OUP user menu

IdeE, an IgG-endopeptidase of Streptococcus equi ssp. equi

Jonas Lannergård, Bengt Guss
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00404.x 230-235 First published online: 1 September 2006


Streptococcus equi ssp. equi is the causative agent of strangles, a highly contagious and serious disease in the upper respiratory tract of horses. The present study describes the characterization of IdeE, a homolog of the secreted IgG-specific protease IdeS/Mac of Streptococcus pyogenes. The activity of IdeE is compared with the activity of IdeZ, the corresponding enzyme of the closely related S. equi ssp. zooepidemicus. A study of the proteolytic activity of recombinant IdeE and IdeZ on IgG from a selection of mammals shows that only antibodies containing the substrate site of IdeS/Mac are cleaved, indicating that the specificities of these enzymes are similar. Interestingly, IgG from horse is less effectively cleaved than IgG from e.g. dog or humans, as the dominating IgG isotype in horse sera (IgG4) lacks a distinct substrate site for IdeE/IdeZ. IgG-degradation is observed when S. equi ssp. equi is grown in the presence of horse serum, but not when grown with purified IgG. As the fraction of degraded IgG contains IgG4, the observed activity might be due to the expression of an unknown enzyme rather than IdeE. In a similar assay, no proteolysis of IgG was detected in the growth media of S. equi ssp. zooepidemicus.

  • Streptococcus equi
  • Streptococcus pyogenes
  • IdeS
  • Mac
  • immunoglobulin
  • virulence


Bacterial pathogens have evolved mechanisms to interfere with the immune response triggered by the host during infection. In order to invade and proliferate, the pathogen has to avoid recognition by immunoglobulins, key components of the adaptive immune response. IdeS, also known as Mac, is a cysteine protease secreted by the important human pathogen Streptococcus pyogenes. IdeS/Mac cleaves the hinge region of human IgG with high specificity, and thereby interferes with phagocytic killing (Leiet al,2001, 2002; von Pawel-Rammingenet al,2002; von Pawel-Rammingen & Björck, 2003). Enzymatic characterization of IdeS/Mac suggests that the high substrate specificity is most likely due to a secondary interaction with a binding site located in the CH2 or CH3 domains of IgG (Vincentset al,2004). A recent study demonstrated that antibodies from patients with S. pyogenes infections could neutralize the activity of IdeS/Mac and affect streptococcal survival in human blood (Åkessonet al,2006).

Variants of IdeS/Mac from different S. pyogenes isolates have been grouped into two allele families, complex I and II, based on sequence divergence in the middle one-third of the protein (Leiet al,2002). The complex I variants efficiently bind the lower hinge region of human IgG, leading to proteolytic cleavage of the antibody. In contrast, variants representing complex II display weak peptidase activity but instead efficiently bind FcγRII and FcγRIII, and thereby block IgG from recognition by polymorphonuclear leukocytes expressing these IgG Fc-receptors (Agniswamyet al,2004). The conclusion by Agniswamy et al. was that IdeS/Mac mediates host immune response evasion by two separate mechanisms: immunoglobulin degradation and blockage of their receptors. A later study suggests that the reported low peptidase activity of complex II might only apply to a minority of variants with double cysteines that undergo autooxidation (Åkessonet al,2006).

Streptococcus equi ssp. equi is the causative agent of strangles, a highly contagious and serious disease in the upper respiratory tract of horses (Sweeneyet al,2005). Streptococcus equi ssp. equi shares many potentially virulence-related extracellular proteins with related pathogens within the genus Streptococcus (Harringtonet al,2002; Alberet al,2005; Lannergårdet al,2005; Karlströmet al,2006). A gene for a homolog of IdeS/Mac has been found in the genome of S. equi ssp. equi, and has previously been used in bioinformatic comparisons (Leiet al,2002, 2003). The present study describes the cloning of this gene and the recombinant expression and enzymatic characterization of the protein it encodes, which has been given the name IdeE (IgG-degrading enzyme of S. equi ssp. equi). A gene encoding a nearly identical protein can also be found in the genome of the closely related, but less virulent, horse pathogen Streptococcus equi ssp. zooepidemicus. This gene, denoted ideZ, was cloned and the activity of recombinant IdeZ is compared with the activity of IdeE. The ability of S. equi ssp. equi and zooepidemicus to degrade antibodies is studied by determining the proteolysis of IgG in cell cultures.

Materials and methods

Bacterial strains, plasmids, antibodies, sera and growth conditions

Streptococcus equi ssp. equi strain 1866 was obtained from NordVacc Läkemedel AB. Other ssp. equi (n=10) and zooepidemicus (n=10) strains used in this study were obtained from the National Veterinary Institute (SVA), Uppsala, Sweden. Escherichia coli strain ER2566 and the plasmid vector pTYB4 were obtained from New England Biolabs Incet al, MA (NEB). Purified serum IgG of different animals were obtained from Sigma (St Louis, MO), and human IgG from Kabi Vitrum (Stockholm, Sweden). Horse serum was obtained from SVA. Streptococcal strains were grown on cattle blood agar plates (SVA) or in Todd–Hewitt broth (Oxoid, Basingstoke, Hampshire, UK) supplemented with 0.5% yeast extract (THB-yeast). Escherichia coli was cultured in Luria–Bertani (LB) broth, supplemented with ampicillin (100μgmL−1), or on LAA plates [LB-broth with ampicillin and agar (15gL−1)]. Incubations were at 37°C, unless otherwise stated.

Construction of clones and purification of recombinant proteins

Recombinant proteins corresponding to amino acid 35-349 (36kDa) of IdeE and IdeZ were expressed and purified using the NEB IMPACTT7 system. Primer OSEMAC 1:5, 5′-CATGCCATGGACGATTACCAAAGGAATGCT-3′ and primer OSEMAC 2:3, 5′-CCGCTCGAGGCTCAGTTTCTGCCATATGTC-3′, were used to PCR amplify a DNA-fragment of ideE, using ssp. equi strain 1866 DNA as a template. Similarly, a DNA fragment of ideZ was amplified using ssp. zooepidemicus strain ZV DNA as a template. The underlined nucleotide sequences of the primers hybridize to the ideE/ideZ genes and the introduced restriction endonuclease cleavage sites (NcoI and XhoI) are written in bold. The PCR products were digested with NcoI and XhoI and ligated into the pTYB4-vector previously digested with the same restriction endonucleases. After ligation, the plasmids were electrotransformed into E. coli ER2566 and spread on LAA plates. The inserts were sequenced and one clone of each gene was chosen for production of recombinant proteins according to the manufacturer's (NEB) recommendations.

IgG-cleavage assays

Purified IgG (1mgmL−1) from different mammals were incubated with recombinant IdeE or IdeZ (20μgmL−1) in PBS (total volume 20μL) for 2h, or overnight, at 37°C and analyzed by reducing and nonreducing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Cell culture cleavage assays were performed by inoculating S. equi from a blood agar plate to 100μL THB-yeast supplemented with 10% horse serum, or purified IgG from selected mammals (0.3mgmL−1). Following overnight incubation at 37°C, cells were removed by centrifugation and the supernatants were analyzed by SDS-PAGE (supernatants containing serum were diluted 1:10). For the experiment where the supernatant was removed at different time points, a stock solution was made with 60μL overnight culture and 6mL THB-yeast with 10% horse sera, and then divided into six tubes.

SDS-PAGE, Western blot and MS/MS sequencing analysis

All SDS-PAGE analyses were performed under reducing conditions using the PhastSystem (Amersham) with precast 8–25% gradient gels, unless otherwise stated. IgG bands were detected by Western blotting. The proteins were diffusion-blotted onto a Nitrocellulose (NC)-membrane, blocked with PBS supplemented with Tween 20 (0.05% v/v) (PBS-T) and casein (0.5% w/v), incubated with rabbit anti-horse IgG peroxidase conjugate (Sigma) or sheep antihorse IgG4 peroxidase conjugate (Bethyl Laboratories Incet al, Montgomery, TX), washed and developed with 4-chloro-1-naphthol according to standard procedure.


Genes encoding IdeS/Mac-homologs are present in both S. equi ssp. equi and ssp. zooepidemicus

The genome S. equi ssp. equi (http://www.sanger.ac.uk/Projects/S_equi/) contains a 1047 nucleotide long ORF encoding an IdeS/Mac homolog. The corresponding gene from S. equi ssp. equi strain 1866 was PCR amplified, cloned and sequenced. This gene, denoted ideE, encodes a protein of 349 amino acids, termed IdeE. The first 34 amino acids were predicted as a signal peptide, but no putative cell wall/membrane-spanning regions, or anchoring motifs, can be found and IdeE is therefore predicted as a secreted protein. The mature IdeE protein (i.e. excluding signal peptide) shows 72% identities and 85% positives when aligned to IdeS of the S. pyogenes serotype M12. Furthermore, an ideE homolog can be found in the genome of S. equi ssp. zooepidemicus strain 4047 (http://www.sanger.ac.uk/Projects/S_zooepidemicus/). This gene, IdeZ, contains two nucleotide dissimilarities, leading to two amino acid substitutions (Glu131 to Asp and Arg314 to Gly) when compared with ideE. IdeZ was cloned from S. equi ssp. zooepidemicus strain ZV, sequenced and expressed. Differences in the ideZ gene of strain ZV compared with strain 4047 result in four amino acid substitutions. The ideE and ideZ genes are localized in the same genome context in ssp. equi and zooepidemicus and they are not flanked by phage-associated or mobile genetic elements that would indicate that they have been acquired by a recent horizontal gene transfer.

The presence of ideE was confirmed in 10 of 10 strains of S. equi ssp. equi by PCR, using primer OSEMAC 1:5 combined with OSEMAC 2:3 (data not shown). Fragments of the same length were also found in 10 of 10 S. equi ssp. zooepidemicus strains tested. The GenBank accession number for ideE of strain 1866 is DQ508733 and the accession number for ideZ of strain ZV is DQ826037.

IdeE and IdeZ are functional IgG-specific endopeptidases

The activity of recombinant IdeE was analyzed by incubation with purified IgG and compared with an untreated control by SDS-PAGE (Fig. 1). A unique band from the IdeE-treated sample was isolated and the N-terminal end was identified as GPSVFIFPP[K/N]PK by tandem mass spectrometry sequencing. The N-terminal residues correspond to the previously known IdeS/Mac-cleavage site in human IgG (Agniswamyet al,2004; Weniget al,2004).

Figure 1

Cleavage of IgG by recombinant IdeE. Purified horse IgG (1mgmL−1) was incubated overnight in PBS (1) or with recombinant IdeE (20μgmL−1) (2). Samples were separated by 8–25% SDS-PAGE under nonreducing conditions without prior boiling of the samples. The cleavage product (arrow) was identified by tandem mass spectrometry sequencing.

In a similar experiment, recombinant IdeE and IdeZ were incubated with purified IgG from a selection of animal species to analyze the substrate specificity (Fig. 2). Interestingly, the antibodies from humans, guinea-pig and dog were clearly cleaved more efficiently than those of horse. No cleavage of IgG from pig could be observed. The relative effect of IdeE and IdeZ seemed identical. To correlate the observed cleavage to differences in the hinge region of the IgG substrates, the IdeS/Mac cleavage site was aligned to the corresponding sites of pig, dog and horse (Fig. 3a). In pig, all subclasses lack the Leu-Leu-Gly motif of human IgG, explaining why IgG from this species was not cleaved. In dog, in contrast, all subclasses display a sequence similar or identical to human IgG. In horse, seven IgG subtypes are expressed, three of which are present in different allelic forms (Wagner, 2006). IgG1-3 display sequences identical, or similar, to the IdeS/Mac substrate site, whereas the other subclasses contain amino acid variations in this region that make cleavage unlikely. The lack of a Leu-Leu-Gly motif in IgG4, which constitute >60% of the total IgG concentration in horse serum (Sheoranet al,2000), could explain the low degree of proteolysis of total horse IgG. To test this hypothesis, a Western blot was performed on IdeE-treated horse IgG, using IgG4-specific antibodies. As can be seen in Fig. 3b, IgG4 can only be detected in the fraction of intact IgG heavy chains, and thus seem resistant to proteolysis of recombinant IdeE.

Figure 2

Purified IgG (1mgmL−1) from a selection of species were incubated overnight with recombinant IdeE or IdeZ (20μgmL−1) and separated by 8–25% SDS-PAGE. Lanes marked E and Z show samples incubated with IdeE and IdeZ, respectively, with the level of cleavage indicated: no cleavage (−), weak cleavage (+) or strong cleavage (++). Lanes on the left-hand side show negative controls incubated without IdeE/IdeZ.

Figure 3

(a) IdeS/Mac substrate site in human IgG aligned to the corresponding sites of the pig, dog and horse IgG subtypes. The cleavage site is indicated (arrow). Residues identical to the IdeS/Mac substrate site are in gray. (b) IdeE-treated horse IgGγ (untreated control to the left) was separated by SDS-PAGE (1) and the IgG4 fraction was detected in a Western blot, using IgG4-specific antibodies (2).

Growth in the presence of serum induces IgG degradation in S. equi ssp. equi, but not in ssp. zooepidemicus

The expression of IgG proteolysis in S. equi ssp. equi was analyzed by growing cells overnight in THB-yeast, supplemented either with horse serum or purified horse IgG. As controls, horse serum and IgG were incubated without bacteria. After incubation, the supernatant was analyzed by SDS-PAGE (Fig. 4). The supernatant from bacteria grown in the presence of serum contained a protein band similar in length to the previously identified cleavage product from IgG incubated with recombinant IdeE. A Western blot confirmed that this was an IgG heavy-chain cleavage product not present in the serum control (Fig. 4b). Further analysis by tandem mass spectrometry sequencing confirmed that the band represented a heavy-chain IgG sequence with an N-terminal end identical to the cleavage product of recombinant IdeE. A complementary Western blot with IgG4-specific antibodies was performed and the result shows that the band contains IgG of this subtype (Fig. 4c). Interestingly, no corresponding band could be seen, or detected by Western blots, from the supernatant of the bacteria grown with purified horse IgG. Together, these experiments suggest that no proteolytic activity on horse IgG is present when S. equi ssp. equi cells are grown under standard in vitro conditions, but that cells grown in the presence of horse serum express an active IgG protease. This protease, however, degrades a larger fraction of the total IgG content than does recombinant IdeE. The serum growth experiments were performed with nine different S. equi ssp. zooepidemicus isolates, five of which were isolated from horse and the remaining four isolated from cat, dog, sheep and cattle, but no differences from the negative control could be observed on gel (Fig. 4d) or by Western blots (data not shown). In addition, the growth experiments with isolates of ssp. zooepidemicus were reproduced using serum from goat and humans, two hosts of this more general pathogen, but no proteolysis of IgG could be observed (data not shown). In order to determine when IgG degradation is induced by ssp. equi during growth in the presence of serum, supernatants of cell cultures were collected at different time points and analyzed by SDS-PAGE. Proteolysis of IgG could be detected after 330min, during the logarithmic phase of growth (Fig. 4e).

Figure 4

IgG-cleavage assays in Streptococcus equi supernatants. (a) Purified horse IgG (0.3mgmL−1) was incubated overnight in growth medium alone (1) or in medium inoculated with S. equi ssp. equi cells (2). Ten percent horse sera were incubated in medium alone (3) or in medium inoculated with S. equi ssp. equi cells (4). Samples were separated on 8-25% SDS-PAGE. Molecular size markers are indicated. The culture supernatant from S. equi ssp. equi cells grown with horse sera contains an IgG cleavage product (arrow), identical to the previously identified cleavage product from IgG incubated with IdeE. (b) An anti-horse IgG Western blot confirms that the protein band in lane 4 is an IgG cleavage product not present in the other samples. (c) A Western blot shows that lane 4 contains degraded IgG4. (d) Ten percent horse sera were incubated in medium alone (1) or in medium inoculated with nine different S. equi ssp. zooepidemicus isolates (three examples shown, lanes 2–4), without generating visible IgG-degradation products. (d) Six tubes containing 10% horse serum were inoculated with S. equi ssp. equi cells. After the indicated time (in minutes), the supernatant was collected and filtered through a 0.45-μm filter. Following overnight incubation, the samples were analyzed by SDS-PAGE.


As shown previously (Leiet al,2002), IdeS/Mac allele variants from different GAS isolates can be grouped into two families, complex I and II, where the S. equi IdeS/Mac homolog is more closely related to complex II than I by direct sequence comparison. However, a key difference from the previously studied complex II variants is that IdeE and IdeZ contain only one cysteine, and therefore efficiently cleaves IgG under nonreducing conditions. Furthermore, the residues that have been proven critical (Cys94, His262) or important (Asp284 and Asp286) for IdeS/Mac endopeptidase activity by site-directed mutagenesis (Leiet al,2002, 2003) are all conserved in IdeE and IdeZ.

A study of the proteolytic activity of recombinant IdeE and IdeZ on purified IgG from a selection of mammals shows that only antibodies containing the substrate site of IdeS/Mac are cleaved, indicating that the specificities of the two enzymes are very similar. Interestingly, the proteolytic effect on IgG from horse is less effective than on IgG from e.g. dog or humans, as the hinge region of the dominating IgG isotype in horse serum (IgG4) is resistant to cleavage due to the lack of a distinct substrate site for IdeE/IdeZ.

Cell culture studies show that an active IgG protease is expressed by S. equi ssp. equi during the logarithmic phase when grown in the presence of horse serum, but that no activity is present when grown with purified IgG under standard laboratory growth conditions. No proteolysis of IgG was detected in the growth media of S. equi ssp. zooepidemicus. The fraction of ssp. equi-degraded IgG was cleaved at the same site as from the activity of IdeE; however, the presence of degraded IgG4 (Fig. 4c) might indicate that this is in fact due to another, as yet unidentified, IgG protease only expressed by this subspecies. An alternative explanation would be that the native IdeE produced in growth cultures more effectively degrades horse IgG than experiments with recombinant enzymes would suggest, for example due to post-translational modification of the enzyme or the combined effect with another bacterial factor, such as the S. equi homolog of the IgG-glycan hydrolase EndoS of S. pyogenes (Collin & Olsén, 2003). This hypothesis would make sense considering that ideE is evolutionarily conserved in an obligate horse pathogen, and IgG4 is regarded the most important IgG isotype in the humoral immune response of the horse (Sheoranet al,2000).


This work was supported by the AgriFunGen programme at the Swedish University of Agricultural Sciences and Stiftelsen Svensk Hästforskning (SSH:0447057). The authors also thank Erik Mattson and Ann-Louise Mickelsen for preparation of recombinant IdeE protein and initial cleavage experiments, Åke Engström for tandem mass spectrometry analysis and Karin Jacobsson for critically reading the manuscript.


View Abstract