titration curve with a pK(F) of 6.1. The vicinity of Trp with Cys was indicated by resonance energy transfer between the intrinsic fluorophore (Trp) and 5-iodoacetamide-fluorescein labeled Cys (extrinsic fluorophore). Our results on the proximity of Trp with essential His and Cys thus confirm the presence of Trp in the active site of the alkaline protease.     

The use of intrinsic protein fluorescence to quantitate enzyme-bound persulfide and to measure equilibria between intermediates in rhodanese catalysis

The intrinsic fluorescence of the enzyme rhodanese is quenched by as much as 30% when sulfur is transferred to the free enzyme form, E, giving the sulfur-substituted enzyme, ES. This fluorescence change (lambda ex = 295 nm and lambda em = 335 nm) has been used to quantitate the E and ES forms which are isolatable, obligatory intermediates in rhodanese catalysis. Fluorescence titration was performed using cyanide to irreversibly remove sulfur from ES. The results show a stoichiometry corresponding to 1 bound sulfur/molecule of the ES form of rhodanese (Mr = 33,000). The fluorescence changes were used to measure the concentrations of E and ES when these were in reversible equilibria induced by interactions with the substrates S2O3(2-) and SO3(2-). These results were compared with an equilibrium constant derived from published kinetic studies for the reaction (formula; see text) The very close agreement between the physical and kinetic methods indicate that there are no significant concentrations of intermediates other than E and ES. Overall, the results are compatible with the formation of a persulfide intermediate in rhodanese catalysis and are consistent with conclusions from x-ray crystallography and absorption spectroscopy. In addition, these procedures offer a facile method to measure equilibria between catalytic intermediates in the rhodanese reaction using functionally relevant concentrations.     

Tryptophan mediated photoreduction of disulfide bond causes unusual fluorescence behaviour of Fusarium solani pisi cutinase.

The fluorescence signal of the single tryptophan residue (Trp69) of Fusarium solani pisi cutinase is highly quenched. However, prolonged irradiation of the enzyme in the tryptophan absorption band causes an increase of the tryptophan fluorescence quantum yield by an order of magnitude. By using a combination of NMR spectroscopy and chemical detection of free thiol groups with a sulfhydryl reagent we could unambiguously show that the unusual fluorescence behaviour of Trp69 in cutinase is caused by the breaking of the disulfide bond between Cys31 and Cys109 upon irradiation, while the amide-aromatic hydrogen bond between Ala32 and Trp69 remains intact. This is the first example of tryptophan mediated photoreduction of a disulfide bond in proteins.