- Split View
-
Views
-
Cite
Cite
Keita Iyori, Junzo Hisatsune, Tetsuji Kawakami, Sanae Shibata, Nobuo Murayama, Kaori Ide, Masahiko Nagata, Tsuneo Fukata, Toshiroh Iwasaki, Kenshiro Oshima, Masahira Hattori, Motoyuki Sugai, Koji Nishifuji, Identification of a novel Staphylococcus pseudintermedius exfoliative toxin gene and its prevalence in isolates from canines with pyoderma and healthy dogs, FEMS Microbiology Letters, Volume 312, Issue 2, November 2010, Pages 169–175, https://doi.org/10.1111/j.1574-6968.2010.02113.x
- Share Icon Share
Abstract
Staphylococcal exfoliative toxins are involved in some cutaneous infections in mammals by targeting desmoglein 1 (Dsg1), a desmosomal cell–cell adhesion molecule. Recently, an exfoliative toxin gene (exi) was identified in Staphylococcus pseudintermedius isolated from canine pyoderma. The aim of this study was to identify novel exfoliative toxin genes in S. pseudintermedius. Here, we describe a novel orf in the genome of S. pseudintermedius isolated from canine impetigo, whose deduced amino acid sequence was homologous to that of the SHETB exfoliative toxin from Staphylococcus hyicus (70.4%). The ORF recombinant protein caused skin exfoliation and abolished cell surface staining of Dsg1 in canine skin. Moreover, the ORF protein degraded the recombinant extracellular domains of canine Dsg1, but not Dsg3, in vitro. PCR analysis revealed that the orf was present in 23.2% (23/99) of S. pseudintermedius isolates from dogs with superficial pyoderma exhibiting various clinical phenotypes, while the occurrence in S. pseudintermedius isolates from healthy dogs was 6.1% (3/49). In summary, this newly found orf in S. pseudintermedius encodes a novel exfoliative toxin, which targets a cell–cell adhesion molecule in canine epidermis and might be involved in a broad spectrum of canine pyoderma.
Introduction
Exfoliative toxins produced by staphylococci are reported to be involved in some cutaneous infections in mammals. In humans with bullous impetigo and staphylococcal scaled-skin syndrome, virulent strains of Staphylococcus aureus produce two serotypes of exfoliative toxins (ETs), ETA and ETB, which cause skin exfoliation when injected into neonatal mice (Kondo et al., 1975; Nishifuji et al., 2008). The deduced amino acid sequences and crystallographic analyses of ETA and ETB revealed that they have the structure of glutamate-specific serine proteases containing a catalytic triad, and a putative active site comprising serine, histidine and aspartic acid (Dancer et al., 1990; Vath et al., 1997, 1999; Papageorgiou et al., 2000). Moreover, ETD, another serotype of ET produced by S. aureus, has been isolated primarily from humans with deep pyoderma (Yamaguchi et al., 2002; Yamasaki et al., 2006). Recent studies revealed that these three isoforms of ET are serine proteases, whose catalytic serine selectively hydrolyzes a single peptide bond in the extracellular domains of human and mouse desmoglein 1 (Dsg1), a desmosomal cell–cell adhesion molecule (Amagai et al., 2000, 2002; Hanakawa et al., 2002; Yamaguchi et al., 2002).
In Staphylococcus hyicus, five isoforms of ETs (SHETB, ExhA, ExhB, ExhC and ExhD) have been reported, whose amino acid sequences are homologous to those of S. aureus ETs (Sato et al., 1999; Ahrens & Andresen, 2004). Among the five isoforms of S. hyicus ETs, the four Exh isoforms were primarily isolated from S. hyicus causing exudative epidermitis in pigs (Futagawa-Saito et al., 2007). These Exh isoforms were also reported to induce skin exfoliation in piglets, selectively digesting the extracellular domains of swine Dsg1 (Fudaba et al., 2005; Nishifuji et al., 2005). Digestion of Dsg1 by ETs causes the disruption of cell–cell adhesion of keratinocytes predominantly in the upper spinous and granular layers, wherein loss of the adhesive function of Dsg1 is not compensated for by other Dsg isoforms (Amagai, 1999; Stanley & Amagai, 2006; Nishifuji et al., 2008).
Recently, an exfoliative toxin gene (exi) was identified in the chromosomal DNA of a strain of Staphylococcus pseudintermedius isolated from canine pyoderma. The deduced amino acid sequence of the exi gene product (EXI) was homologous to known ETs and caused intraepidermal splitting in mouse skin (Futagawa-Saito et al., 2009). In addition, our laboratory found recently that EXI selectively digests the extracellular domain of canine Dsg1 and causes intraepidermal splitting in canine skin (K. Iyori & K. Nishifuji, manuscript in preparation). However, it has not yet been revealed whether S. pseudintermedius strains harbor other ET encoding genes. During genome sequence analysis of a clinically isolated EXI-negative S. pseudintermedius strain, an orf showing significant similarity to ET genes was identified. In the current study, the exfoliative toxicity and substrate specificity of this orf product were characterized and the prevalence of this orf was investigated among clinical isolates from cases of canine superficial pyoderma exhibiting various clinical phenotypes, and healthy carriers.
Materials and methods
Bacterial strains, cloning and sequence analysis
Staphylococcus pseudintermedius MS5134 isolated from a pustule of a dog with impetigo was used for genome sequencing and cloning of an orf encoding a putative ET gene. To investigate the occurrence of this orf, 148 S. pseudintermedius isolates (99 from dogs diagnosed with superficial pyoderma and 49 from healthy dogs) collected from three veterinary institutes in Japan (Tokyo University of Agriculture and Technology, ASC Dermatology Service and Gifu University) were analyzed in this study. The S. pseudintermedius isolates were collected either from the skin lesions of dogs with superficial pyoderma or from the nasal cavity of healthy dogs. Identification of S. pseudintermedius was carried out on the basis of Gram-staining results, catalase, coagulase and β-galactosidase activities, and multiplex PCR to distinguish the thermonuclease (nuc) gene of S. pseudintermedius from that of other staphylococcal species (Sasaki et al., 2010). Chromosomal DNA was isolated from S. pseudintermedius strains using the UltraClean Microbial DNA Isolation Kit (MO BIO, Carlsbad, CA). The genome of strain MS5134 was sequenced using the whole-genome random-sequencing method as described previously at the Graduate School of Frontier Sciences, The University of Tokyo, Japan (Toh et al., 2010). From these DNA sequences, a putative ET gene homolog was identified based on sequence homology to ETD of S. aureus. The sequence data were analyzed using the blastp program (http://blast.ncbi.nlm.nih.gov/Blast.cgi) for amino acid homology searches, the clustalx program (http://www.clustal.org) for multisequence alignments and phylogenic analysis and the signalp 3.0 program (http://www.cbs.dtu.dk/services/SignalP) for signal peptide prediction. The theoretical pI and molecular weight were calculated using the expasy proteomics server (http://us.expasy.org/tools/pi_tool.html).
Production and purification of the recombinant ORF protein
The primers 5′-gggcatgcacatatgatgaagcc-3′ (forward primer) and 5′-ccagatctatcttctgattcagc-3′ (reverse primer) were used to amplify the novel orf by PCR. The PCR product was digested with SphI and BglII, subcloned into SphI–BglII-digested pQE70 vector (Qiagen, Valencia, CA) and transfected into Escherichia coli M15 cells (Qiagen). The final construct was confirmed by sequence analysis and was designated pQE-neworf-His. The recombinant protein encoded by the novel orf was harvested from the cytoplasmic soluble fraction of the transfected M15 cells using lysis buffer (BugBuster Protein Extraction Reagent; Novagen, Madison, WI), purified with TALON metal affinity resin (BD Biosciences Clontech, Mountain View, CA) and dialyzed against phosphate-buffered saline (PBS). The presence of the new ORF protein in the purified fraction was confirmed by immunoblotting with anti-His tag antisera (MBL, Nagoya, Japan) (data not shown).
In vivo exfoliative toxicity assay
To evaluate the exfoliative toxicity of the novel ORF protein in canine skin, 250 μg of the protein dissolved in 50 μL of PBS, or PBS alone as a negative control, were injected intradermally into inguinal areas of three healthy Beagles. Skin samples of the injected areas were biopsied under local anesthesia 12 and 24 h after injection. All of the animal experiments were approved by the animal research committee at Tokyo University of Agriculture and Technology.
Formalin-fixed, paraffin-embedded skin samples were subjected to histopathological analysis. Frozen skin samples were subjected to immunofluorescence analysis for Dsg1 and Dsg3 using a human pemphigus foliaceus serum containing anti-Dsg1 IgG (Amagai et al., 1994, 1999; Ishii et al., 1997) and an AK15 mouse monoclonal antibody against Dsg3 (Tsunoda et al., 2003) (kind gifts from Dr Masayuki Amagai, Keio University School of Medicine, Tokyo, Japan), respectively. The anti-Dsg1 IgG serum and the AK15 monoclonal antibody were detected with fluorescein isothiocyanate-conjugated goat anti-human IgG (MP Biomedicals, Solon, OH) and Alexa Fluor 546 goat anti-mouse IgG (Invitrogen Corp., Carlsbad, CA), respectively.
In vitro enzymatic assay to digest canine Dsg proteins
Secreted forms of the recombinant proteins representing the entire extracellular domain of canine Dsg1 (cDsg1) and Dsg3 (cDsg3), fused with the hinge region of human IgG1, an E tag and a His tag at their carboxyl termini, were produced using the baculovirus-expression system as described previously (Nishifuji et al., 2003a, b). Five insect cell supernatants containing recombinant cDsg1 or cDsg3 were mixed with 10 μg of purified new ORF protein or PBS alone, and incubated at 37 °C for 12 h. Immunoblotting with rabbit anti-E-tag polyclonal antibody (Bethyl Laboratories, Montgomery, TX) was performed to detect intact and/or degraded cDsg proteins.
Detection of the novel orf in various S. pseudintermedius strains
PCR was performed with the primers 5′-gcggcatgcctaaaacatatgatgaagccgaa-3′ (forward primer) and 5′-tctctatttacattcagagag-3′ or 5′-tctggatccatcttctgattcagctctttttttcaaa-3′ (reverse primers) to amplify two partial regions of the orf gene. The PCR products were resolved by electrophoresis through a 1.2% (w/v) agarose gel, and visualized by the application of the SYBR safe DNA gel stain (Invitrogen Corp.).
Nucleotide sequence accession numbers
Nucleotide sequence data obtained in this study are available in the DDBJ, EMBL and GenBank nucleotide sequence databases under accession number AB569087.
Results and discussion
Identification of an orf encoding a novel ET gene in S. pseudintermedius
During genome sequencing analysis of S. pseudintermedius strain MS5134, an orf with significant homology to a previously reported ET gene was identified. This orf consisted of 843 bp and was predicted to encode a protein of 279 amino acid residues, including a putative signal peptide in the first 32 amino acids (Fig. 1). The mature protein derived from an orf consisting of 247 amino acid residues with an N-terminal sequence beginning with KTYDEAEIIKK, and a predicted molecular weight and pI of 26.9 kDa and 5.86, respectively.
The deduced amino acid sequence of the orf was compared with previously isolated ETs including S. aureus ETs (ETA, ETB and ETD), S. hyicus ETs (SHETB, ExhA, ExhB, ExhC and ExhD) and S. pseudintermedius EXI. Significant homology was detected with these known ETs (38.4–70.4% identity), particularly with SHETB (70.4%), ETD (66.1%) and EXI (56.9%). In addition, the predicted amino acid sequence of the orf possessed the conserved catalytic triad, His-99 (H), Asp-147 (D) and Ser-221 (S), which is known to comprise the active site of S. aureus ETA, ETB and ETD needed to digest Dsg1 (Fig. 1) (Hanakawa et al., 2004). Phylogenic analysis of the ETs revealed that the orf was most similar to SHETB in its primary structure (Fig. 2).
Recombinant protein of the novel ORF targeted Dsg1 and caused superficial epidermal splitting in canine skin
To investigate whether the novel orf gene product conferred exfoliative toxicity in canine skin, purified recombinant protein of the orf product (new ORF) or PBS was injected into the skin of three healthy Beagles. Macroscopically, the novel ORF protein induced skin exfoliation at 24 h postinjection, whereas no apparent changes were observed with PBS alone (Fig. 3a and b). The injection site was evaluated histopathologically 12 h after injection. Intraepidermal splitting at the level of the granular layer was observed at the site of injection of the new ORF protein, while no changes were observed at the PBS injection site (Fig. 3c and d). Splitting was also observed in the granular layer of the follicular infundibulum (data not shown).
To determine the effect of the new ORF protein on Dsg1 in canine skin, immunofluorescence analysis of Dsg1 and Dsg3 was performed using cryosections of the canine skin described above. In normal canine skin, Dsg1 is reportedly expressed throughout the entire epidermal layer, while Dsg3 is only expressed in the lower epidermis (Nishifuji et al., 2007). We found that cell surface staining for Dsg1 was abolished in canine skin injected with the new ORF protein, whereas staining was retained in the skin injected with PBS (Fig. 3e and f). In the same area, the cell surface staining for Dsg3 was not altered by the presence or absence of the recombinant toxins (Fig. 3g and i). To further investigate the direct degradation of the extracellular domains of canine Dsg1 by the novel ORF protein, baculovirus cDsg1 and cDsg3 proteins were incubated with the purified ORF protein or PBS alone in vitro. Immunoblot analysis showed that cDsg1, but not cDsg3, was degraded into smaller peptides by the novel ORF protein (Fig. 4). The exfoliative toxicity of the new ORF protein demonstrated in this study, namely the selective digestion of Dsg1, was similar to that seen with previously isolated ETs (Amagai et al., 2000, 2002; Yamaguchi et al., 2002; Fudaba et al., 2005; Nishifuji et al., 2005), including S. pseudintermedius EXI (K. Iyori & K. Nishifuji, manuscript in preparation).
The new orf gene was detectable in S. pseudintermedius isolates from dogs with superficial pyoderma exhibiting various clinical phenotypes
The occurrence of the orf gene was determined among Japanese isolates of S. pseudintermedius from the cutaneous lesions of dogs with superficial pyoderma exhibiting various clinical phenotypes and from the nasal cavities of healthy dogs without any skin lesions. PCR analysis revealed that the novel orf was present in 23.2% (23/99) of S. pseudintermedius isolates from dogs with superficial pyoderma (Table 1). This rate of occurrence was similar to that of the S. pseudintermedius exi gene (23.3%) reported previously (Futagawa-Saito et al., 2009). The 23 clinical isolates positive for the orf were collected from dogs exhibiting pustules (15), erythema (5), scales/epidermal collarettes (1) and crusts (2) (Table 1). In contrast, the rate of occurrence of the orf gene in S. pseudintermedius isolates from healthy dogs was 6.1% (3/49) (Table 1).
Clinical phenotype (number of isolates tested) | Number (%) of positive isolates |
Superficial pyoderma (99) | 23 (23.2) |
Pustule (62) | 15 (24.2) |
Erythema (7) | 5 (71.4) |
Scales/epidermal collarette (13) | 1 (7.7) |
Crusts (17) | 2 (11.8) |
No skin lesions (49) | 3 (6.1) |
Clinical phenotype (number of isolates tested) | Number (%) of positive isolates |
Superficial pyoderma (99) | 23 (23.2) |
Pustule (62) | 15 (24.2) |
Erythema (7) | 5 (71.4) |
Scales/epidermal collarette (13) | 1 (7.7) |
Crusts (17) | 2 (11.8) |
No skin lesions (49) | 3 (6.1) |
Clinical phenotype (number of isolates tested) | Number (%) of positive isolates |
Superficial pyoderma (99) | 23 (23.2) |
Pustule (62) | 15 (24.2) |
Erythema (7) | 5 (71.4) |
Scales/epidermal collarette (13) | 1 (7.7) |
Crusts (17) | 2 (11.8) |
No skin lesions (49) | 3 (6.1) |
Clinical phenotype (number of isolates tested) | Number (%) of positive isolates |
Superficial pyoderma (99) | 23 (23.2) |
Pustule (62) | 15 (24.2) |
Erythema (7) | 5 (71.4) |
Scales/epidermal collarette (13) | 1 (7.7) |
Crusts (17) | 2 (11.8) |
No skin lesions (49) | 3 (6.1) |
It has been reported previously that the S. aureus etd gene could be isolated from various cutaneous infections in humans, including cutaneous abscesses, furuncles and finger pulp infections. Conversely, the isolation rate of the etd gene was much lower than that of the eta and etb genes in humans with bullous impetigo, a dermatological disorder that exhibits intraepidermal blisters due to the disruption of cell–cell adhesion of epidermal keratinocytes (Kanzaki et al., 1997; Yamaguchi et al., 2002; Yamasaki et al., 2006). Similar to these findings in humans, the novel orf gene from S. pseudintermedius was detectable in dogs with superficial pyoderma exhibiting various clinical phenotypes. Because the orf product targets a cell–cell adhesion molecule of keratinocytes in superficial epidermis and follicular infundibulum, there is an intriguing possibility that this effect may be facilitating the invasion and spread of staphylococci into the epidermis and hair follicles of dogs, resulting in a broad spectrum of canine pyoderma.
In summary, a novel orf encoding a second ET was identified in S. pseudintermedius, and its product disrupted a single cell–cell adhesion molecule in canine epidermis. With respect to the nomenclature of Exhs, we propose that S. pseudintermedius EXI be renamed ExpA and the novel ORF protein reported here be named ExpB (T. Olivry, pers. commun.). Further epidemiological analysis of ExpA- and ExpB-positive S. pseudintermedius strains and a comparative genomic analysis will help to identify the pathogenic involvement of these Exp proteins in cutaneous infections in mammals. It will also be interesting to raise antibodies against Exp proteins for the detection of Exps at the protein level in cutaneous lesions, and to compare the histopathological patterns of Exp-positive and -negative skin lesions of pyoderma in future studies.
Authors' contribution
K.I. and J.H. contributed equally to this study.
Acknowledgements
We are grateful to Dr Masayuki Amagai, Department of Dermatology, Keio University School of Medicine, Tokyo, Japan, for kindly providing human pemphigus foliaceus serum and AK15 mouse anti-Dsg3 monoclonal antibody, and to Keiko Furuya, Chie Shindo, Hiromi Inaba, Erika Iioka, Kanako Motomura and Yasue Hattori (The University of Tokyo, Japan) for technical assistance. We also thank Dr Thierry Olivry, Center for Comparative Medicine and Translational Research, College of Veterinary Medicine (North Carolina State University, Raleigh, NC), for his insightful suggestions regarding the nomenclature of S. pseudintermedius exfoliative toxins. This work was supported by Grants-in-Aid for Scientific Research (to K.N. and J.H.) and by Grants-in-Aid for Scientific Research on Priority Areas ‘Applied Genomics’ (to M.S.) and ‘Comprehensive Genomics’ (to M.H.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
References
Author notes
Editor: Jan-Ingmar Flock