Staphylococcal exfoliative toxins: “Molecular scissors” of bacteria that attack the cutaneous defense barrier in mammals

Bullous impetigo and its generalized form, staphylococcal scalded-skin syndrome (SSSS), are highly contagious, blistering skin diseases caused by Staphylococcus aureus infection. Virulent strains of the bacteria produce exfoliative toxins (ETs) that cause the loss of keratinocyte cell-cell adhesion in the superficial epidermis. Recent studies have indicated that the three isoforms of ETs, i.e., ETA, ETB, and ETD, are glutamate-specific serine proteases that specifically and efficiently cleave a single peptide bond in the extracellular region of human and mouse desmoglein 1 (Dsg1), a desmosomal intercellular adhesion molecule. In addition, four isoforms of S. hyicus exfoliative toxin, ExhA, ExhB, ExhC, and ExhD, cleave swine Dsg1, resulting in skin exfoliation similar to that observed in pigs with exudative epidermitis. In this review, we describe recent advances in our knowledge of the mechanisms of action of staphylococcal exfoliative toxins, which act as "molecular scissors" to facilitate percutaneous bacterial invasion of mammalian skin by cleavage of keratinocyte cell-cell adhesion molecules. The species-specificity of staphylococcal exfoliative toxins to cleave Dsg1 in certain mammalian species is discussed.

Staphylococcal scalded skin syndrome I. Purification of exfoliatin and maternal transmission of neutralizing ability against exfoliatin

Exfoliatin, an exotoxin produced by phage group 2 staphylococci, was separated and was analyzed. The toxin was separated in a single peak by Sephadex G-100 gel filtration and subsequent DEAE Sephadex chromatography. By acrylamide gel disc electrophoresis this single peak was found to contain 2 components both of which posessed exfoliative activity and the same antigenic properties. Their molecular weight, however, differed; the major component had a molecular weight of 30,000 and the minor component had 27,000. The toxin was stable when heated at 60 degrees C for 40 minutes, and labile at 100 degrees C for 20 minutes. The toxin was immunogenic to mice and produced antibody which could be transmitted to neonatal mice to neutralize exfoliative activity of the toxin. Innoculation of the bacteria caused the production of antibodies against both exfoliatin and alpha-hemolysin in mice. The antibodies were also transmitted to neonatal mice to neutralize the exfoliative activity of exfoliatin.

Differential structural properties and expression patterns suggest functional significance for multiple mouse desmoglein 1 isoforms

The four isoforms of desmosomal cadherin desmogleins (Dsg1-4) are expressed in epithelial tissues in a differentiation-specific manner. Extensive sequencing of the human genome has revealed only one copy of the Dsg1 gene. However, we recently cloned two novel additional mouse Dsg1 genes, Dsg1-beta and -gamma, which flank the original Dsg1-alpha on chromosome 18. Sequence conservation between the Dsg1 isoforms diverged significantly at exon 11, particularly in the region that encodes for the extracellular anchoring (EA) domains. Computational analysis revealed very low hydrophilic potential of the Dsg1-gamma EA compared with the corresponding sequences of Dsg1-alpha and -beta, suggesting that the Dsg1-gamma EA domain may have a stronger affinity to the cell membrane. We generated antibodies using synthetic peptides or recombinant proteins localized within the EA domains. These antibodies were tested for their specificity and were then used to demonstrate expression of Dsg1 isoforms in various tissues. In the epidermis, all Dsg1 isoforms were differentially expressed in the differentiating cell layers. In the hair follicle, all Dsg1 isoforms were present throughout the entire process of its development and cycling but the expression of Dsg1 isoforms is subject to significant hair cycle-dependent changes. Dsg1-beta and -gamma, but not Dsg1-alpha, were detected in the sebaceous gland epithelium and the stratified epithelium of the stomach. Finally, Dsg1-alpha and Dsg1-beta, but not Dsg1-gamma, are proteolytically cleaved by exfoliative toxin A. These results suggest that the developmental complexity of mouse tissues, including skin and hair, may play a significant role in the evolutionary driving force to maintain multiple Dsg1 genes in mouse.

Exudative epidermitis in pigs caused by toxigenic Staphylococcus chromogenes

Staphylococcus chromogenes is closely related to Staphylococcus hyicus, which is recognised as the causative agent of exudative epidermitis (EE) in pigs. S. chromogenes is part of the normal skin flora of pigs, cattle and poultry and has so far been considered non-pathogenic to pigs. A strain of S. chromogenes producing exfoliative toxin type B, ExhB, was identified by the use of a multiplex PCR specific for the exfoliative toxins from S. hyicus. The exfoliative toxin from S. chromogenes reacted in immunoblot analysis with polyclonal and monoclonal antibodies specific to ExhB from S. hyicus and had an apparent molecular weight of 30 kDa. Sequencing the gene encoding the exfoliative toxin from S. chromogenes revealed that the molecular weight of the toxin with the signal peptide and the mature toxin was 30,553 and 26,694 Da, respectively. Comparison of the exhB genes from S. chromogenes strain VA654 and S. hyicus strain 1289D-88 showed differences in seven base pairs of the DNA sequences and in two amino acid residues in the deduced amino acid sequences. Pigs were experimentally inoculated with S. chromogenes strain VA654. By clinical observations and histopathological evaluation of the skin alterations, all pigs revealed development of generalized exudative epidermitis. No toxin producing S. hyicus was isolated from the pigs and all ExhB-positive bacterial isolates were identified as S. chromogenes. This confirmed that the disease-causing agent was the inoculated S. chromogenes strain VA654. The results of this study show that S. chromogenes may cause exudative epidermitis in pigs.

Cloning and sequence analysis of genes encoding Staphylococcus hyicus exfoliative toxin types A, B, C, and D

Exfoliative toxins produced by certain strains of Staphylococcus hyicus mediate exudative epidermitis in pigs. In this study the genes coding for four different exfoliative toxin from S. hyicus (ExhA, ExhB, ExhC, and ExhD) were cloned and sequenced. The coding sequence of the four toxin genes ranged from 816 to 834 bp. The amino acid sequences of these four toxins were homologous to the earlier described exfoliative toxins SHETB from S. hyicus and ETA, ETB, and ETD from Staphylococcus aureus. The homology between the S. hyicus toxins was at the same level as the homology to the exfoliative toxins from S. aureus. The toxins showed similarity to serine proteases, including preservation of the catalytic tract in ExhA, ExhB, and ExhC. However, in ExhD, Asp in the putative catalytic tract was replaced with Glu. The recombinant toxins could be expressed in Escherichia coli, and three of the four toxins were recognized by monoclonal antibodies raised against native exfoliative toxins.

Understanding the mechanism of action of the exfoliative toxins of Staphylococcus aureus

The exfoliative toxins of Staphylococcus aureus are responsible for the staphylococcal scalded skin syndrome, a blistering skin disorder that particularly affects infants and young children, as well as adults with underlying disease. Their three-dimensional structure is similar to other glutamate-specific trypsin-like serine proteases with two substrate-binding domains and a serine-histidine-aspartate catalytic triad that forms the active site. However, unlike other serine proteases, the exfoliative toxins possess a highly charged N-terminal alpha-helix and a unique orientation of a critical peptide bond, which blocks the active site of the toxins so that, in their native state, they do not possess any significant enzymatic activity. The target for the toxins has recently been identified as desmoglein-1, a desmosomal glycoprotein which plays an important role in maintaining cell-to-cell adhesion in the superficial epidermis. It is speculated that binding of the N-terminal alpha-helix to desmoglein-1 results in a conformation change that opens the active site of the toxin to cleave the extracellular domain of desmoglein-1 between the third and fourth domains, resulting in disruption of intercellular adhesion and formation of superficial blisters. Elucidating the mechanism of action of the toxins and identifying desmoglein-1 as their specific epidermal substrate has not only given us an insight into the pathogenesis of the staphylococcal scalded skin syndrome, but also provided us with useful information on normal skin physiology and the pathogenesis of other toxin-mediated diseases. It is hoped that this knowledge will lead to development of rapid screening and diagnostic tests, and new antitoxin strategies for the treatment and prevention of the staphylococcal scalded skin syndrome in the near future.

Structural similarities and differences in Staphylococcus aureus exfoliative toxins A and B as revealed by their crystal structures

Staphylococcal aureus epidermolytic toxins (ETs) A and B are responsible for the induction of staphylococcal scalded skin syndrome, a disease of neonates and young children. The clinical features of this syndrome vary from localized blisters to severe exfoliation affecting most of the body surface. Comparison of the crystal structures of two subtypes of ETs-rETA (at 2.0 A resolution), rETB (at 2.8 A resolution), and an active site variant of rETA, Ser195Ala at 2.0 A resolution has demonstrated that their overall topology resembles that of a "trypsin-like" serine protease, but with significant differences at the N- and C-termini and loop regions. The details of the catalytic site in both ET structures are very similar to those in glutamate-specific serine proteases, suggesting a common catalytic mechanism. However, the "oxyanion hole," which is part of the catalytic sites of glutamate specific serine proteases, is in the closed or inactive conformation for rETA, yet in the open or active conformation for rETB. The ETs contain a unique amphipathic helix at the N-terminus, and it appears to be involved in optimizing the conformation of the catalytic site residues. Determination of the structure of the rETA catalytic site variant, Ser195Ala, showed no significant perturbation at the active site, establishing that the loss of biological and esterolytic activity can be attributed solely to disruption of the catalytic serine residue. Finally, the crystal structure of ETs, together with biochemical data and mutagenesis studies, strongly confirms the classification of these molecules as "serine proteases" rather than "superantigens."

Mutational analysis of the superantigen staphylococcal exfoliative toxin A (ETA)

Exfoliative toxin A (ETA) is known to be a causative agent of staphylococcal scalded skin syndrome (SSSS). Although relatively little is known about exactly how the exfoliative toxins (ETs) cause SSSS, much has been discovered recently that may help elucidate the mechanism(s) by which ETA exhibits activities such as lymphocyte mitogenicity and epidermolytic activity. Here, we have shown that highly purified ETA does have T lymphocyte mitogenic activity in that wild-type ETA induced T cell proliferation whereas several single amino acid mutants lacked significant activity. Neither wild-type ETA nor any single amino acid mutants were proteolytic for a casein substrate, yet esterase activity was detected in wild-type ETA and several mutants, but eliminated in other mutants. A mutation in aa 164 (Asp to Ala) showed a 9-fold increase in esterase activity as well. Finally, we correlated esterase activity with epidermolytic activity. All mutants that lost esterase activity also lost epidermolytic activity. Conversely, mutants that retained esterase activity also retained exfoliative activity, implicating serine protease or serine protease-like activity in the causation of SSSS. Moreover, the mutants that displayed markedly reduced T cell superantigenic activity retained their epidermolytic activity (although some of these mutants required higher doses of toxin to cause disease), which suggests an ancillary role for this activity in SSSS causation.

The crystal structure of exfoliative toxin B: a superantigen with enzymatic activity

The exfoliative toxins (ETs) cause staphylococcal scalded skin syndrome, a disease characterized by specific separation of layers of the skin. Evidence suggests that the toxins act as serine proteases, though the specific substrate and mode of action are not known for certain. The crystal structure of exfoliative toxin A (ETA) was reported earlier and shown to be similar to that of the chymotrypsin-like serine proteases. Here, we report the 2.4 A resolution crystal structure of the other exfoliative toxin, ETB, which is 40% identical to ETA. The overall structures of ETA and ETB are similar including the positions of key residues within the active site. The structure of ETB supports the previous findings that the ETs are serine proteases that cleave substrates after glutamic acid residues. In this study we also discuss a number of structural differences including a large 14 residue loop insertion which may be a key feature involved in the differing biological properties of the ETs, particularly the pyrogenic and lethal activities of ETB not shared by ETA.

Unique superantigen activity of staphylococcal exfoliative toxins

Certain strains of Staphylococcus aureus express one or both of two related, but immunologically distinct, exfoliative toxins (ETA and ETB). These toxins induce the symptoms associated with staphylococcal scalded skin syndrome. Both ETs have been shown to stimulate T cell proliferation. Recently, it was reported that ETA is a superantigen that stimulates T cells bearing human Vbeta2 or several murine Vbetas. However, other investigators have proposed that the superantigenicity reported for ETA resulted from contaminants in commercial preparations. This present study addresses those conflicting reports by assessing the biological and immunologic activities of highly purified rETs. ETA and ETB required APCs to induce selective polyclonal expansion of several human Vbetas (huVbetas), although, neither toxin expanded huVbeta2. ETB induced expansion of murine T cells bearing Vbetas 7 and 8, those that have the highest homology to the huVbetas expanded by ETA and ETB. Although flow cytometry of ETB-stimulated T cells matched PCR results, stimulation by ETA reduced percentages of T cells positive for several huVbetas that had been shown to have increased levels of mRNA transcripts. ETA and ETB induced contrasting reactions in vivo. In rabbits, ETB was moderately pyrogenic and enhanced susceptibility to lethal shock, while ETA lacked both activities. Predictions based on comparisons with other superantigens suggest molecular regions potentially involved in receptor binding in the ETA crystal structure and a modeled ETB three-dimensional structure. These results show that ETs are superantigens with unique properties that could account for the discrepancies reported.

Differentiation and distribution of three types of exfoliative toxin produced by Staphylococcus hyicus from pigs with exudative epidermitis

Exfoliative toxins of approximately 30 kDa produced by Staphylococcus hyicus strains NCTC 10350, 1289D-88 and 842A-88 were purified and specific polyclonal antisera were raised against each of the toxins. It was shown by immunoblot analysis and ELISA that three exfoliative toxins from S. hyicus were antigenically distinct. The three toxins were designated ExhA, ExhB and ExhC. From 60 diseased pigs, each representing an outbreak of exudative epidermitis, a total of 584 isolates of S. hyicus were phage typed and tested for production of exfoliative toxin. ExhA-, ExhB- and ExhC-producing S. hyicus isolates were found in 12 (20%), 20 (33%) and 11 (18%), respectively, of the 60 pig herds investigated. Production of the different types of exfoliative toxin was predominantly associated with certain phage groups. However, toxin production was found in all of the six phage groups defined by the phage typing system. Some changes in the distribution of isolates between phage groups were observed when the results of this study were compared to previous investigations. In this study two new antigenically distinct exfoliative toxins were isolated and tools for in vitro detection of toxin producing S. hyicus isolates and for further studies on the exfoliative toxins from S. hyicus have been provided.

Distant homology recognition using structural classification of proteins

Protein structure prediction is arguably the biggest unsolved problem of structural biology. The notion of the number of naturally occurring different protein folds being limited allows partial solution of this problem by the use of fold recognition methods, which "thread" the sequence in question through a library of known protein folds. The fold recognition methods were thought to be superior to the distant homology recognition methods when there is no significant sequence similarity to known structures. We show here that the Structural Classification of Proteins (SCOP) database, organizing all known protein folds according their structural and evolutionary relationships, can be effectively used to enhance the sensitivity of the distant homology recognition methods to rival the "threading" methods. In the CASP2 experiment, our approach correctly assigned into existing SCOP superfamilies all of the six "fold recognition" targets we attempted. For each of the six targets, we correctly predicted the homologous protein with a very similar structure; often, it was the most similar structure. We correctly predicted local alignments of the sequence features that we found to be characteristic for the protein superfamily containing a given target. Our global alignments, extended manually from these local alignments, also appeared to be rather accurate.

Fold recognition using predicted secondary structure sequences and hidden Markov models of protein folds

We present an analysis of the blind predictions submitted to the fold recognition category for the second meeting on the Critical Assessment of techniques for protein Structure Prediction. Our method achieves fold recognition from predicted secondary structure sequences using hidden Markov models (HMMs) of protein folds. HMMs are trained only with experimentally derived secondary structure sequences of proteins having similar fold, therefore protein structures are described by the models at a remarkably simplified level. We submitted predictions for five target sequences, of which four were later found to be suitable for threading. Our approach correctly predicted the fold for three of them. For a fourth sequence the fold could have been correctly predicted if a better model for its structure was available. We conclude that we have additional evidence that secondary structure information represents an important factor for achieving fold recognition.

Protein folds from pair interactions: a blind test in fold recognition

We submitted nine predictions to CASP2 using our fold recognition program ProFIT. Two of these structures were still unsolved by the end of the experiment, six had a recognizable fold, and one fold was new. Four predictions of the six recognizable folds were correct. Two models were excellent in terms of alignment quality (T0031, T0004): in one the alignment was partially correct (T0014), and one fold was correctly identified (T0038). We discuss improvements of the program and analyze the prediction results.

Novel features of serine protease active sites and specificity pockets: sequence analysis and modelling studies of glutamate-specific endopeptidases and epidermolytic toxins

With a view to obtaining a better understanding of the structural determinants of P1 glutamate specificity in glutamate-specific endopeptidases (GSEs), the active sites and specificity pockets of such enzymes from Bacillus licheniformis (gse-bl), Bacillus subtilis (mpr) and Staphylococcus aureus (v8 protease) were modelled. This approach was extended to the epidermolytic toxins (ETs), responsible for the staphylococcal scalded skin syndrome. We identify a canonical structure for the S1 subsite, composed of H213 and T190, both of which we predict to interact directly with the P1 glutamate. The possible importance of R30 (for gse-bl and mpr) and of the N-terminus (for gse-bl, mpr and v8 protease) was also examined. In the case of mpr, a G193C substitution is predicted to participate in a novel disulphide bridge which stabilizes C193 in such a way as to maintain the oxyanion hole. In v8, the loss or substitution of several important structural components around D102 of the catalytic triad probably explains its reduced catalytic efficiency in comparison with other GSEs. In the case of the epidermolytic toxins K216 may be important for the previously reported phospholipase C-like activity, since the model predicts that it may stabilize the negative charge on the phosphonyl group.

The epidermolytic (exfoliative) toxins of Staphylococcus aureus

Two epidermolytic toxins, produced by different strains of Staphylococcus aureus, split human skin at a site in the upper epidermis. Clinical effects are most common in infants, but adults are susceptible. Epidermolysis may also be observed in the mouse, in vivo and in vitro, and in a few other mammals. Recent in vitro experiments have demonstrated an inhibition by chelators and point to metal-ion, possibly Ca2+, involvement. The epidermolysis effect is insensitive to a wide range of other metabolic inhibitors. The toxin amino acid sequences are similar to that of staphylococcal proteinase, and new experiments by chemical modification and site-directed mutagenesis have shown that toxicity depends on 'active serine' residues of a catalytic triad similar to that found in serine proteases. Furthermore the toxins possess esterolytic activity, also dependent on the 'active serine' sites. However, the toxins have low or undetectable activity towards a range of peptide or protein substrates. In histological and related studies, the toxins bound selectively to an intracellular skin protein, profilaggrin, but there was no evidence that the toxin can enter intact epidermal cells. Therefore, although the circumstantial evidence that the toxins act by proteolysis is convincing, a specific skin proteolytic substrate for the toxin has not been identified.

Functional evidence that the Ser-195 residue of staphylococcal exfoliative toxin A is essential for biological activity

The substitution of the serine 195 residue of staphylococcal exfoliative toxin A by a cysteine residue led to a biologically inactive protein. This result is consistent with the hypothesis that exfoliative toxin A could be a protease or a lipase. However, no protease or lipase activity was detected with the native toxin.

The role of the serine protease active site in the mode of action of epidermolytic toxin of Staphylococcus aureus

The sequences of the epidermolytic toxins and V8 serine proteinase share about 25% identity, including the catalytic triad at the proteinase active centre. Here we have altered the putative ETA active-site serine-195 to glycine by site-directed mutagenesis. No epidermolytic activity was detected when up to 100-fold greater amounts of the homogeneous mutant ETA were injected subcutaneously into neonatal mice showing that serine-195 is required for toxicogenesis.