ARTICLE cro

Author’s Choice

Identification of a staphylococcal complement inhibitor with

broad host specificity in equid Staphylococcus aureus strains
Received for publication, October 24, 2017, and in revised form, January 25, 2018 Published, Papers in Press, February 5, 2018, DOI 10.1074/jbc.RA117.000599
Nienke W. M. de Jong‡1, Manouk Vrieling‡§1, Brandon L. Garcia¶, Gerrit Koop!, Matt Brettmann¶, Piet C. Aerts‡,
Maartje Ruyken‡, Jos A. G. van Strijp‡, Mark Holmes**, Ewan M. Harrison‡‡, Brian V. Geisbrecht¶2,
and Suzan H. M. Rooijakkers‡2,3
From the ‡Department of Medical Microbiology, University Medical Center Utrecht, Utrecht University, 3584 CX Utrecht, The
Netherlands, the §Roslin Institute, University of Edinburgh, Easter Bush Campus, Midlothian EH25 9RG, United Kingdom, the
¶
Department of Biochemistry and Molecular Biophysics, Kansas State University, Manhattan, Kansas 66506, the !Department of
Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, 3584 CL Utrecht, The Netherlands, the **Department of
Veterinary Medicine, University of Cambridge, Cambridge CB3 0ES, United Kingdom, and the ‡‡Department of Medicine, University
of Cambridge, Addenbrooke’s Hospital, Cambridge CB2 0QQ, United Kingdom
Edited by Chris Whitfield

Staphylococcus aureus is a versatile pathogen capable of caus- The Gram-positive bacterium Staphylococcus aureus has
ing a broad range of diseases in many different hosts. S. aureus become a major risk for both human and animal health.

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

can adapt to its host through modification of its genome (e.g. by Whereas the bacterium harmlessly colonizes 30% of healthy
acquisition and exchange of mobile genetic elements that adults, it can also cause severe infections ranging from
encode host-specific virulence factors). Recently, the prophage abscesses to endocarditis, sepsis, and toxic shock syndrome (1).
!Saeq1 was discovered in S. aureus strains from six different Furthermore, S. aureus also colonizes and infects a broad range
clonal lineages almost exclusively isolated from equids. Within of animal species, including cows, sheep, goats, poultry, rabbits,
this phage, we discovered a novel variant of staphylococcal com- and horses (2). For example, S. aureus is a major cause of intra-
plement inhibitor (SCIN), a secreted protein that interferes with mammary infections in dairy cows and therefore a major eco-
activation of the human complement system, an important line nomic burden for the dairy industry (3). In horses, S. aureus can
of host defense. We here show that this equine variant of SCIN, cause both community-onset infections (joint, skin, or soft tis-
eqSCIN, is a potent blocker of equine complement system acti- sue infections (4)) and severe surgical site infections in hospi-
vation and subsequent phagocytosis of bacteria by phagocytes. talized horses.
Mechanistic studies indicate that eqSCIN blocks equine com- Previous studies on human S. aureus isolates have shown
plement activation by specific inhibition of the C3 convertase that this pathogen is highly versatile and can successfully adapt
enzyme (C3bBb). Whereas SCIN-A from human S. aureus iso- to its host and various host tissues via production of specific
lates exclusively inhibits human complement, eqSCIN repre- virulence factors. For instance, S. aureus expresses both sur-
sents the first animal-adapted SCIN variant that functions in a face-attached and secreted proteins that allow the bacterium to
broader range of hosts (horses, humans, and pigs). Binding anal- adhere to various human tissues, directly kill host cells, or block
yses suggest that the human-specific activity of SCIN-A is important immune mechanisms (5). Some of these factors are
related to amino acid differences on both sides of the SCIN-C3b encoded by mobile genetic elements, such as bacteriophages
interface. These data suggest that modification of this phage- and pathogenicity islands, which allow more rapid acquisition
encoded complement inhibitor plays a role in the host adapta- and exchange of virulence genes between strains. In the past,
tion of S. aureus and are important to understand how this our group was involved in the identification of a S. aureus bac-
pathogen transfers between different hosts. teriophage that encodes four human-specific immune evasion
molecules: the staphylococcal complement inhibitor (SCIN,4
referred to as SCIN-A hereafter), chemotaxis inhibitory
This work was supported by National Institutes of Health Grants AI113552 proteins of staphylococci, staphylokinase, and staphylococcal
and GM121511 (to B. V. G.) and by ERC Starting Grant 639209 (to enterotoxin A. This prophage !Sa3, which specifically inserts
S. H. M. R.), and this publication presents independent research supported
by the Health Innovation Challenge Fund (WT098600, HICF-T5-342), a par- in the "-hemolysin gene, was only found in human S. aureus
allel funding partnership between the Department of Health and Well- isolates (prevalence of 90% from a genetically diverse clinical
come Trust. The authors declare that they have no conflicts of interest with S. aureus strain collection (6)) and encodes secreted proteins
the contents of this article. The content is solely the responsibility of the
authors and does not necessarily represent the official views of the that solely block human immune functions. The fact that this
National Institutes of Health. phage (and its immune evasion cluster) is lost in animal-asso-
Author’s Choice—Final version free via Creative Commons CC-BY license. ciated strains (7, 8) suggests that this cluster is important for
This article contains Table S1 and Figs. S1–S3.
1
These authors contributed equally to this work.
S. aureus infection in humans (6). The gene with the highest
2
These authors shared in supervision of this work.
3
To whom correspondence should be addressed: Dept. of Medical Microbi-
4
ology, University Medical Center Utrecht and Utrecht University, Heidel- The abbreviations used are: SCIN, staphylococcal complement inhibitor; HI,
berglaan 100, 3584 CX, Utrecht, The Netherlands. Tel.: 31-887559962; heat-inactivated; AP, alternative pathway; SPR, surface plasmon reso-
E-mail: s.h.m.rooijakkers@umcutrecht.nl. nance; ANOVA, analysis of variance.

4468 J. Biol. Chem. (2018) 293(12) 4468 –4477

© 2018 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.
 Identification of staphylococcal equine SCIN

Figure 1. eqSCIN is a novel SCIN variant encoded by prophage !Saeq1. A, location of the scn-eq gene on prophage !Saeq1 of equine CC133 reference
isolate 3711. B, amino acid alignment of eqSCIN (strain 3711) with SCIN-A (Newman) shows that these SCIN variants share 57.8% sequence identity. Conserved
regions are highlighted.

prevalence on !Sa3 is scn, which encodes SCIN, a 9.8-kDa is located downstream of lukQ on the outer end of the prophage
#-helical molecule that specifically blocks activation of the near the attL phage attachment site (Fig. 1A). Previously, we
human complement system (9). found lukPQ and !Saeq1 to be associated with equid strains of

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Complement is a protein network in plasma that labels bac- S. aureus. Here, we set out to investigate the distribution of
terial cells for phagocytosis by human immune cells. This label- scn-eq. BLASTn analysis of our collection of sequenced isolates
ing is achieved via specific activation of the complement reac- (14) revealed that scn-eq was present in 29 strains from five
tion on the bacterial cell (via pattern recognition molecules) different clonal complexes (CC1, CC133, CC522, CC350, and
that subsequently drives formation of surface-bound protease CC1660) isolated in geographically distinct locations (Table
complexes (C3 convertases) that cleave C3, the major comple- S1). The majority of these strains were cultured from equid
ment protein in blood. Conversion of C3 into the reactive C3b hosts, but the gene was also present in a few ruminant isolates.
fragment results in covalent surface deposition of C3b mole- All isolates that harbored scn-eq also encoded lukQ and lukP,
cules that are recognized by complement receptors on phago- indicating that these virulence genes are prone to occur
cytic cells (10). The SCIN-A molecule, which is secreted by together. Even in isolates that do not encode a complete !Saeq1
S. aureus, specifically binds and blocks C3 convertases via a prophage, as observed in two isolates from Brazilian buffalo
unique inhibitory mechanism that is well-characterized via (14), scn-eq, lukQ, and lukP are found in close proximity to each
structural studies (11–13). Functional studies showed that other on the chromosome. The sequence identity of scn-eq was
SCIN-A is highly specific for human complement, because it highly conserved (!95%), and only a few SNPs were observed,
does not block complement activation in sera of other animals which were associated with clonal lineage (Table S1). The
(mouse, rat, dog, sheep, guinea pig, goat, and cow) (9). scn-eq sequence of the CC1660 isolate showed the largest
Here we identify a novel variant of SCIN that is present in degree of variability and contained an insertion of nine base
S. aureus isolates from horses and encoded by an equid- pairs, adding three amino acids (VKA) to the N terminus of
specific prophage (termed !Saeq1). This equine SCIN mol- eqSCIN. Altogether, we here describe a novel phage-encoded
ecule (eqSCIN) represents the first animal-adapted SCIN var- variant of SCIN associated with equid S. aureus isolates. The
iant, because it inhibits C3 convertases of horses and other scn-eq gene co-localizes with the lukPQ genes and was found in
animals. These data suggest that S. aureus can adapt to other all isolates encoding this equine-specific leukocidin.
hosts by modification of phage-encoded complement evasion
proteins. eqSCIN blocks complement activation in equine serum
To test whether eqSCIN indeed functions as a complement
Results
inhibitor, we first cloned and expressed the scn-eq gene (with-
eqSCIN is encoded on prophage !Saeq1 out the signal sequence) in Escherichia coli and purified the
Recently, a 45-kb prophage !Saeq1 was discovered in recombinant protein using nickel-affinity chromatography.
S. aureus clonal complex (CC)133 isolates from horses and Then we analyzed whether eqSCIN could block complement
donkeys that encodes the equine-specific leukocidin LukPQ activity in equine serum. Because reagents for functional com-
(14), a bi-component pore-forming toxin consisting of LukP plement analyses in equine sera are limited, we developed a C3b
and LukQ subunits (Fig. 1A). Interestingly, LukPQ has an opsonization assay on S. aureus in equine serum using flow
enhanced ability to kill equine host cells as compared with cytometry. S. aureus (commonly used laboratory strain Wood)
related staphylococcal leukocidins located elsewhere in the was incubated with equine serum, and deposition of C3b mol-
genome. Sequence analysis of CC133 reference isolate 3711 ecules onto staphylococci was determined via staining with
revealed that !Saeq1 also encodes a novel variant of SCIN-A anti-human C3b antibodies and flow cytometry. Representative
(termed eqSCIN) that shares 57.8% amino acid identity with flow plots of the assay are shown in Fig. S1. Western blotting
!Sa3-encoded SCIN-A (Fig. 1B). The gene for eqSCIN (scn-eq) was used to verify that polyclonal anti-human C3b antibodies

J. Biol. Chem. (2018) 293(12) 4468 –4477 4469

Identification of staphylococcal equine SCIN
cells in the first line of defense against S. aureus (17). To this
end, GFP-labeled S. aureus strain Wood was preincubated with
equine serum as a complement source, after which freshly iso-
lated neutrophils were added for 15 min. We used human and
equine neutrophils for our phagocytosis assays and showed that
equine complement was competent for promoting phagocyto-
sis by both human and equine neutrophils (Fig. 2, B and C).
Again, we used heat-inactivated (HI) serum as a control to show
that bacterial phagocytosis was indeed complement-depen-
dent. Previous studies have established that SCIN-A from
human isolates potently inhibits phagocytosis in the presence
of human serum (9). In concordance with the C3b deposition
experiments, we likewise observed that this SCIN-A does not
interfere with phagocytosis in the presence of equine serum.
However, the eqSCIN variant inhibited phagocytosis in equine
serum (Fig. 2, B and C). Altogether, these data show that
eqSCIN, in contrast to human SCIN-A, blocks complement
activity in equine serum.

eqSCIN interferes with equine C3 convertases

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Previous investigations from our groups showed that
SCIN-A blocks C3b deposition via a specific interaction with
the C3 convertase enzymes (11). In the alternative pathway
(AP), the C3 convertase consists of a transiently stable complex
between surface-attached C3b and protease Bb (C3bBb) (18).
For C3 cleavage to occur in this context, C3bBb first forms a 1:1
complex with its C3 substrate, and then Bb reorients itself into
proximity with the scissile bond of C3 (by virtue of Bb binding a
flexible domain of C3b). From structural studies of SCIN-A in
complex with C3bBb (11), C3b, and its C3c fragment (19), we
now know that the SCIN-A molecule makes "1,400 Å2 of con-
tact with both C3b and protease Bb. This bipartite interaction
blocks movement of Bb toward its substrate, thus rendering the
enzymatic complex inactive.
To study whether eqSCIN blocks the equine AP C3 conver-
tase in a similar manner, we first purified C3, C3b, and factor B
from equine plasma using a modified purification protocol for
Figure 2. eqSCIN blocks complement activation in equine serum. A, human complement components. We then developed an assay
eqSCIN inhibits C3b deposition on S. aureus in equine serum. Twenty percent
horse serum was incubated with 30 nM SCIN-A or 37 nM eqSCIN, after which in which activity of equine AP C3 convertase could be mea-
C3b deposition on cells of S. aureus strain Wood was measured via staining sured using these purified components. In this system, C3 con-
with anti-human C3b antibodies and flow cytometry. The extent of C3b dep- version was assessed by adding equine C3 (eqC3) as a substrate
osition is shown as the geometric mean fluorescence of all gated bacteria. B
and C, eqSCIN inhibits phagocytosis of S. aureus strain Wood by human (B) to equine AP C3 convertase (eqC3bBb) that had been produced
and equine (C) neutrophils after opsonization with a concentration range of by mixing a solution of equine C3b (eqC3b), equine factor B,
horse complement. Preincubation of serum with SCIN-A shows no effect on
phagocytosis. The amount of phagocytosis is shown as the ratio of mean FITC and human factor D. Inhibition of equine C3 cleavage was
with 0.5% serum with buffer control set as 1 for B and mean FITC for C. Error seen with eqSCIN, whereas SCIN-A failed to block eqC3
bars, S.D.; n # 3. Significance relative to buffer control was determined by conversion (Fig. 3A). We further characterized the dose-de-
one-way ANOVA with Bonferroni post-test correction for multiple compari-
son. ns, non-significant; ***, p $ 0.001; ****, p % 0.0001. pendent inhibition of eqC3bBb across eqSCIN concentra-
tions ranging from 0.04 to 10 $M, which yielded a half-max-
cross-react with horse C3 (data not shown). In 20% equine imal inhibitory concentration (IC50) of 0.61 $M (Fig. 3B).
serum, the antibodies detected C3b molecules on the bacterial These data strongly suggest that eqSCIN binds and inter-
surface, which was specific, because no signal was observed in feres with activation of eqC3bBb.
serum that was heat-treated at 56 °C (to abrogate complement Because C3bBb has a short half-life that causes rapid disso-
activity (15)). SCIN-A from human isolates effectively blocks ciation of Bb, this protease is not normally detected on target
C3b labeling in human serum (9, 16), but not in equine serum surfaces following opsonization. The binding of SCIN to C3bBb
(Fig. 2A). In contrast, eqSCIN strongly inhibited C3b labeling was found to stabilize this otherwise labile complex (11). We
by equine serum. therefore examined whether eqSCIN, like human SCIN-A,
We next studied whether eqSCIN could affect the phagocy- could also stabilize C3bBb enzymes on bacterial surfaces. To
tosis of S. aureus by neutrophils, which are critical immune address this question, we incubated S. aureus cells with equine

4470 J. Biol. Chem. (2018) 293(12) 4468 –4477

 Identification of staphylococcal equine SCIN

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Figure 3. eqSCIN interferes with equine C3 convertases. A, the ability of eqSCIN or SCIN-A to inhibit the fluid-phase equine AP convertase was assessed. An
equimolar solution of eqC3b and equine factor B was mixed with human factor D to form an equine AP convertase (eqC3bBb). Equine C3 was mixed in the
presence or absence of 1 $M eqSCIN or SCIN-A for 1 h at 37 °C. The conversion of eqC3 #-chain to #% was monitored on a reducing SDS-polyacrylamide gel. In
the presence of 1 $M eqSCIN, this conversion is inhibited, whereas in the presence of SCIN-A, the reaction proceeds in a manner similar to buffer control. B, a
2-fold serial dilution of eqSCIN (0.04 –10 $M) was incubated with eqC3bBb, and the band corresponding to the eqC3 #%-chain was quantified by densitometry
using ImageJ, where the invariant eqC3 "-chain was used to normalize each lane. These data indicate that whereas eqSCIN inhibits equine AP convertases,
SCIN-A does not. All assays were performed in duplicate, and the half-maximal inhibitory concentration (IC50) was calculated using variable non-linear
regression in GraphPad Prism version 5. C, eqSCIN stabilizes Bb on the surface of bacteria opsonized with 10 and 20% equine serum, and SCIN-A does not. Buffer
only and 1 $M SCIN-A showed background levels of Bb, whereas Bb stabilization was increased by adding 1 $M eqSCIN. One representative experiment is
shown of three independent experiments.

serum in the presence of eqSCIN, and, after washing, surface- ND) but quite highly to human C3b (KD # 140 nM) (Fig. 4 (C
bound Bb molecules were detected using Western blotting. and D) and Table 1).
Whereas human SCIN-A did not affect eqC3bBb stability at the These C3b-binding analyses suggest that SCIN-A is specific
bacterial surface, we found that eqSCIN led to stabilization of for human complement and that eqSCIN might have activity
convertases on S. aureus cells (Fig. 3C). Together, these data against both human and equine complement. To investigate
indicate that eqSCIN blocks complement activation in equine this further, we examined the ability of eqSCIN to inhibit the
serum by stabilization of an inhibited form of the equine AP C3 human complement system. We first measured C3b deposition
convertases. on S. aureus cells using human and horse sera in the presence or
absence of eqSCIN and SCIN-A (Fig. 4, E and F). In agreement
eqSCIN binds equine and human C3b with the SPR results, 100 nM eqSCIN fully inhibited (AP-depen-
Following verification that eqSCIN indeed functions as a dent) C3b deposition in human and horse serum (Fig. 4E),
convertase inhibitor in horses, we wondered whether this mol- whereas an identical concentration of SCIN-A only blocked
ecule has similar host-restricted specificity as human SCIN-A C3b deposition by human serum (Fig. 4F). Examination of
molecules. We therefore performed surface plasmon resonance eqSCIN and SCIN-A across a range of concentrations revealed
(SPR) to compare the affinity of eqSCIN with C3b molecules that eqSCIN (IC50 # 5.1 nM) (Fig. 4E) was 5 times more potent
from equine and human origin (Fig. 4 (A and B) and Table 1). as an inhibitor of human complement compared with SCIN-A
C3b was non-covalently captured on streptavidin biosensor (IC50 # 26.1 nM) (Fig. 4F).
chips using a previously described method where biotin is
site-specifically linked to the thioester domain of C3b that eqSCIN blocks human, equine, and pig complement
normally anchors C3b to a target surface (12, 13). We found The C3b-binding data described above suggested that
that eqSCIN binds with a comparably high affinity to both eqSCIN is capable of inhibition of the complement system of
eqC3b and human C3b (KD # 61 and 230 nM, respectively) multiple host species. To further explore the species specificity
(Fig. 4, A and B). In concordance with previous reports, we of eqSCIN, we used an assay of complement-dependent eryth-
observed that SCIN-A binds very weakly to eqC3b (KD # rocyte lysis to also test sera of other animals (mouse, rat, sheep,

J. Biol. Chem. (2018) 293(12) 4468 –4477 4471

Identification of staphylococcal equine SCIN

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Figure 4. eqSCIN binds equine and human C3b. A–D, characterization of eqSCIN (A and B) or SCIN-A (C and D) binding to human C3b or eqC3b by SPR. eqSCIN
binds to both equine and human C3b, whereas SCIN-A binds to only human C3b. E, eqSCIN inhibits C3b deposition on S. aureus in both 20% human and 20%
horse serum, whereas SCIN-A (F) inhibits C3b deposition only in human serum. As control, HI serum from both human and horses does not lead to C3b
deposition.

Table 1
Surface plasmon resonance: SCIN/C3b binding parameters
KDSPR,kin, equilibrium dissociation constant derived from rate constants (kd/ka); KDSPR,ss, equilibrium dissociation constant derived from steady-state fitting; ND, not
determined; could not be calculated.
Immobilized ligand Analyte KDSPR,kin ka kd KDSPR,ss
&1 &1
nM M s s&1 nM
Equine C3b eqSCIN 61 ' 3.4 (1.8 ' 1.0) ( 106 (1.1 ' 0.65) ( 10&1 220 ' 6.0
SCIN-A ND ND ND 29,000 ' 1700
Human C3b eqSCIN 230 ' 10 (5.5 ' 0.01) ( 105 (1.3 ' 0.04) ( 10&1 300 ' 12
SCIN-A 140 ' 8.0 (6.3 ' 2.6) ( 105 (9.2 ' 4.3) ( 10&2 140 ' 4.0

guinea pig, goat, cow, and pig). We initially verified that eqSCIN sheep, guinea pig, goat, and cow) revealed that eqSCIN only
could inhibit complement activation in the presence of both blocked equine, human, and pig complement, whereas SCIN-A
equine (Fig. 5A) and human serum (Fig. 5B) in this assay, inhibited the human complement system exclusively (Fig. 5D).
whereas SCIN-A only blocked human complement effectively. As a control, HI serum did not show any hemolysis (Fig. S2).
We also found that eqSCIN inhibited complement activity in Together, these data established that eqSCIN is not specific for
pig serum (Fig. 5C), albeit at only the highest concentration of equine complement but also targets the AP C3 convertase of
inhibitor tested. Testing a broader range of animals (mouse, rat, humans and pigs as well.

4472 J. Biol. Chem. (2018) 293(12) 4468 –4477

 Identification of staphylococcal equine SCIN

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Figure 5. eqSCIN blocks hemolysis in horses, humans, and pigs. eqSCIN is able to inhibit AP-dependent hemolysis of erythrocytes caused by the comple-
ment system of pigs (A), horses (B), and humans (C), whereas SCIN-A inhibits the complement system of only humans (C). As a control, HI serum did not lead to
hemolysis of the erythrocytes. D, both SCIN-A and eqSCIN (concentration of 1 $M was used) do not significantly block hemolysis in mouse, rat, sheep, guinea
pig, goat, and cow serum. Error bars, S.D.; n # 3. Significance relative to buffer control was determined by one-way ANOVA with Bonferroni post-test correction
for multiple comparisons.

Discussion human host jumps of these S. aureus lineages. Further research

In this study, we identified a variant of the S. aureus virulence may be able to address these questions. Genome sequencing of
factor SCIN-A in equid isolates. Interestingly, we observed that wider collections of S. aureus strains isolated from equids over
eqSCIN was 5 times more potent as an inhibitor of human time will help to further investigate the evolution of equid-as-
complement compared with SCIN-A (Fig. 4, E and F). This sociated S. aureus lineages and their virulence factors.
could be explained by the fact that SCIN has two major inter- Whereas SCIN-A and eqSCIN are only 58% identical (Fig.
action sites in the C3 convertase (C3bBb). In contrast to 1B), human C3b and eqC3b share nearly 80% identity at the
SCIN-A orthologues found in human S. aureus isolates that amino acid sequence level. We were interested in understand-
only block human complement (16), this newly identified ing whether differences in amino acids on either side of the
eqSCIN acts on a broader range of host species and is able to SCIN-A/C3b interface may provide an underlying structural
inhibit equine, human, and pig complement. These data on explanation for the human-specific activity of SCIN-A. Previ-
eqSCIN support previously observed broad species specificities ous work in our laboratories has produced crystal structures for
of equine-adapted S. aureus immune evasion proteins (14, 20). SCIN-A, SCIN-A–C3bBb, SCIN-A–C3b, and SCIN-A–C3c
Previous work on staphylococcal leukocidins and clotting fac- (11, 22, 23). These studies have been extended to naturally
tors in equid S. aureus isolates has identified that equine- occurring SCIN variants found in human-associated staphylo-
adapted immune evasion proteins are not strictly horse-specific coccal isolates and include atomic-resolution structures of an
but have the ability to interact with immune mediators of mul- inactive SCIN protein known as SCIN-D (12, 24). We leveraged
tiple host species. For example, the equine-adapted leukocidin this information to model a putative SCIN-A– eqC3b complex.
LukPQ kills equine neutrophils with higher efficiency than its In the context of our previous structural and biochemical stud-
closest relative in human isolates (LukED) (14). However, ies on SCIN-A– human C3b this approach afforded several
LukPQ also has the ability to lyse neutrophils of other animals potential insights into the human-specific nature of SCIN-A.
(humans and cows), albeit less efficiently, in contrast to the First, we noted that a key SCIN-A/C3b interaction involving
!Sa2-encoded leukocidin PVL that specifically kills human SCIN-A Arg-42 (25) is conserved in the eqSCIN– eqC3b model
neutrophils (21). Similarly, a broader host range was observed (Fig. 6A). Despite its relative importance in driving overall C3b
for the equine variant of the von Willebrand factor– binding affinity, multiple lines of evidence from our studies involving
protein compared with similar molecules from ruminant SCIN variants indicate that residues other than Arg-42 are also
strains (20). The observed broader species specificity of equine critical in mediating C3b interaction and complement inhibi-
virulence factors like eqSCIN as opposed to the human-specific tory activity (25). This point is underscored by the observation
activity of molecules like SCIN-A is interesting and may result that SCIN-D, which possesses an equivalently positioned Arg
from similarities between the complement proteins of these residue, fails to interact with C3b and does not inhibit comple-
hosts or possibly result from more frequent/recent host equine/ ment (11, 12). Next, we examined the C3b side of the interface.

J. Biol. Chem. (2018) 293(12) 4468 –4477 4473

Identification of staphylococcal equine SCIN

Figure 6. Modeling a putative SCIN-A/eqC3b interface. To gain insight into the human-specific nature of SCIN-A activity, the co-crystal structures of SCIN-A
(cyan) in complex with human C3b (gray) were used to model an interaction between equine C3b and SCIN-A. A, SCIN-A Arg-42 forms hydrogen bonds with
three human C3b residues (Pro-555, Ser-741, and Asp-775, marked in yellow) and is critical for mediating high-affinity SCIN-A/C3b interaction. An arginine
residue is encoded at an equivalent position in eqSCIN, and C3b residues that directly contact Arg-42 are conserved between equine and human C3b. B, in
contrast to the conserved Arg-42–mediated interaction, the Gln-49 residue of SCIN-A forms a salt bridge with human C3b residues Ala-735 and Asn-738 (shown
in yellow), whereas in equine C3b (C), this interaction would be abrogated by A735P and N738D substitutions. The equivalent eqSCIN position encodes a Tyr
residue rather than a Gln.

Of the 12 residues that bury more than 10% of total surface area eqC3b relative to SCIN-A or human C3b, respectively. To this

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

in the SCIN-A–C3c crystal structure (as judged by the EBI PISA point, the crystal structure of the inactive SCIN-D protein
server (26) using Protein Data Bank entry 3L3O), four are non- revealed a change in the secondary structure of the second
conserved in eqC3b. These include human to horse substitu- #-helix relative to SCIN-A as well as a significantly different
tions of V554L, A735P, N738D, and V740I. Whereas Val-554 surface charge profile (12). We cannot rule out the possibility
contacts SCIN-A Tyr-39, Val-740 makes contact with side- that higher-order structural differences contribute to the C3b
chain atoms of SCIN-A Lys-41 and Arg-42. Each of these specificity observed for SCIN-A versus eqSCIN. Thus, future
SCIN-A residues is conserved in eqSCIN, and analysis of a puta- structure/function studies involving eqSCIN and eqC3b are
tive SCIN-A/eqC3b interface reveals that the similar hydro- needed to address the underlying structural basis for the
phobic contacts mediated by each C3b residue are likely to be human-specific activity of SCIN-A more comprehensively.
formed even with the V554L and V740I equine substitutions. When considered as a whole, our functional experiments
Together, these observations suggest that these positions are suggest an important role for eqSCIN in the evasion of the
unlikely to significantly alter the affinity of SCIN-A for eqC3b. innate immune defense against S. aureus in horses. The close
In contrast, residues Ala-735 and Asn-738 mediate a salt-bridge association of eqSCIN with LukPQ on phage !Saeq1 implies
interaction with SCIN-A Gln-49 (Fig. 6B). This interaction is that these molecules may act in concert to evade the equine
abrogated in the SCIN-A– eqC3b model, as Pro-735 lacks the neutrophil response and contribute to S. aureus infections in
backbone amide and Asp-738 lacks the side-chain amino group horses. However, further evaluation of the clinical impact of
that participate in the Gln-49 salt bridge (Fig. 6C). Interestingly, both LukPQ and eqSCIN is clearly required to test this idea.
eqSCIN encodes a Tyr residue at this position and thus may be Importantly, zoonotic transmission of equine isolates has been
able to maintain high affinity for eqC3b via compensatory documented between hospitalized horses and their personnel
interactions. (27–29). In Europe, an epidemic subclone of CC398 MRSA of
It seems likely that subtle changes in structure resulting from almost exclusively spa-type t011 was shown to spread within
altered sequences of SCIN proteins along with corresponding and between equine hospitals (29). Interestingly, we demon-
sequence differences in C3b give rise to their differing specific- strated in our previous work that the prevalence of LukPQ in
ity profiles. Although our model seemingly implicates the equid isolates of this spa type (t011) was relatively high (14). We
A735P/N738D substitutions and the corresponding SCIN-A to can speculate that molecules like LukPQ and eqSCIN that block
eqSCIN Q49Y substitution, this alone does not explain the large innate immune responses in horses and humans may facilitate
differences in affinity that we observe for SCIN-A–C3b versus zoonotic transmission of S. aureus in settings like the veterinary
SCIN-A– eqC3b (data not shown). It is important to note sev- hospital. The same molecules may also facilitate host jumps
eral limitations inherent in this analysis. First, whereas we have between horses and other susceptible host species, like pigs
only considered the primary SCIN-A–C3b site, a second SCIN- (30). A follow-up study focusing on the population distribution
A–C3b– binding site involving residues of the N terminus of of prophage !Saeq1 in S. aureus isolates of horses, humans, and
SCIN-A is present in all SCIN-A–C3b or C3c crystal structures. pigs may reveal whether there is an association between !Saeq1
We have shown previously that this site contributes to the over- and S. aureus host switching, as was previously shown for other
all affinity of SCIN proteins for C3b (24), and candidate inter- staphylococcal mobile genetic elements (31, 32). Although our
actions analogous to the one described above can be identified understanding of S. aureus host adaptation is still largely
(i.e. where substitutions on both sides of the SCIN/C3b inter- incomplete, this study provides new insights into the ability of
face are present). Second, an assumption of this type of analysis S. aureus to evade immune responses in a host-adaptive
is that larger-scale structural changes are absent in eqSCIN or manner.

4474 J. Biol. Chem. (2018) 293(12) 4468 –4477

 Identification of staphylococcal equine SCIN
Experimental procedures analysis as geometric mean fluorescence of the gated bacteria.
Bacterial strains and genomic analysis The buffer used for C3b deposition was veronal-buffered saline
(140 mM NaCl), pH 7.4, 5 mM MgCl2, 10 mM EGTA, and 0.05%
The collection of sequenced S. aureus genomes used in this
(v/v) BSA.
study has been described earlier in our previous work (14). In
short, strains used in this study were collected as part of routine Conversion of C3
surveillance in the United Kingdom or were part of previous/
An equimolar solution of eqC3b and equine factor B (250 nM)
ongoing studies in Switzerland (33), Tunisia (34, 35), and Brazil
was mixed with 100 nM human factor D to form an equine AP
(36). BLASTn analysis was performed to identify the scn-eq
convertase (eqC3bBb). Equine C3 (2 $M) was mixed in the pres-
gene in the genome collection, and the scn-eq sequence of iso-
ence or absence of 1 $M eqSCIN or SCIN-A for 1 h at 37 °C in 20
late 3711 was used as a reference throughout the study. The
mM HEPES (7.3), 140 mM NaCl, 5 mM NiCl2. The conversion of
complete sequence of the !Saeq1 phage of isolate 3711 is avail-
eqC3 #-chain to #% was monitored on a reducing SDS-poly-
able in the Sequence Read Archive database of the European
acrylamide gel. A 2-fold serial dilution of eqSCIN (0.04 –10 $M)
Nucleotide Archive (accession number LT671578). We fur-
was incubated with eqC3bBb, and the band corresponding to
thermore used Wood 46 (ATCC-10832) for phagocytosis, C3b
the eqC3 #%-chain was quantitated by densitometry using
deposition, and convertase stabilization.
ImageJ with the invariant eqC3 "-chain to normalize each lane.
All assays were performed in duplicate, and an IC50 was calcu-
Proteins
lated using variable non-linear regression in GraphPad Prism
Recombinant SCIN with and without a His tag was prepared version 5.
as described before (9). The protein without a His tag was used

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

in the phagocytosis assay, whereas the His-tagged protein was Convertase stabilization
used for the C3b deposition and hemolysis assay. A gel with the Convertase C3bBb stabilization experiments were done as
purified proteins is shown in Fig. S3. The eqSCIN coding described before (9). In short, 2.5 ( 107 cells of S. aureus Wood
sequence excluding the signal peptide sequence was amplified were incubated for 20 min at 37 °C in veronal-buffered saline,
from strain 3711 using Phusion polymerase (Thermo Scien- pH 7.4, containing 1 mM MgCl2 and 1 mM CaCl2 plus 0.1% (v/v)
tific). PCR products were ligated into a slightly modified BSA with different concentrations of horse serum and 1 $M
expression vector pRSETB (Invitrogen Life Technologies) with SCIN-A, eqSCIN, or buffer control). After centrifugation, bac-
a non-cleavable N-terminal His6 tag. Plasmids were trans- terially associated proteins were separated by SDS-PAGE, fol-
formed in E. coli (Rosetta-gami(DE3)pLysS (Novagen, Merck lowed by immunoblot. Horse Bb was detected with goat anti-
Biosciences)), and protein expression was induced using 1 mM human factor B (Complement Technology) followed by
isopropyl "-D-1-isogalactopyranoside. Bacterial pellets were peroxidase-conjugated anti-goat IgG (Santa Cruz Biotechnol-
lysed using 200 $g/ml lysozyme and three freeze-thaw cycles in ogy, Inc.).
20 mM sodium phosphate (pH 7.8) to isolate the eqSCIN pro-
tein. His-tagged protein was purified using nickel-affinity chro- Hemolysis assay
matography (HiTrap chelating, HP, GE Healthcare) with an The alternative pathway hemolytic assay was performed by
imidazole gradient ranging from 10 to 250 mM (Sigma-Aldrich). incubating 20% serum of different animals (30% for horse
Finally, purified eqSCIN was dialyzed to PBS and stored at serum) with various concentrations of SCIN-A or eqSCIN (0 –1
&20 °C. $M) or one fixed concentration of 1 $M and 2 ( 107 rabbit
erythrocytes (Biotrading) for 1 h at 37 °C in veronal-buffered
Phagocytosis and C3b deposition on S. aureus saline containing 5 mM MgCl2 and 10 mM EGTA. A450 nm was
Informed consent for blood draw was obtained from all sub- measured from supernatant of lysed cells.
jects, in accordance with the Declaration of Helsinki. Approval
from the medical ethics committee of the University Medical Protein-protein interaction assays by SPR
Center Utrecht was attained (METC-protocol 07-125/C Direct binding of SCIN-A and eqSCIN to human or equine
approved on March 1, 2010). Blood was collected immediately C3b was assessed by SPR on a Biacore T200 instrument. All
upon death from four healthy horses during the slaughter pro- experiments were performed at 25 °C in a running buffer of 20
cess in tubes containing 3 mM EDTA anticoagulant. Equine mM HEPES (pH 7.3), 140 mM NaCl, 0.005% (v/v), Tween 20
neutrophils were isolated using 70% Percoll gradients as (HBS-T) using a flow rate of 30 $l min&1. Site-specifically bioti-
described before (37). Phagocytosis assays were carried out as nylated human or equine C3b was prepared using protocols
described before (38). In short, 2.5 ( 105 cells of S. aureus described previously (12, 13). C3b surfaces were prepared by
Wood expressing GFP were incubated with 1 $M SCIN-A or capturing biotinylated C3b on a CMD-200 sensor chip (Xantec)
eqSCIN (or buffer control without SCIN-A/eqSCIN), 0 – 0.5% that had previously been coupled with neutravidin (Sigma-Al-
(v/v) serum, and freshly isolated neutrophils for 15 min at 37 °C. drich). eqC3b-biotin was captured at 3,000 resonance units and
For C3b deposition, 20% horse and 20% human serum were human C3b-biotin at 4,200 response units. A reference surface
preincubated with 30 nM SCIN-A or 37 nM eqSCIN and added was generated by injecting with biotin only. SCIN-A or eqSCIN
to 2.5 ( 107 washed cells of S. aureus Wood WT. Surface- was injected for 2 min over each C3b surface in 2-fold serial
bound C3b was stained with anti-C3 conjugated with FITC dilutions ranging from 2.5 to 2,500 nM. The dissociation of
(Protos Immunoresearch) and measured by flow cytometry SCIN–C3b complexes was monitored for 3 min. Baseline signal

J. Biol. Chem. (2018) 293(12) 4468 –4477 4475

Identification of staphylococcal equine SCIN
was achieved without the use of regeneration solutions. Each 11. Rooijakkers, S. H. M., Wu, J., Ruyken, M., van Domselaar, R., Planken,
injection series was performed in triplicate. Data were fit to a K. L., Tzekou, A., Ricklin, D., Lambris, J. D., Janssen, B. J. C., van Strijp, J. A.,
and Gros, P. (2009) Structural and functional implications of the alterna-
1:1 kinetic model of binding (red lines) as well a steady-state
tive complement pathway C3 convertase stabilized by a staphylococcal
binding model (inset plots) using BiaEvaluation T200 software. inhibitor. Nat. Immunol. 10, 721–727 CrossRef Medline
In the case of SCIN-A– eqC3b, data could not be fit to a kinetic 12. Garcia, B. L., Summers, B. J., Lin, Z., Ramyar, K. X., Ricklin, D., Kamath,
model, and thus a steady-state affinity was estimated. As the D. V., Fu, Z. Q., Lambris, J. D., and Geisbrecht, B. V. (2012) Diversity in the
concentrations used for SCIN-A– eqC3b were subsaturat- C3b convertase contact residues and tertiary structures of the staphylo-
ing, this was achieved by fitting these data using a constant coccal complement inhibitor (SCIN) protein family. J. Biol. Chem. 287,
maximal observable signal (Rmax) based on the eqSCIN/ 628 – 640 CrossRef Medline
13. Ricklin, D., Tzekou, A., Garcia, B. L., Hammel, M., McWhorter, W. J.,
eqC3b interaction.
Sfyroera, G., Wu, Y.-Q., Holers, V. M., Herbert, A. P. Barlow, P. N., Geis-
brecht, B. V., and Lambris, J. D. (2009) A molecular insight into comple-
Statistical analyses
ment evasion by the staphylococcal complement inhibitor protein family.
Statistical analysis was performed with GraphPad Prism ver- J. Immunol. 183, 2565–2574 CrossRef Medline
sion 7. Statistical significance was calculated using one-way 14. Koop, G., Vrieling, M., Storisteanu, D. M. L., Lok, L. S. C., Monie, T., van
ANOVA and Student’s t test. Wigcheren, G., Raisen, C., Ba, X., Gleadall, N., Hadjirin, N., Timmerman,
A. J., Wagenaar, J. A., Klunder, H. M., Fitzgerald, J. R., Zadoks, R., et al.
(2017) Identification of LukPQ, a novel, equid-adapted leukocidin of
Author contributions—N. W. d. J. and M. V. conceptualization;
Staphylococcus aureus. Sci. Rep. 7, 40660 CrossRef Medline
N. W. d. J., M. V., B. L. G., M. B., P. C. A., M. R., and B. V. G. data
15. Joisel, F., Leroux-Nicollet, I., Lebreton, J. P., and Fontaine, M. (1983) A
curation; N. W. d. J., M. V., B. L. G., G. K., J. A. v. S., M. H., E. M. H., hemolytic assay for clinical investigation of human C2. J. Immunol. Meth-
B. V. G., and S. H. R. formal analysis; N. W. d. J., M. V., B. L. G., G. K., ods 59, 229 –235 CrossRef Medline

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

M. B., P. C. A., M. R., J. A. v. S., M. H., E. M. H., B. V. G., and S. H. R. 16. Jongerius, I., Köhl, J., Pandey, M. K., Ruyken, M., van Kessel, K. P. M., van
investigation; N. W. d. J., M. V., B. L. G., M. B., P. C. A., M. R., and Strijp, J. A., and Rooijakkers, S. H. M. (2007) Staphylococcal complement
B. V. G. methodology; N. W. d. J., M. V., B. L. G., G. K., B. V. G., and evasion by various convertase-blocking molecules. J. Exp. Med. 204,
S. H. R. writing-original draft; N. W. d. J. and S. H. R. project admin- 2461–2471 CrossRef Medline
istration; B. L. G. software; G. K., J. A. v. S., M. H., E. M. H., B. V. G., 17. Nauseef, W. M. (2007) How human neutrophils kill and degrade mi-
and S. H. R. writing-review and editing; J. A. v. S., B. V. G., and S. H. R. crobes: an integrated view. Immunol. Rev. 219, 88 –102 CrossRef Medline
supervision; E. M. H., B. V. G., and S. H. R. funding acquisition. 18. Xu, Y., Narayana, S. V., and Volanakis, J. E. (2001) Structural biology of the
alternative pathway convertase. Immunol. Rev. 180, 123–135 CrossRef
Acknowledgments—We thank Dr. Bryan Wee for providing bio-infor- Medline
19. Garcia, B. L., Tzekou, A., Ramyar, K. X., McWhorter, W. J., Ricklin, D.,
matic advice and Prof. Ross Fitzgerald for critical review of the
Lambris, J. D., and Geisbrecht, B. V. (2009) Crystallization of human com-
manuscript.
plement component C3b in the presence of a staphylococcal comple-
ment-inhibitor protein (SCIN). Acta Crystallogr. Sect. F Struct. Biol. Cryst.
References Commun. 65, 482– 485 CrossRef Medline
20. Viana, D., Blanco, J., Tormo-Más, M. A., Selva, L., Guinane, C. M., Baselga,
1. Lowy, F. D. (1998) Staphylococcus aureus infections. N. Engl. J. Med. 339,
R., Corpa, J. M., Lasa, I., Novick, R. P., Fitzgerald, J. R., and Penadés, J. R.
520 –532 CrossRef Medline
(2010) Adaptation of Staphylococcus aureus to ruminant and equine hosts
2. Fitzgerald, J. R. (2012) Livestock-associated Staphylococcus aureus: origin,
involves SaPI-carried variants of von Willebrand factor-binding protein.
evolution and public health threat. Trends Microbiol. 20, 192–198
Mol. Microbiol. 77, 1583–1594 CrossRef Medline
CrossRef Medline
3. Bradley, A. (2002) Bovine mastitis: an evolving disease. Vet. J. 164, 21. McCarthy, A. J., Witney, A. A., and Lindsay, J. A. (2012) Staphylococcus
116 –128 CrossRef Medline aureus temperate bacteriophage: carriage and horizontal gene transfer is
4. Weese, J. S., and van Duijkeren, E. (2010) Methicillin-resistant Staphylo- lineage associated. Front. Cell Infect. Microbiol. 2, 6 Medline
coccus aureus and Staphylococcus pseudintermedius in veterinary medi- 22. Rooijakkers, S. H. M., Milder, F. J., Bardoel, B. W., Ruyken, M., van Strijp,
cine. Vet. Microbiol. 140, 418 – 429 CrossRef Medline J. A., and Gros, P. (2007) Staphylococcal complement inhibitor: structure
5. Spaan, A. N., Surewaard, B. G. J., Nijland, R., and van Strijp, J. A. (2013) and active sites. J. Immunol. 179, 2989 –2998 CrossRef Medline
Neutrophils versus Staphylococcus aureus: a biological tug of war. Annu. 23. Garcia, B. L., Ramyar, K. X., Tzekou, A., Ricklin, D., McWhorter, W. J.,
Rev. Microbiol. 67, 629 – 650 CrossRef Medline Lambris, J. D., and Geisbrecht, B. V. (2010) Molecular basis for comple-
6. van Wamel, W. J. B., Rooijakkers, S. H. M., Ruyken, M., van Kessel, ment recognition and inhibition determined by crystallographic studies of
K. P. M., and van Strijp, J. A. G. (2006) The innate immune modulators the staphylococcal complement inhibitor (SCIN) bound to C3c and C3b.
staphylococcal complement inhibitor and chemotaxis inhibitory protein J. Mol. Biol. 402, 17–29 CrossRef Medline
of Staphylococcus aureus are located on "-hemolysin-converting bacte- 24. Garcia, B. L., Summers, B. J., Ramyar, K. X., Tzekou, A., Lin, Z., Ricklin, D.,
riophages. J. Bacteriol. 188, 1310 –1315 CrossRef Medline Lambris, J. D., Laity, J. H., and Geisbrecht, B. V. (2013) A structurally
7. Cuny, C., Abdelbary, M., Layer, F., Werner, G., and Witte, W. (2015) dynamic N-terminal helix is a key functional determinant in staphylococ-
Prevalence of the immune evasion gene cluster in Staphylococcus aureus cal complement inhibitor (SCIN) proteins. J. Biol. Chem. 288, 2870 –2881
CC398. Vet. Microbiol. 177, 219 –223 CrossRef Medline CrossRef Medline
8. Sung, J. M. L., Lloyd, D. H., and Lindsay, J. A. (2008) Staphylococcus aureus 25. Summers, B. J., Garcia, B. L., Woehl, J. L., Ramyar, K. X., Yao, X., and
host specificity: comparative genomics of human versus animal isolates by Geisbrecht, B. V. (2015) Identification of peptidic inhibitors of the alter-
multi-strain microarray. Microbiology 154, 1949 –1959 CrossRef Medline native complement pathway based on Staphylococcus aureus SCIN pro-
9. Rooijakkers, S. H. M., Ruyken, M., Roos, A., Daha, M. R., Presanis, J. S., teins. Mol. Immunol. 67, 193–205 CrossRef Medline
Sim, R. B., van Wamel, W. J. B., van Kessel, K. P. M., and van Strijp, J. A. G. 26. Krissinel, E., and Henrick, K. (2007) Inference of macromolecular assem-
(2005) Immune evasion by a staphylococcal complement inhibitor that blies from crystalline state. J. Mol. Biol. 372, 774 –797 CrossRef Medline
acts on C3 convertases. Nat. Immunol. 6, 920 –927 CrossRef Medline 27. Cuny, C., Strommenger, B., Witte, W., and Stanek, C. (2008) Clusters of
10. Gasque, P. (2004) Complement: a unique innate immune sensor for dan- infections in horses with MRSA ST1, ST254, and ST398 in a veterinary
ger signals. Mol. Immunol. 41, 1089 –1098 CrossRef Medline hospital. Microb. Drug Resist. 14, 307–310 CrossRef Medline

4476 J. Biol. Chem. (2018) 293(12) 4468 –4477

 Identification of staphylococcal equine SCIN
28. van Duijkeren, E., Moleman, M., Sloet van Oldruitenborgh-Oosterbaan, 33. Sieber, S., Gerber, V., Jandova, V., Rossano, A., Evison, J. M., and Perreten,
M. M., Multem, J., Troelstra, A., Fluit, A. C., van Wamel, W. J. B., Houw- V. (2011) Evolution of multidrug-resistant Staphylococcus aureus infec-
ers, D. J., de Neeling, A. J., and Wagenaar, J. A. (2010) Methicillin-resistant tions in horses and colonized personnel in an equine clinic between 2005
Staphylococcus aureus in horses and horse personnel: an investigation of and 2010. Microb. Drug Resist. 17, 471– 478 CrossRef Medline
several outbreaks. Vet. Microbiol. 141, 96 –102 CrossRef Medline 34. Gharsa, H., Ben Sallem, R., Ben Slama, K., Gómez-Sanz, E., Lozano, C.,
29. Abdelbary, M. M. H., Wittenberg, A., Cuny, C., Layer, F., Kurt, K., Wieler, Jouini, A., Klibi, N., Zarazaga, M., Boudabous, A., and Torres, C. (2012)
L. H., Walther, B., Skov, R., Larsen, J., Hasman, H., Fitzgerald, J. R., Smith, High diversity of genetic lineages and virulence genes in nasal Staphylo-
T. C., Wagenaar, J. A., Pantosti, A., Hallin, M., et al. (2014) Phylogenetic coccus aureus isolates from donkeys destined to food consumption in
analysis of Staphylococcus aureus CC398 reveals a sub-lineage epidemio- Tunisia with predominance of the ruminant associated CC133 lineage.
logically associated with infections in horses. PLoS One 9, e88083 BMC Vet. Res. 8, 203 CrossRef Medline
CrossRef Medline
35. Gharsa, H., Ben Slama, K., Gómez-Sanz, E., Lozano, C., Zarazaga, M.,
30. Islam, M. Z., Espinosa-Gongora, C., Damborg, P., Sieber, R. N., Munk, R.,
Messadi, L., Boudabous, A., and Torres, C. (2015) Molecular characteriza-
Husted, L., Moodley, A., Skov, R., Larsen, J., and Guardabassi, L. (2017)
tion of Staphylococcus aureus from nasal samples of healthy farm animals
Horses in Denmark are a reservoir of diverse clones of methicillin-resist-
and pets in Tunisia. Vector Borne Zoonotic Dis. 15, 109 –115 CrossRef
ant and -susceptible Staphylococcus aureus. Front. Microbiol. 8, 543
Medline
Medline
31. Guinane, C. M., Ben Zakour, N. L., Tormo-Mas, M. A., Weinert, L. A., 36. Aires-de-Sousa, M., Parente, C. E. S. R., Vieira-da-Motta, O., Bonna,
Lowder, B. V., Cartwright, R. A., Smyth, D. S., Smyth, C. J., Lindsay, J. A., I. C. F., Silva, D. A., and de Lencastre, H. (2007) Characterization of Staph-
Gould, K. A., Witney, A., Hinds, J., Bollback, J. P., Rambaut, A., Penadés, ylococcus aureus isolates from buffalo, bovine, ovine, and caprine milk
J. R., and Fitzgerald, J. R. (2010) Evolutionary genomics of Staphylococcus samples collected in Rio de Janeiro State, Brazil. Appl. Environ. Microbiol.
aureus reveals insights into the origin and molecular basis of ruminant 73, 3845–3849 CrossRef Medline
host adaptation. Genome Biol. Evol. 2, 454 – 466 CrossRef Medline 37. Siemsen, D. W., Schepetkin, I. A., Kirpotina, L. N., Lei, B., and Quinn,
32. Lowder, B. V., Guinane, C. M., Ben Zakour, N. L., Weinert, L. A., Conway- M. T. (2007) Neutrophil isolation from nonhuman species. Methods Mol.
Biol. 412, 21–34 CrossRef Medline

Downloaded from http://www.jbc.org/ by guest on April 12, 2018

Morris, A., Cartwright, R. A., Simpson, A. J., Rambaut, A., Nübel, U., and
Fitzgerald, J. R. (2009) Recent human-to-poultry host jump, adaptation, 38. Rooijakkers, S. H. M., van Wamel, W. J. B., Ruyken, M., van Kessel,
and pandemic spread of Staphylococcus aureus. Proc. Natl. Acad. Sci. K. P. M., and van Strijp, J. A. (2005) Anti-opsonic properties of staphyloki-
U.S.A. 106, 19545–19550 CrossRef Medline nase. Microbes Infect. 7, 476 – 484 CrossRef Medline

J. Biol. Chem. (2018) 293(12) 4468 –4477 4477