The rate of allograft use in orthopaedics is on the
rise1. While
allograft bone and tendon is generally considered safe, there is a potential
risk of disease transmission whenever tissue is transplanted from one
individual to
another2-6.
Rigorous donor screening, although not 100% effective, is currently thought to
be the best available method to improve allograft
safety1,6.
Despite these efforts, viral and bacterial transmission has been documented
following transplantation of connective tissue allografts from infected
donors2-6.
This has led to the development of a number of sterilization procedures
designed to eliminate possible viral and/or bacterial
transmission1,6,7-10.
While such procedures are effective for sterilizing connective tissue
allografts in which cell viability is neither desired nor needed for
successful host incorporation and remodeling (such as bone and tendon), this
is not the case for articular cartilage allografts. An experimental study has
suggested that the functional outcome of articular cartilage transplants is
directly related to the number of viable donor cells present at the time of
transplantation11.
Therefore, the tissue sterilization techniques that have been advocated to
ensure the sterility and safety of bone and tendon allografts may not be
appropriate for articular cartilage allografts.
While the transmission of infectious retrovirus through the transplantation
of bone and tendon allografts has been documented clinically and
experimentally2-4,12-14,
there is some debate regarding the ability of cartilage to harbor and transmit
infective retrovirus. In one study, the investigators were unable to
demonstrate human immunodeficiency virus (HIV)-1 DNA in cartilage from
HIV-positive
patients15. This
observation led the authors to conclude that HIV is not present in the
cartilage of HIV-infected patients, making HIV transmission through cartilage
allografts improbable. However, another, similar study demonstrated HIV DNA in
postmortem cartilage samples from nine of ten patients who were
HIV-antibody-positive16.
To add to this confusion, in vitro studies have produced conflicting results
regarding the ability of the HIV retrovirus to infect human
chondrocytes17,18.
Our laboratory has developed and validated in vivo and in vitro models to
test the infectivity of musculoskeletal allografts systemically infected with
feline leukemia virus
(FeLV)14. Feline
leukemia virus, a retrovirus with a structure and replication cycle similar to
that of HIV, is widely used as an animal model for HIV. Previous experimental
studies from our laboratory have demonstrated that a variety of
musculoskeletal tissues, including ligament, bone, and menisci, from
systemically FeLV-infected cats transmit infectious retrovirus following
allotransplantation12.
The purpose of the current study was to determine, with use of our in vitro
test system, the ability of articular cartilage fragments and isolated
articular cartilage chondrocytes from donors systemically infected with FeLV
to transmit retrovirus. We hypothesized that intact articular cartilage
fragments, but not isolated articular cartilage chondrocytes, from cats
systemically infected with FeLV are capable of transmitting the infectious
retrovirus.
Graft Donors
Following institutional approval of the animal care and use protocols, five
ten-week-old, specific pathogen-free cats were infected with the Rickard
strain of FELV with use of a previously described
method14. The
animals were monitored weekly from two to eight weeks after inoculation for
the presence of the FeLV p27 antigen in their plasma with use of an
enzyme-linked immunosorbent assay
(ELISA)12. All five
animals tested positive for systemic infection with FeLV by three weeks and
remained positive through eight weeks. At that time, the animals were humanely
killed and musculoskeletal tissues (bones, tendons, ligaments, menisci, and
articular cartilage) were harvested under sterile conditions for the current
and future studies.
Articular Cartilage Fragments and Isolated Chondrocytes
Immediately after the animals were killed, fresh articular cartilage
segments were harvested from each FeLV-positive animal under sterile
conditions with use of a ring curet. The segments were placed in culture media
at room temperature. Care was taken to limit cartilage harvest to the level of
calcified cartilage, avoiding subchondral bone and noncartilaginous tissues.
This was determined by visual inspection. The harvested cartilage segments
from each animal were then divided into three groups, which were treated as
follows: (1) the segments were minced into approximately 1-mm3
pieces and were cocultured with a felinev embryonic fibroblast cell line, (2)
they were processed to isolate articular cartilage chondrocytes, or (3) they
were fixed in 4% paraformaldehyde for immunohistochemistry studies. Cortical
bone segments were also placed in culture media at room temperature and were
minced into pieces approximately 1 mm3 in size immediately prior to
introduction into feline embryonic fibroblast cell cultures to serve as a
positive control for each donor animal.
Chondrocyte Isolation
Fresh articular cartilage fragments from each cat were minced into small
(approximately 1-mm3) pieces under sterile conditions and were
placed into individual culture wells containing 5 mL of
collagenase-phosphate-buffered saline solution (0.5 mg/mL of clostridial
collagenase [Sigma, St. Louis, Missouri]). The specimens were gently rocked
overnight at 37°C and 10% CO2. The collagenase solution
containing chondrocytes from each cat was centrifuged at 1500 rpm for ten
minutes, and the supernatant was discarded. The cells were then washed three
times in chondrocyte media containing Dulbecco minimal essential medium
(DMEM)-F12, 2 µg/mL of gentamicin, 2% glutamine, 15% fetal bovine serum,
100 units/mL of pencillin G, 100 µg/mL of streptomycin, and 0.25 µg/mL
of amphotericin. The cells were then resuspended in the above media and were
plated into individual flasks. The media were changed every two to three days,
and the cells were expanded to passage 2, at which time confluent cells were
trypsinized and frozen into aliquots of 2.0 × 106 cells/mL in
10% dimethyl sulfoxide (DMSO)/DMEM-F12.
Cell Cultures
Five flasks of feline embryonic fibroblast cells per group (articular
cartilage, isolated chondrocytes, or cortical bone) were cultured in 10 mL of
feline embryonic fibroblast medium containing DMEM, 15% fetal bovine serum, 2%
glutamine, 10 µg/mL of Gentocin (gentamicin sulfate), and 10 µg/mL of
enrofloxacin. Cultures of feline embryonic fibroblast cells were grown to
confluence and divided at a dilution of 1:3 eighteen hours before challenge.
Diethylaminoethyl (DEAE) dextran (0.03 mg/mL) was added to the media for
thirty minutes before challenge to enhance susceptibility of the feline
embryonic fibroblast cells to FeLV infection. The DEAE dextran medium was
replaced with feline embryonic fibroblast medium just prior to challenge with
one of the following three experimental conditions: (1) fresh articular
cartilage fragments (approximately 1 cm3) systemically infected
with FeLV, (2) isolated chondrocytes (1:1 ratio with feline embryonic
fibroblast cells) from fresh articular cartilage systemically infected with
FeLV, and (3) fresh bone fragments (approximately 1 cm3)
systemically infected with FeLV to serve as a positive control. Chondrocytes
and feline embryonic fibroblast cells were cocultured in a 50:50 mixture of
feline embryonic fibroblast medium and chondrocyte medium. The five replicates
in each of the three coculture groups were composed of individual samples from
each of the five donor animals. Media and cells from feline embryonic
fibroblast stock cultures were confirmed negative for FeLV antigen and
provirus, respectively. Cocultured feline embryonic fibroblast cells were
trypsinized and introduced into new flasks at a 1:10 dilution on reaching
confluence, approximately every three to six days, for a total of four
passages to facilitate exponential cellular and viral replication. At each
passage, media were collected to test for FeLV antigen. In addition, DNA was
extracted from feline embryonic fibroblast cells at passages 3 and 4 and was
amplified by real-time quantitative polymerase chain reaction to test for FeLV
provirus. Bone fragments from the systemically FeLV-infected donor cat served
as a positive control for each animal because we had documented that cortical
bone from FeLV-infected cats always transmits the retrovirus in vitro and in
vivo1,12-14.
Positive and Negative Control Cultures
Five replicate flasks were assigned to the positive control group. After
thirty minutes of DEAE dextran treatment, the DEAE dextran medium was removed
and 100 µL of filtered supernatant fluid from a FeLV-infected cell line
along with 10 mL of feline embryonic fibroblast medium was added to each
flask. The five negative control replicates received DEAE dextran treatment
with feline embryonic fibroblast medium replacement only.
Quantification of FeLV p27 Antigen
A commercial kit (Feline Leukemia Virus Antigen Test Kit, ViraCHEK/FeLV
Laboratory Pack; Synbiotics, San Diego, California) was used to detect FeLV
p27 antigen in culture supernatant fluids. The optical density (OD) was
measured at 630 nm with use of a spectrophotometer (Bio-Tek Instruments,
Winooski, Vermont), and a standardized OD ratio (SODR) to correct for
interassay variations was calculated with use of the following formula: SODR =
(OD of sample — OD of negative control)/(OD of positive control —
OD of negative control). A SODR value of =0.1 that increased over time was
considered positive.
Quantification of FeLV Proviral DNA
Proviral FeLV DNA from the feline embryonic fibroblast cell cocultures was
amplified and quantified after passages 3 and 4, with use of a real-time
quantitative polymerase chain reaction assay. Specifically, a 65-base-pair
fragment of exogenous FeLV was amplified by real-time quantitative polymerase
chain reaction (7500 Fast Real-Time PCR System; Applied Biosystems, Foster
City, California) under the absolute quantification mode from 100 ng of EcoRI
linearized chromosomal DNA extracted from feline embryonic fibroblast cells
cocultured with tissue or cells (articular cartilage segments, bone fragments,
or isolated chondrocytes). Primers and probes were unique to the U3 region of
exogenous FeLV. Serial tenfold calibration standard dilutions and negative
controls were run in parallel with test samples. Additionally, assays for each
sample were run in triplicate with the mean reported as copy number per 100 ng
of DNA. Copy number was determined by comparison with standard dilutions.
Results of >75 copies/100 ng DNA were considered positive. In addition,
FeLV proviral nucleic acids were quantified by quantitative polymerase chain
reaction in DNA extracted from isolated chondrocytes.
Statistical Analysis
Feline leukemia virus p27 antigen SODR and FeLV proviral copy number were
assigned either a positive or negative value (according to the criteria given
above) and were compared with use of a two-tailed Fisher exact test.
Differences between treatment groups were considered significant at p <
0.05.
Immunohistochemistry
Paraffin-embedded sections of articular cartilage fragments, as well as
fixed preparations of isolated chondrocytes from FeLV-infected animals, were
stained for the FeLV p27 antigen (ViroStat, Portland, Maine). Cartilage
sections and isolated chondrocytes from specific pathogen-free cats that were
not infected with FeLV were used as negative controls. A commercial
preparation containing both FeLV-positive and negative feline kidney cells
(CrFK; VMRD, Pullman, Washington) also served as both a positive and negative
control.
Tissue samples were fixed in 4% paraformaldehyde and embedded in paraffin.
Sections were deparaffinized, dehydrated in alcohol, and antigenic sites were
unmasked through digestion with 0.03% hyaluronidase (Sigma) for two hours at
37°C, 1 mg/mL of collagenase (Sigma) for thirty minutes at 37°C, and
incubation with citrate buffer (Antigen Unmasking Solution; Vector
Laboratories, Burlingame, California) for one hour at 65°C. Nonspecific
protein-binding was blocked with horse serum followed by Avidin/Biotin
blocking solutions (Vector). Slides were incubated overnight with goat
anti-FeLV p27 antibody (dilution 1:100) at 4°C. After the sections were
rinsed, they were incubated with a biotinylated rabbit anti-goat antibody
(1:400; Vector) followed by Texas Red Avidin DCS (1:100; Vector). Nuclei were
counterstained with Vector mount with 4',6-diamidino-2-phenylindole (DAPI;
Vector), a fluorescent stain that binds strongly to DNA. Sections were viewed
with use of a Confocal Laser Scanning microscope (LSM 510 META; Carl Zeiss,
Oberkochen, Germany).
Isolated chondrocytes grown on chamber slides (Nalge Nunc, Rochester, New
York) were fixed in 4% paraformaldehyde, rinsed in distilled water, and
air-dried. Prior to staining, slides were rehydrated in phosphate-buffered
saline solution and permeabilized with 0.2% Triton X-100. After they were
quickly rinsed with phosphate-buffered saline solution, antigen retrieval was
performed with use of citrate buffer (Antigen Unmasking Solution; Vector) at
65°C for forty minutes. Subsequently, the same protocol as for the tissue
samples was followed (see above). Slides were viewed with a Zeiss Axioscope
microscope (Carl Zeiss).
Donor Animals
Plasma from all five donor animals tested positive for the FeLV p27 antigen
with use of an ELISA assay by the third week postinoculation and remained
positive through the remainder of the eight-week period.
Articular Cartilage Fragments
Conditioned media from feline embryonic fibroblast cell and fresh articular
cartilage fragment cocultures from each of the five FeLV-infected cats tested
positive for the FeLV p27 antigen by the third passage
(Fig. 1). This indicates active
viral replication.
In addition, the feline embryonic fibroblast cells that were expanded
following coculture with articular cartilage fragments tested positive for
proviral FeLV DNA by quantitative polymerase chain reaction, indicating
transmission of infective virus from the articular cartilage fragments.
Immunohistochemical staining of articular cartilage fragments from
FeLV-infected cats demonstrated the presence of p27 antigen throughout the
extracellular matrix. Staining was especially intense in the immediate
pericellular matrix region (Fig.
2-A). However, there did not appear to be any p27 antigen within
the chondrocytes. Articular cartilage samples from specific pathogen-free cats
were negative for the FeLV p27 antigen
(Fig. 2-B).
Isolated Articular Cartilage Chondrocytes
Conditioned media from feline embryonic fibroblast cell and isolated
chondrocyte cocultures from each of the five FeLV-infected cats remained
negative for the FeLV p27 antigen throughout all four passages
(Fig. 1). Similarly, proviral
FeLV DNA was not documented in isolated chondrocytes or the expanded feline
embryonic fibroblast cells that had been cocultured with isolated
chondrocytes.
Immunohistochemical staining of isolated chondrocyte preparations from each
of the FeLV-infected animals was negative for FeLV p27 antigen
(Fig. 3-A). Similarly, all
isolated chondrocyte preparations from specific pathogen-free cats were
negative for the FeLV p27 antigen.
The control (FeLV-infected and noninfected kidney cells) revealed strong
intracytoplasmic staining for the FeLV p27 antigen in approximately 30% of the
cells (Fig. 3-B).
Cortical Bone Fragments
Conditioned media from feline embryonic fibroblast cell and cortical bone
fragment cocultures from each of the five FeLV-infected cats were positive for
FeLV p27 antigen at passage 1 and remained positive throughout all four
passages (see Fig. 1).
Likewise, the expanded feline embryonic fibroblast cells that had been
cocultured with bone fragments from each of these five FeLV-infected cats
tested positive for proviral FeLV DNA by quantitative polymerase chain
reaction, indicating transmission of infective virus from the cortical bone
fragments.
Negative and Positive Controls
The five negative control replicates receiving only DEAE dextran treatment
were negative for FeLV p27 antigen throughout all four passages. All positive
control replicates were positive for FeLV p27 antigen at passages 2, 3, and 4.
Negative control replicates were negative at all passages, while positive
control replicates were positive for FeLV provirus tested at passages 3 and
4.
While the potential for retroviral transmission through the
allotransplantation of tendon and bone is well documented, the ability of
articular cartilage to harbor and transmit infectious retrovirus is less
clear-cut. A study that examined postmortem cartilage samples from eight
HIV-infected patients found no proviral HIV-1 DNA in any of the
specimens15. This
led the authors to conclude that because HIV proviral DNA is not present in
the cartilage of HIVinfected patients, the risk of HIV transmission through
cartilage allografts is very
low15. Importantly,
this study did not test for the presence of infectious retrovirus (RNA) within
the extracellular matrix of the articular cartilage. Thus, the ability of
cartilage allografts from HIV-infected individuals to transmit infectious
retrovirus was not rigorously examined.
The results of the current study demonstrate that while FeLV proviral DNA
or p27 antigen could not be identified within isolated chondrocytes from
systemically FeLV-infected animals, intact cartilage fragments from these same
animals contained large amounts of p27 antigen in the extracellular matrix.
While the presence of the FeLV p27 antigen has been associated with the
presence of FeLV viral particles, the antigen itself is not
infectious19,20.
However, increasing quantities of the FeLV p27 antigen detected in our in
vitro test system, produced by means of active viral replication by the feline
embryonic fibroblast cells, is indicative of retroviral transmission from the
intact infectious cartilage fragments.
The presence of FeLV p27 antigen and infectious retrovirus in the
extracellular matrix of the articular cartilage of FeLV-infected cats, but not
within the isolated chondrocytes, may be explained by entry of p27 antigen and
viral particles into the extracellular articular cartilage matrix by means of
external sources such as synovial fluid. While the presence of the FeLV p27
antigen in synovial fluid has never been investigated, the antigen has been
identified in serum, saliva, and tears from FeLV-infected
cats21. Infectious
retrovirus has been isolated from synovial
fluid22 and
arthroscopic effluents of HIV-positive
patients23. A
previous study has suggested that cell-free retroviral infection of articular
cartilage is possible through the diffusion of viral particles into the
extracellular
matrix16. Indeed,
retroviral transmission through acellular mechanisms is well known. For
example, prior to the institution of viral inactivation procedures,
hemophiliac patients were at extremely high risk of acquiring HIV infection
through the administration of clotting factor
proteins24,25.
Therefore, it is possible that infectious retrovirus could be present within
the extracellular matrix in the absence of proviral DNA within the
chondrocytes.
It is also possible that the infectious retrovirus transmitted to feline
embryonic fibroblast cells during coculture with intact articular cartilage
fragments may have come from contamination by foreign (blood or
nonchondrocytic) cells even though these fragments were harvested with extreme
care. In one study, HIV-DNA was detected by polymerase chain reaction in
cadaver cartilage samples from nine of ten HIV-antibody-positive
individuals16. The
reason for this apparent discrepancy with the results of other
investigators15 is
unclear. However, the relatively small number of infected cells identified
within the cartilage specimens (0.001% to 1% of the total cells) may reflect
contamination of the samples with noncartilage cells from blood,
perichondrium, or synovium rather than infected
chondrocytes15,22.
The presence of p27 antigen demonstrated by immunohistochemistry in the
extracellular matrix of the cartilage segments in the current study suggests
that infection through a cell-free pathway may be a more likely mechanism of
viral transmission in our model system.
The basis for the lack of FeLV proviral DNA or p27 antigen within the
chondrocytes of systemically infected cats was not within the scope of the
present study. Possible explanations include lack of a receptor or phagocytic
activity, preventing the retrovirus from entering chondrocytes. A previous
study has suggested that because human chondrocytes lack the CD4 HIV receptor
molecule, normal cartilage cells could not be infected by
HIV17,26.
This concept was challenged later when cultured chondrocytes were successfully
infected with HIV-1 or HIV-2, leading the authors to suggest that viral entry
into chondrocytes could occur by means of a non-CD4 receptor mechanism, such
as binding, fusion, or
uncoating18.
However, because the chondrocytes used in these studies were passaged more
than ten times, the phenotypic changes known to occur in cultured chondrocytes
during serial passages could call into question the relevance of these
observations27. It
is possible that the dedifferentiation of chondrocytes during serial monolayer
cultures results in morphologic and metabolic
changes27 that may
permit viral entry through non-CD4 dependent pathways. Additional research is
needed to determine chondrocyte susceptibility to various viruses.
The results of the current study support the conclusion that articular
cartilage fragments can readily transmit infectious retrovirus even in the
absence of proviral DNA or viral antigen within the articular cartilage
chondrocytes. In addition, because donor-cell viability markedly improves the
long-term functional outcome of articular cartilage allografts (precluding the
use of sterilization procedures currently employed for some tendon and bone
allografts), the results of the current study also underscore the importance
of rigorous donor screening when the use of articular cartilage allografts is
being considered.?