J. Comp. Path. 2012, Vol. 147, 343e353
Available online at www.sciencedirect.com
www.elsevier.com/locate/jcpa
EXPERIMENTALLY INDUCED DISEASE
Pathology and Biofilm Formation in a Porcine
Model of Staphylococcal Osteomyelitis
L. K. Johansen*, J. Koch*, D. Frees*, B. Aalbæk*, O. L. Nielsen*,
P. S. Leifsson*, T. M. Iburg*,†, E. Svalastoga‡, L. E. Buelund‡,
T. Bjarnsholtx,k, N. Høibyx,k and H. E. Jensen*
* Department of Veterinary Disease Biology, Faculty of Life Sciences, University of Copenhagen, † Department of Pathology
and Wildlife Diseases, Veterinary Institute, Uppsala, Sweden, ‡ Department of Small Animal Clinical Sciences, Faculty of Life
Sciences, University of Copenhagen, x Department of International Health, Immunology and Microbiology, Faculty of Health
Sciences, University of Copenhagen and k Department of Clinical Microbiology, Rigshospitalet, Copenhagen, Denmark
Summary
A porcine model was used to examine the potential of human and porcine Staphylococcus aureus isolates to induce
haematogenously spread osteomyelitis. Pigs were inoculated in the right femoral artery with one of the following S. aureus strains: S54F9 (from a porcine lung abscess; n ¼ 3 animals), NCTC-8325-4 (a laboratory strain of
human origin; n ¼ 3 animals) and UAMS-1 (a human osteomyelitis isolate; n ¼ 3 animals). Two pigs were
sham inoculated with saline. At 11 or 15 days post infection the animals were scanned by computed tomography before being killed and subjected to necropsy examination. Osteomyelitis lesions were present in the right
hind limb of all pigs inoculated with strain S54F9 and in one pig inoculated with strain NCTC-8325-4. Microscopically, there was extensive loss of bone tissue with surrounding granulation tissue. Sequestrated bone trabeculae were intermingled with colonies of S. aureus as demonstrated immunohistochemically. By peptide
nucleic acid fluorescence in situ hybridization bacterial aggregates were demonstrated to be embedded in an
opaque matrix, indicating that the bacteria had formed a biofilm. Development of experimental osteomyelitis
was therefore dependent on the strain of bacteria inoculated and on the formation of a biofilm.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords: biofilm; osteomyelitis; pathology; pig model; Staphylococcus aureus
Introduction
Staphylococcus aureus is the most common cause of
haematogenously spread osteomyelitis (HO), a disease which occurs primarily in children (Lew and
Waldvogel, 1997). An episode of bacteraemia leads
to settling of bacteria in bones and initiation of a focus
of osteomyelitis (Lew and Waldvogel, 1997). A reliable animal model of HO should therefore be based
on inoculation of bacteria directly into the bloodstream. A variety of such models has been described
(Table 1). Some models use bacterial strains isolated
from the same animal species, while others use human
strains sometimes isolated from patients with osteomyelitis. The origin of the bacterial strain is likely to
Correspondence to: L. K. Johansen (e-mail: [email protected]).
0021-9975/$ - see front matter
doi:10.1016/j.jcpa.2012.01.018
influence the outcome of the lesions induced in
a model, due to potential host specificity and differential expression of virulence factors (Sung et al., 2008;
Wright and Nair, 2010). One of the main
virulence factors associated with the development of
staphylococcal osteomyelitis, and in particular the
development of chronic disease, is the formation of
a bacterial biofilm (Gristina et al., 1985; Brady et al.,
2008). Biofilms develop on inert surfaces or dead
tissue and constitute a protected environment for
the bacteria in which they may avoid the effects of
antibiotics and host immune defence (Hoiby et al.,
2010). The biofilm itself therefore induces an ongoing
inflammatory reaction (Costerton et al., 1999).
Intra-arterial inoculation of pigs with a lower number of a porcine S. aureus strain (S54F9) than used in
other animal models of HO results in the development
Ó 2012 Elsevier Ltd. All rights reserved.
344
Table 1
Animal models of osteomyelitis based on haematogenous injection of Staphylococcus aureus
Route of
inoculation
Intravenous
Animal species
Inoculum
Results
Strain of S. aureus
Origin
CFU injected*
Osteomyelitis
frequency (%)
Primary systemic effects
Rabbit
No
Micrococcus
NR
NR
NR
Rabbit, dog
Dog
Rabbit
No
No
No
Attenuated S. aureus
NR
OH 172
Human osteomyelitis
Human and canine
Human osteomyelitis
NR
NR
NR
NR
NR
58.8
Sepsis and non-osseous
abscesses
None
Rabbit
(immunized)
Rabbit
Rabbit†
Rabbit†
Rabbit†
Chicken
Chicken
Chicken
Rabbit
Rabbit
Yes
NR
Human osteomyelitis
NR
50
Sepsis and non-osseous
abscesses
Sepsis
No
No
No
No
No
No
No
Yes
Yes
S. aureus
S. aureus
S. aureus
S. aureus
6/42E/53/77/83A/84‡
6/42E/53/77/83A/84‡
6/42E/53/77/83A/84‡
ATCC-25932
ATCC-25932
Human osteomyelitis
NR
NR
NR
Avian infection
Avian infection
Avian infection
Laboratory strain
Laboratory strain
NR
NR
NR
NR
107b
108e109
108
107x
106k
38
NR
NR
NR
100
NR
NR
100
100
Non-osseous abscesses
e
e
e
None
None
None
Not examined
Not examined
Chicken
Rat
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
No
Yes
No
Yes
Yes
No
No
No
Type 8 capsular isolate
S. aureus Phillips
M138
LS-1
LS-1
UAMS-1
UAMS-237
RN6390
107
108
106
107
107
107
107
108
100
100
100
100
100
70
5
100
None
None
Non-osseous abscesses
Not examined
Not examined
Not examined
Not examined
Not examined
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Pig
No
No
No
No
No
No
No
UAMS-957
UAMS-1
UAMS-155
UAMS-929
UAMS-930
UAMS-969
S54F9
Human arthritis
Human osteomyelitis
Human infection
Mouse pathogen
Mouse pathogen
Human osteomyelitis
(Mutation of UAMS-1)
NCTC-8325-4, human
infection
RN6390
Human osteomyelitis
(Mutation of UAMS-1)
(Mutation of UAMS-1)
(Mutation of UAMS-1)
(Mutation of UAMS-1)
Porcine lung abscess
108
108
108
108
108
108
108
44.4
96.3
61.8
37
18.7
90
100
Not examined
Not examined
Not examined
Not examined
Not examined
Not examined
Non-osseous
microabscesses
References
Rodet (1884), 1973
Lexer, 1894
Starr, 2002
Thompson and Dubos,
1938
Scheman et al., 1941
Weaver and Tayler, 1943
Kadyrov et al., 1966
Koiunderliev, 1971
Holland, 1972
Emslie and Nade, 1983
Speers and Nade, 1985
Alderson et al., 1986
Whalen et al., 1988
Morrissy and Haynes,
1989
Daum et al., 1990
Hienz et al., 1995
Matsushita et al., 1997
Chadha et al., 1999
Yoon et al., 1999
Elasri et al., 2002
Elasri et al., 2002
Blevins et al., 2003
Blevins et al., 2003
Blevins et al., 2003
Blevins et al., 2003
Blevins et al., 2003
Blevins et al., 2003
Blevins et al., 2003
Jensen et al., 2010
L.K. Johansen et al.
Bone trauma
or artificial
necrosis
Porcine Model of Osteomyelitis
Animals
NR, not recorded.
*
The reported number of CFU injected is an approximate number based on the information obtained in the original paper.
†
The information is obtained from a review (An et al., 2004).
‡
Phage type.
x
CFU/kg BW.
k
CFU/100 g BW.
100
33
0
104x
104x
104x
Porcine lung abscess
Human infection
Human osteomyelitis
S54F9
NCTC-8325-4
UAMS-1
No
No
No
Group B
Group C
Group D
Present study
Pig
Yes
No
52, 80, 81‡
S54F9
Human infection
Porcine lung abscess
NR
100
None
None
None
Materials and Methods
105
104x
Sepsis, pneumonia
None
Deysine et al., 1976
Johansen et al., 2011
of osteomyelitis only in bones supplied by the artery
into which the injection is made (Johansen et al.,
2011). The aims of the present study were (1) to compare the infection potential of the porcine strain
(S54F9) with two S. aureus strains of human origin
in this model and (2) to examine the development
of HO with a focus on pathology and the localization
and microenvironment of S. aureus.
Dog
Pig
Intra-arterial
345
Eleven healthy 12-week-old female Yorkshiree
Landrace-cross pigs, with a body weight (BW) of approximately 30 kg were obtained from a specific
pathogen free (SPF) herd (Harris and Alexandra,
1999). On arrival, the animals were allowed to acclimatize for 14 days. The animals were fed a commercial pig diet (Svine Erantis Brogaarden ApS, Lynge,
Denmark) ad libitum and had free access to tap water.
The Danish Animal Experimental Act approved the
protocol (licence number 2008/561-37).
Experimental Design
The animals were randomly assigned into four groups
(A, B, C and D) housed in separate pens (Table 1).
The control group of two animals (group A) was inoculated with saline while the three infected groups of
three animals each were inoculated with S. aureus
strains S54F9 (group B), NCTC-8325-4 (group C)
or UAMS-1(group D). Premedication was performed
with an intramuscular injection (1 ml/10 kg BW) of
a solution containing a mixture of 125 mg zolezepam,
125 mg tiletamine, 6.25 ml xylazine (20 mg/ml),
1.25 ml ketamine (100 mg/ml), 2 ml butorphanol
(10 mg/ml) and 2 ml metadon (10 mg/ml). A catheter
was inserted into a lateral ear vein and anaesthesia
was maintained with propofol (10 mg/ml) 1 ml/kg
BW/h. While anaesthetized, the pigs were inoculated
in the direction of the blood flow into the right femoral artery. The animals were killed at 11 or 15 days
post inoculation (dpi) by 20% pentobarbital given intravenously. Blood samples for screening of bacteraemia were taken from the jugular vein prior to
inoculation, 20 min after inoculation and again at 1,
7, 10 and 15 dpi.
Bacterial Isolates
The S. aureus strains of human origin were UAMS-1
and NCTC-8325-4, which have been used previously
for induction of osteomyelitis in mice (Blevins et al.,
2003). NCTC-8325-4 is derived from a case of human
sepsis (Cassat et al., 2005). UAMS-1 is a human
346
L.K. Johansen et al.
osteomyelitis isolate that has also become a preferred
strain for studying the virulence of S. aureus (Cassat
et al., 2005). The S. aureus isolate S54F9 was obtained
from a chronic embolic pulmonary abscess in a Danish
slaughter pig (Leifsson et al., 2010). The strains of S.
aureus were prepared as described by Johansen et al.
(2011) and diluted with sterile isotonic saline 0.9%
to obtain an inoculation dose of 10,000 colony forming units (CFU)/kg BW in a volume of 1 ml.
Clinical Examination
All animals underwent clinical examination during
the acclimatization period and subsequently twice
daily after inoculation. In cases of clinical signs of
disease, the frequency of clinical examinations was
increased and lame animals were treated with intramuscular injections (0.1 mg/kg BW) of buprenorfin (Temgesic 0.3 mg/ml; Schering-Plough,
Heist-op-den-Berg, Belgium) every 6 h. The BW
of the pigs was measured before inoculation and
at 8 dpi.
Blood Samples
For screening of bacteraemia, 4 ml blood samples
were collected and analysed as previously described
(Johansen et al., 2011).
Computed Tomography
The location of osteomyelitic lesions was evaluated by
computed tomography (CT). All pigs were scanned
under anaesthesia (see Experimental Design) to prevent movement, before inoculation and again immediately after death. The scanning was performed with
a single slide CT scanner (Siemens Somatom Emotion, Erlangen, Germany). The pigs were positioned
in ventral recumbency with the hind limbs pulled
caudally and scanned in the craniocaudal direction
from the hip to the digits with a slide thickness of
5 mm (kV ¼ 130 and mAs ¼ 55). The scans were reconstructed using a standard soft tissue algorithm
(B40s) and evaluated on a dedicated workstation
(Siemens Leonardo, Erlangen, Germany).
Pathology
The femoral and tibial bones were sectioned sagittally
and scored using a classification system modified from
that described by Rissing et al. (1985): 0, no lesion; 1,
abscessation; 2, abscessation with penetration
through the cortical bone. Samples of the dorsal border of the left diaphragmatic lung lobe, right kidney,
spleen and liver were taken for microscopical examination. Samples from the distal growth plate of the
right and left femur and the proximal growth plate
of the right and left tibia were collected using an oscillating saw (IM-MAX MEDICAL, Frederiksberg,
Denmark). Due to the fragility and small size of the
bones in the distal hindlimbs (distal to the tibial diaphysis), these were fixed in toto. All tissues were fixed
in 10% neutral buffered formalin for 3 days, after
which the osseous tissues were decalcified (Johansen
et al., 2011). After decalcification, bones of the distal
hindlimbs were sectioned sagittally and scored as described above.
After fixation and/or decalcification, tissues were
processed routinely and embedded in paraffin wax.
Sections (4e5 mm) were stained with haematoxylin
and eosin (HE). Immunohistochemistry (IHC) was
used for in situ identification of S. aureus. The primary
antibody was a murine monoclonal antibody specific
for S. aureus (ab37644; Abcam, Cambridge, UK)
(Johansen et al., 2011).
Microbiology
Tissue samples from the lungs and the distal metaphyseal area of the right and left femur and the proximal
metaphyseal area of the right and left tibia were also
collected for quantitative bacteriology. Metaphyseal
bone marrow tissue without gross lesions was drilled
out with a 13 mm drill bit sterilized in boiling water
and mounted on an electrical drill. The samples were
analysed as described by Jensen et al. (2010). Swabs
were taken from the bone abscesses and streaked on
Luria-Bertani (LB) agar and incubated for 24 h at
37 C. All isolates were characterized using API ID
32 Staph (Biomerieux Inc., Marcy-l’Etoile, France).
Peptide Nucleic Acid Fluorescence In situ Hybridization
Paraffin wax-embedded femoral bone lesions from all
pigs of group B and pig C2 were selected for peptide
nucleic acid fluorescence in situ hybridization (PNA
FISH). Sections (4 mm) were mounted on adhesive
slides (Superfrost R plus, Menzel-Glaser, Germany)
and dewaxed (Bjarnsholt et al., 2009a). A Texas
red-labelled, S. aureus-specific PNA probe in hybridization solution (AdvanDx Inc., Woburn, Massachusetts, USA) was added to each section and hybridized
on a PNA FISH workstation at 55 C for 90 min. The
slides were washed for 30 min at 55 C in wash solution
(AdvanDx Inc.). VECTASHIELD mounting media
with 40 ,6-diamidino-2-phenylindole (DAPI) (Vector
laboratories, Burlingame, California, USA) was applied and a cover slip was fixed to each slide. Slides
were examined using a fluorescence microscope
equipped with a Texas red and DAPI filter
(Bjarnsholt et al., 2009a).
Porcine Model of Osteomyelitis
347
Results
Clinical Observations
All animals were healthy prior to inoculation. At
5 dpi all pigs in group B showed lameness of the inoculated limb. These animals displayed pain on palpation of the affected bones and had intermittent mildly
elevated body temperature. Clinical signs were not
observed in pigs of groups A, C and D. Although
pigs from group B had a lower weight gain compared
with the other groups, all animals gained weight during the experiment.
Blood Samples
No bacteria were isolated from blood at any time
point of the experiment.
Computed Tomography
Bone changes were not observed in the hindlimbs of
any animal prior to inoculation. At the end of the experiment, no changes were seen in the non-inoculated
(left) hindlimb, nor in pigs of group A (controls). In
the inoculated (right) hindlimb, changes were only
present in animals of group B and in pig 2 from
group C (Fig. 1). These changes consisted of osteolytic
areas in which one or more central foci of increased
opacity were identified. Observations on individual
animals are described below.
Pig B1: Distally in the femur, a lesion extended from
the metaphysis and growth plate centrally in the bone
to the medial aspect of the distal epiphysis (i.e. the medial femoral condyle) (Fig. 2). In the proximal tibia
from the metaphysis to the epiphysis, including the
growth plate, a lesion was present only in the trabecular bone. In the distal part of the third metatarsal
bone a small osteolytic lesion was noted within the
trabecular and cortical bone and an associated soft
tissue swelling was present.
Pig B2: Two lucent areas were seen in the femur.
The lesions were situated medially in the trabecular
bone of the metaphysis with extension through the
growth plate and just into the epiphysis. In one of
the lesions further medial extension into the cortical
bone of the metaphysis was seen.
Pig B3: Distally in the femur, a lesion was observed
at the medial aspect of the metaphysis in close approximation to the growth plate. A second lesion was present at the level of the lateral femoral condyle in the
trabecular bone and the dorsal aspect of the cortical
bone. Two tibial lesions were observed in the medial
and lateral aspect of the proximal metaphysis and
the growth plate and epiphysis, respectively. One of
these lesions showed further medial extension into
Fig. 1. Schematic representation of all lesions observed by CT
scanning in the femur and tibia of pigs B1, B2, B3 and
C2. In general, the largest part of a lesion was located in
the metaphyseal area, with a smaller proportion involving
the growth plate and epiphysis. F, femur and T, tibia.
and through the cortical bone of both the metaphysis
and epiphysis. In the distal part of calcaneus and in
the fourth and fifth metatarsal bones, small osteolytic
lesions with associated soft tissue swellings were
observed.
Pig C2: In the distal femur a trabecular osteolytic
area was observed in the plantar aspect of the metaphysis close to the growth plate. A comparable lesion
was recognized proximally in the tibia.
Gross Pathology
Apart from bone changes observed by CT, other
lesions were not detected grossly. The scoring of
lesions in the right femoral, tibial, tarsal and metatarsal bones is presented in Fig. 3. All bones of the left
hindlimbs (non-inoculated leg) were scored 0. In
group B the inflammation sometimes penetrated
into and through the cortical bone, leading to the formation of soft tissue abscesses (Fig. 4A). Femoral abscesses of animals B1 and B3 extended through the
articular cartilage, leading to secondary fibrinous
and purulent arthritis in the stifle joints. Soft pink
Fig. 2. CT scanning of the femoral lesion (right inoculated limb) in animal B1. (A) The different planes in which the lesions were visualized.
Red, sagittal plane; green, transverse plane. (B) Sagittal plane. An osteolytic area extends from the metaphysis and growth plate
centrally in the bone to the medial aspect of the distal epiphysis (i.e. the medial femoral condyle). In the metaphysis and growth
plate the lesion was irregular and the normal contour of the medial cortex cannot be seen. In the medial femoral condyle, both trabecular and cortical bone are involved. (C) Transverse plan, 10 mm sections. Multiple central foci of increased opacity are seen
throughout the lesion. No lesions are present in the non-inoculated left limb.
Score 2
Score 1
Score 0
A1
A2
A1
A2
A1
A2
A1
A2
F
Ti
T
M
Group A
Control (saline)
n=2
B1
B3
B1
B3
B2
B2
F
Ti
B3
B1
B3
C2
C2
B1
B2
B2
C1
C3
C1
C3
C1
C2
C3
C1
C2
C3
D1
D2
D3
D1
D2
D3
D1
D2
D3
D1
D2
D3
T
M
F
Ti
T
M
F
Ti
T
M
Group B
S54F9 (porcine)
n=3
Group C
NCTC8325-4
(human)
n=3
Group D
UAMS-1(human)
n=3
Fig. 3. Gross bone lesion score in the right hindlimb of pigs inoculated with saline (group A) or three different strains of S. aureus (groups B,
C and D) into the right femoral artery. Bone lesions of: F, femur; Ti, tibia; T, tarsus; M, metatarsus. Scores: 0, no lesion; 1, abscessation; 2, abscessation with penetration through cortical bone. In cases where two lesions were seen in the same bone, only the highest score is included in the figure.
Porcine Model of Osteomyelitis
349
Fig. 4. Right metatarsal bone of animal B3 infected with S. aureus strain S54F9. (A) Sagittally sectioned metaphyseal abscess (MA). The
inflammation penetrated through the distal growth plate (DGP) and the cortical bone (CB); the latter resulted in formation of a soft
tissue abscess (SA). B, C, area from which Fig. 4B and 4C are taken. E, area from which Fig. 4E is taken. (B) The content of the
abscess is mainly of neutrophils (arrows). HE. Bar, 60 mm. (C) Necrotic and almost broken bone trabeculae (B) with empty osteocyte lacuna (arrow) surrounded by colonies of bacteria (arrowhead). HE. Bar, 100 mm. The inset shows a bone resorbing osteoclast. HE. Bar, 80 mm. (D) The lesion is surrounded by granulation tissue (GT) and the bordered, vital bone trabeculum is lined by
osteoblasts (arrow). HE. Bar, 120 mm. (E) The distal growth plate (DGP) is fractured due to penetration of the inflammation resulting in fragmentation of primary spongiosa (PS). HE. Bar, 240 mm. (F) Higher magnification of Fig. 4E. Bacteria (arrow) are
located in the capillary loops of the fragmented primary spongiosa (PS). HE. Bar, 120 mm.
granulation tissue surrounded the purulent material
of all lesions; however, this was most pronounced in
animals B2 and C2. In the femoral abscesses of animal
B2, granulation tissue and surrounding dense fibrotic
tissue was present. Gross lesions were not observed in
the spleen, liver, lungs or kidneys.
Histopathology
One microabscess in the distal metaphyseal area of
the fifth left metatarsal bone in animal B1 was de-
tected microscopically in addition to the lesions identified grossly. All bone abscesses contained central
neutrophils surrounded by granulation tissue. Viable
neutrophils (Fig. 4B) were intermingled with necrotic
inflammatory and tissue cells collectively forming
amorphous granular debris. Multiple fragmented
and necrotic bone trabeculae, with empty lacunae,
were often seen in the centre of abscesses (Fig. 4C).
Sometimes, the necrotic trabeculae were surrounded
by fibrin and numerous activated osteoclasts
(Fig. 4C), while bone trabeculae outside the
350
L.K. Johansen et al.
granulation tissue were bordered by a layer of osteoblasts (Fig. 4D). Proliferation of fibroblasts and dense
fibrosis made up the surrounding granulation tissue
(Fig. 4D). The layer of granulation tissue associated
with the femoral and tibial lesions of animals B2
and C2 was more prominent and thicker when compared with all other abscesses in group B animals.
Fragments of dead primary spongiosa were also present among the neutrophils due to penetration of the
growth plate by the abscesses (Figs. 4A, E and F).
At sites of inflammatory extension through cortical
bone and articular cartilage, the normal histological
architecture of bone and cartilage was completely
destroyed and replaced by neutrophils and necrotic
debris. All bacteria in HE-stained sections (Fig. 4F)
were identified by IHC as S. aureus (Fig. 5A). In the
bone abscesses, bacteria were observed both as single
organisms and in colonies, within and around the
capillary loops at the point of endochondral ossification (Figs. 4F, 5A).Colonies of S. aureus were also detected centrally in bone and soft tissue abscesses
(Fig. 4C). The spleen, liver and kidneys of all animals
were free of microscopical lesions, with the exception
of pig B1 in which a secondary, acute microabscess
(a small focus of neutrophils) was seen in the lung.
Microbiology
S. aureus was re-isolated from all bone abscesses. Bacteriological examination of lung tissue revealed that
only pig B1 had secondary embolic seeding of S. aureus
(200 CFU/g lung tissue). S. aureus bacteria were detected only in bone tissue drilled out from animal
B1 (900 CFU/g bone tissue).
Peptide Nucleic Acid Fluorescence In situ Hybridization
PNA FISH-positive S. aureus was identified in the femoral abscesses of all group B animals and pig C2. The
bacteria were present as aggregates of small looselypacked cocci. The aggregates were embedded in an
opaque matrix suggesting that the bacteria had
formed a biofilm (Fig. 5B).
Discussion
The virulence of three S. aureus strains of porcine and
human origin was examined in a porcine model of
HO. Bone lesions were present in three, one and
none of the recipients of strains S54F9 (porcine),
NCTC-8325-4 and UAMS-1 (human), respectively.
An aggressive disease process with severe bone destruction and secondary soft tissue abscessation is
sometimes described in children with HO (Lew and
Waldvogel, 1997). Comparable lesions were induced
by the porcine strain (S54F9), but milder lesions
mimicking subacute ‘Brodie abscesses’ (Steer and
Carapetis, 2004) developed in pigs B2 and C2. Therefore, all animals with clinical signs of lameness and/or
bone lesions developed HO in a similar pattern as described for human patients. One of the pigs (B1) that
received the porcine strain showed acute secondary
embolic seeding of S. aureus to the lungs and in the fifth
left metatarsal bone. Sepsis following HO occurs
rarely in paediatric cases (Nade, 1983).
CT scanning provides a superior non-invasive tool
for studying bone abscesses in three planes, giving information on their exact anatomical location and size.
In this pig model, the majority of bone abscesses were
in the metaphyseal area, which corresponds to the
most common location in children with HO (Lew
and Waldvogel, 1997; Steer and Carapetis, 2004).
This characteristic metaphyseal localization is
related to the anatomical arrangement of the
metaphyseal vessels beneath the growth plate
(Trueta, 1959). Inside the osteomyelitis lesions there
were colonies of bacteria within the metaphyseal
Fig. 5. Lesions caused by S. aureus strain S54F9. (A) Identification of S. aureus bacteria (arrow) in the abscess shown in Fig. 4F. IHC. Bar,
80 mm. (B) Within the femoral abscess of pig B2 an aggregate of S. aureus bacteria (arrow) is embedded in an opaque matrix. PNA
FISH. Bar, 10 mm.
Porcine Model of Osteomyelitis
capillary loops, so the present porcine model supports
the theory that the pathogenesis of abscess formation
in bones involves haematogenous spread. CT scanning
has been reported as optimal for detection of small sequestra of devitalized bone tissue (Hernandez, 1985),
which are reported to be present in paediatric osteomyelitis lesions after 10 days of infection (Steer and
Carapetis, 2004). In the present study, devitalized
bone tissue was seen as foci of increased opacity on
the CT scans. In children, damage to the growth plate
due to ischaemia or direct chondrolysis is also seen and
associated with growth disturbance related to the volume of tissue destroyed and the location of the destruction (Nade, 1983). Some of the lesions in group B
animals had an even more aggressive profile and penetrated through the growth plate into the epiphysis
and further through the articular cartilage into the
joint. Similar secondary arthritis is also reported in
children (Nade, 1983; Steer and Carapetis, 2004).
The disease process of children with HO therefore
shows similarities with the pathogenesis and
pathology of the lesions in the pigs of the present study.
The human strains UAMS-1 and a derivative of
strain NCTC-8325-4 have been used previously in
murine models of HO (Table 1). In these studies, animals received injections of 108 CFU into the tail vein,
and while the mice showed no clinical signs of disease,
microscopical signs of osteomyelitis were seen in all
mice at 14 dpi regardless of the strain (Blevins et al.,
2003). However, in that study animals were not examined for primary systemic side-effects. In the present model, pigs inoculated with UAMS-1 did not
develop osteomyelitis and only one pig of three inoculated with NCTC-8325-4 had osteomyelitis. In comparison to the murine model we used a much lower
inoculation dose (104 CFU/kg BW). This inoculation
dose was based on a previous porcine dose-response
study using the porcine strain (Johansen et al.,
2011). Presumably, host specificity of the strains is important for the outcome of the infection. Therefore,
the inoculation dose of the human strains may need
to be increased in order to observe the same lesions
as for the porcine strain. In all previous studies, the intravenously inoculated models received bacterial
doses above 106 CFU to obtain a high frequency of osteomyelitis when inoculated with human strains.
In addition to host specificity, the pathogenesis of
S. aureus osteomyelitis may also be associated with
a variety of bacterial virulence factors including
both structural and secreted products (Wright and
Nair, 2010). However, the diversity of pathogenicity
between the different S. aureus strains may arise from
a diverse array of these virulence factors, which also
may contribute to the different outcomes of the present study. The nature of the inoculated S. aureus strain
351
and the species and individual resistance of the host
determine the occurrence and progression of experimentally induced osteomyelitis, and it should be considered that animal models may sometimes be
unreliable for prediction of bacterial virulence in infections because of host specificity. The strain and
host should therefore be considered when comparing
the different animal models of HO and in future studies based on the porcine model.
One of the main virulence factors associated with
development of staphylococcal osteomyelitis, and in
particular the development of chronic disease, is formation of a bacterial biofilm (Gristina et al., 1985;
Brady et al., 2008). In cases where the bacteria
succeed in colonization followed by biofilm
formation within the host, the infection often turns
out to be untreatable and becomes chronic (Brady
et al., 2008; Hoiby et al., 2010). Numerous in vitro
studies of biofilms have characterized a series of
events that lead to biofilm formation (Gristina et al.,
1985; Otto, 2008). These events include growth and
proliferation of bacterial aggregates (microcolonies)
embedded in a self-made, protective matrix of exopolymers (Gristina et al., 1985; Otto, 2008).
In the present study, bacterial aggregates were seen
in situ within the devitalized bone tissue by PNA FISH
after infection with strains S54F9 and NCTC-8325-4.
The aggregation of bacteria was similar to previous
observations in chronic biofilm-based infections such
as cystic fibrosis (Bjarnsholt et al., 2009a), chronic
wounds (Bjarnsholt et al., 2008) and infections related
to permanent tissue fillers (Bjarnsholt et al., 2009b).
Therefore, and in accordance with the presence of
granulation tissue, the osteomyelitis lesions of the
pigs developed into a pathological and microbiological chronic stage shortly after infection. This observation may explain why antibiotic therapy sometimes
fails in patients diagnosed at an early stage of disease
(Nade, 1983; Lew and Waldvogel, 1997; Hoiby et al.,
2010). The resistance to antibiotics induced by
biofilm formation is mediated through a very low
metabolism and down-regulation of cell division by
the entrenched bacteria (Brady et al., 2008). Furthermore, the biofilm acts as a diffusion barrier to slow
down the infiltration of some antimicrobial agents
(Brady et al., 2008). Together with the resistance
raised by the biofilm, other features of S. aureus are
connected with failure of antibiotic treatment. These
include the intracellular location and presence of mutant forms such as small colony variants, both representing a further complication in the management
of staphylococcal osteomyelitis (Wright and Nair,
2010). S. aureus, which is the most common aetiological agent of osteomyelitis, is a well known potent
biofilm-forming bacterium, leading to the perception
L.K. Johansen et al.
352
that biofilm formation is a critical step in the pathogenesis of osteomyelitis (Brady et al., 2008).The results
of the present study support this theory of
co-occurrence between biofilm formation and chronic
osteomyelitis.
Acknowledgements
We thank B. Andersen and L. Kioerboe for excellent
laboratory assistance. This work was financed by
grant number 271-07-0417 from the Danish Medical
Research Council, the Orthopedic Surgery Foundation in Aarhus, the Fund of King Christian X and
the Beckett Foundation. No conflicts of interest are
declared by the authors.
References
Alderson M, Speers D, Emslie K, Nade S (1986) Acute haematogenous osteomyelitis and septic arthritisea single
disease. An hypothesis based upon the presence of transphyseal blood vessels. Journal of Bone and Joint Surgery
(Britain), 68, 268e274.
An YH, Kang QK, Arciola CR (2004) Animal models of
osteomyelitis. International Journal of Artificial Organs,
29, 407e420.
Bjarnsholt T, Jensen PO, Fiandaca MJ, Pedersen J,
Hansen CR et al. (2009a) Pseudomonas aeruginosa biofilms
in the respiratory tract of cystic fibrosis patients. Pediatric
Pulmonology, 44, 547e558.
Bjarnsholt T, Kirketerp-Moller K, Jensen PO,
Madsen KG, Phipps R et al. (2008) Why chronic
wounds will not heal. A novel hypothesis. Wound Repair
and Regeneration, 16, 2e10.
Bjarnsholt T, Tolker-Nielsen T, Givskov M, Janssen M,
Christensen LH (2009b) Detection of bacteria by fluorescence in-situ hybridization in culture-negative soft
tissue filler lesions. Dermatological Surgery, 35,
1620e1624.
Blevins JS, Elasri MO, Allmendinger SD, Beenken KE,
Skinner RA et al. (2003) Role of sarA in the pathogenesis
of Staphylococcus aureus musculoskeletal infection. Infection
and Immunity, 71, 516e523.
Brady RA, Leid JG, Calhoun JH, Costerton JW,
Shirtliff ME (2008) Osteomyelitis and the role of biofilms in chronic infection. FEMS Immunology and Medical
Microbiology, 52, 13e22.
Cassat JE, Dunman PM, McAleese F, Murphy E,
Projan SJ et al. (2005) Comparative genomics of Staphylococcus aureus musculoskeletal isolates. Journal of Bacteriology, 187, 576e592.
Chadha HS, Fitzgerald RH Jr., Wiater P, Sud S, Nasser S
et al. (1999) Experimental acute hematogenous osteomyelitis in mice. I. Histopathological and immunological findings. Journal of Orthopaedic Research, 17, 376e381.
Costerton JW, Stewart PS, Greenberg EP (1999) Bacterial
biofilms: a common cause of persistent infections. Science,
284, 1318e1322.
Daum RS, Davis WH, Farris KB, Campeau RJ,
Mulvihill DM et al. (1990) A model of Staphylococcus aureus bacteremia, septic arthritis, and osteomyelitis in
chickens. Journal of Orthopaedic Research, 8, 804e813.
Deysine M, Rosario E, Isenberg HD (1976) Acute hematogenous osteomyelitis: an experimental model. Surgery,
79, 97e99.
Elasri MO, Thomas JR, Skinner RA, Blevins JS,
Beenken KE et al. (2002) Staphylococcus aureus collagen
adhesin contributes to the pathogenesis of osteomyelitis.
Bone, 30, 275e280.
Emslie KR, Nade S (1983) Acute hematogenous staphylococcal osteomyelitis. A description of the natural history
in an avian model. American Journal of Pathology, 3,
333e345.
Gristina AG, Oga Masayoshi, Webb LX, Hobgood CD
(1985) Bacterial colonization in the pathogenesis of osteomyelitis. Science, 228, 990e993.
Harris DL, Alexandra TJL (1999) Methods of disease control. In: Diseases of Swine, 8th Edit., BE Straw,
JJ Zimmerman, SD’ Allaire, DJ Taylor, Eds., Iowa
State University Press, Ames, pp. 1077e1110.
Hernandez RJ (1985) Visualization of small sequestra by
computerized tomography. Report of 6 cases. Pediatric
Radiology, 15, 238e241.
Hienz SA, Sakamoto H, Flock JI, Morner AC, Reinholt FP
et al. (1995) Development and characterization of a new
model of hematogenous osteomyelitis in the rat. Journal
of Infectious Disease, 171, 1230e1236.
Hoiby N, Bjarnsholt T, Givskov M, Molin S, Ciofu O
(2010) Antibiotic resistance of bacterial biofilms. International Journal of Antimicrobial Agents, 35, 322e332.
Holland C (1972) Experiments on ultra-sound osteomyelitis.
Zeitschrift fur Orthopadie und ihre Grenzgebiete, 110, 145e159.
Jensen HE, Nielsen OL, Agerholm JS, Iburg T,
Johansen LK et al. (2010) A non-traumatic Staphylococcus
aureus osteomyelitis model in pigs. In Vivo, 24, 257e264.
Johansen LK, Frees D, Aalbaek B, Koch J, Iburg T et al.
(2011) A porcine model of acute, haematogenous, localized osteomyelitis due to Staphylococcus aureus: a pathomorphological study. Acta Pathologica, Microbiologica et
Immunologica Scandinavica, 119, 111e118.
Kadyrov MA, Muratova K, Shakirov DS (1966) On obtaining a model for osteomyelitis. Eksperimental’Naia
Khirurgiia i Anesteziologiia, 11, 32e33.
Koiunderliev P (1971) Experimental studies on the etiopathogenesis of hematogenous osteomyelitis. Khirurgiia
(Sofiia), 24, 172e177.
Leifsson PS, Iburg T, Jensen HE, Agerhom JS, KjeldgaardHansen M et al. (2010) Intravneous inoculation of
Staphylococcus aureus in pigs induces severe sepsis as
indicated by increased hypercoagulability and hepatic
dysfunction. FEMS Microbiology Letters, 309, 208e216.
Lew DP, Waldvogel FA (1997) Osteomyelitis. New England
Journal of Medicine, 336, 999e1007.
Lexer E (1894) Zur experimentellen erzeugung osteomyelitischer herde. Archiv f€ur klinische Chirurgie, 48, 181e200.
Matsushita K, Hamabe M, Matsuoka M, Aoki H,
Miyoshi K et al. (1997) Experimental hematogenous
Porcine Model of Osteomyelitis
osteomyelitis by Staphylococcus aureus. Clincal Orthopaedics
and Related Research, 334, 291e297.
Morrissy RT, Haynes DW (1989) Acute hematogenous
osteomyelitis: a model with trauma as an etiology.
Journal of Pediatric Orthopaedics, 9, 447e456.
Nade S (1983) Acute haematogenous osteomyelitis in
infancy and childhood. Journal of Bone and Joint Surgery
(Britain), 65, 109e119.
Otto M (2008) Staphylococcal biofilms. Current Topics in
Microbiology and Immunology, 322, 207e228.
Rissing JP, Buxton TB, Fisher J, Harris R, Shockley RK
(1985) Aracidonic acid facilitates experimental chronic
osteomyelitis in rats. Infection and Immunity, 49, 141e144.
Rodet A (1973) The classic. An experimental study on infectious osteomyelitis. Clincal Orthopaedics and Related Research, 96, 3e4.
Scheman L, Jonta M, Lewin P (1941) The production of
experimental osteomyelitis: a preliminary report. Journal of the American Medical Association, 117, 1525e1529.
Speers DJ, Nade SM (1985) Ultrastructural studies of adherence of Staphylococcus aureus in experimental acute hematogenous osteomyelitis. Infection and Immunity, 49, 443e446.
Starr CL (2002) Acute hematogenous osteomyelitis.
Clarence L. Starr, M.D. (1868e1928). Clinical Orthopaedics and Related Research, 403, 4e7.
Steer AC, Carapetis JR (2004) Acute hematogenous osteomyelitis in children: recognition and management. Paediatric Drugs, 6, 333e346.
Sung JM, Lloyd DH, Lindsay JA (2008) Staphylococcus aureus host specificity: comparative genomics of human
353
versus animal isolates by multi-strain microarray.
Microbiology, 154, 1949e1959.
Thompson RH, Dubos RJ (1938) Production of experimental osteomyelitis in rabbits by intravenous injection
of Staphylococcus aureus. Journal of Experimental Medicine,
68, 191e206.
Trueta J (1959) The 3 types of acute haematogenous osteomyelitis e a clinical and vascular study. Journal of Bone
and Joint Surgery (Britain), 41, 671e680.
Weaver JB, Tayler MW (1943) Experimental staphylococcaemia and hematogenous osteomyelitis. Journal of Bone
and Joint Surgery (America), 25, 791e802.
Whalen JL, Fitzgerald RH Jr., Morrissy RT (1988) A histological study of acute hematogenous osteomyelitis following physeal injuries in rabbits. Journal of Bone and
Joint Surgery (America), 70, 1383e1392.
Wright JA, Nair SP (2010) Interaction of staphylococci
with bone. International Journal of Medical Microbiology,
300, 193e204.
Yoon KS, Fitzgerald RH Jr., Sud S, Song Z, Wooley PH
(1999) Experimental acute hematogenous osteomyelitis
in mice. II. Influence of Staphylococcus aureus infection on
T-cell immunity. Journal of Orthopaedic Research, 17,
382e391.
October 25th, 2011
½ Received,
Accepted, January 21st, 2012