Gene Therapy (2005) 12, 1171–1179
& 2005 Nature Publishing Group All rights reserved 0969-7128/05 $30.00
www.nature.com/gt
RESEARCH ARTICLE
Enhanced repair of articular cartilage defects in vivo
by transplanted chondrocytes overexpressing
insulin-like growth factor I (IGF-I)
H Madry1, G Kaul1, M Cucchiarini1, U Stein2, D Zurakowski3, K Remberger2, MD Menger4, D Kohn1
and SB Trippel5
1
Laboratory for Experimental Orthopaedics, Department of Orthopaedic Surgery, Saarland University, Homburg, Germany;
Department of Pathology, Saarland University, Homburg, Germany; 3Harvard University, Boston, Massachusetts, USA;
4
Department of Surgery, Saarland University, Homburg, Germany; and 5Indiana University, Indianapolis, Indiana, USA
2
Traumatic articular cartilage lesions have a limited capacity
to heal. We tested the hypothesis that overexpression of a
human insulin-like growth factor I (IGF-I) cDNA by transplanted articular chondrocytes enhances the repair of fullthickness (osteochondral) cartilage defects in vivo. Lapine
articular chondrocytes were transfected with expression
plasmid vectors containing the cDNA for the Escherichia
coli lacZ gene or the human IGF-I gene and were
encapsulated in alginate. The expression patterns of the
transgenes in these implants were monitored in vitro for 36
days. Transfected allogeneic chondrocytes in alginate were
transplanted into osteochondral defects in the trochlear
groove of rabbits. At three and 14 weeks, the quality of
articular cartilage repair was evaluated qualitatively and
quantitatively. In vitro, IGF-I secretion by implants constructed from IGF-I-transfected chondrocytes and alginate
was 123.2722.3 ng/107 cells/24 h at day 4 post transfection
and remained elevated at day 36, the longest time point
evaluated. In vivo, transplantation of IGF-I implants improved
articular cartilage repair and accelerated the formation of the
subchondral bone at both time points compared to lacZ
implants. The data indicate that allogeneic chondrocytes,
transfected by a nonviral method and cultured in alginate, are
able to secrete biologically relevant amounts of IGF-I over a
prolonged period of time in vitro. The data further demonstrate that implantation of these composites into deep
articular cartilage defects is sufficient to augment cartilage
defect repair in vivo. These results suggest that therapeutic
growth factor gene delivery using encapsulated and transplanted genetically modified chondrocytes may be applicable
to sites of focal articular cartilage damage.
Gene Therapy (2005) 12, 1171–1179. doi:10.1038/
sj.gt.3302515; published online 7 April 2005
Keywords: cartilage defects; IGF-I; gene transfer; chondrocytes; alginate; cell transplantation
Introduction
Traumatic articular cartilage defects do not heal spontaneously. Efforts to achieve repair of these lesions have
been limited by the challenge of stimulating the resident
cells to form new cartilage. When cell-based therapy is
employed, there arises the additional challenge of
retaining the transplanted cells in the defect.1 The initial
healing response to an osteochondral articular cartilage
defect is mediated, in part, by cell signaling polypeptides
that act on cells derived either from the joint cavity2 or
the bone marrow.3 Such polypeptides influence the rate
of articular cartilage repair.3–6 Insulin-like growth factor-I
(IGF-I) is a 7.6 kDa polypeptide growth factor that
stimulates both matrix synthesis and cell proliferation
Part of this work was presented at the 49th Annual Meeting of the
Orthopaedic Research Society, February 2–5, 2003, New Orleans,
Louisiana, USA
Correspondence: Dr H Madry, Laboratory for Experimental Orthopaedics,
Department of Orthopaedic Surgery, Saarland University Medical Center,
66421 Homburg/Saar, Germany
Received 25 April 2004; accepted 29 January 2005; published online
7 April 2005
of chondrocytes. In particular, it increases chondrocyte
production of proteoglycan and type-II collagen, two
principal constituents of cartilage.7–9 The exogenous
administration of human IGF-I has been reported to
enhance the cell-based repair of articular cartilage
defects.3,6 Delivery of recombinant human IGF-I protein
to articular cartilage has been achieved by intra-articular
injection,10 by supplementing a fibrin clot with IGF-I3
and by combining composites of chondrocytes and
polymerized fibrin with IGF-I.6
The development of IGF-I as a therapeutic agent for
articular cartilage disorders has been restrained by its
short intra-articular residence time and the intrinsic
paucity of articular chondrocytes to serve as target cells.
Genetically engineered chondrocytes could be used to
both secrete a therapeutic growth factor and to supply a
cell population capable of responding to an exogenous
repair stimulus. We previously reported that overexpression of a human IGF-I cDNA promotes new tissue
formation in an ex vivo model of articular chondrocyte
transplantation11 and enhances tissue engineering of
cartilage.12 When luciferase-transfected chondrocytes
embedded in alginate were implanted into articular
cartilage defects in vivo, reporter gene expression
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1172
persisted in articular defects for at least 32 days.13 It
remains unknown if chondrocyte transfection with a
potentially therapeutic gene such as IGF-I leads to
secretion of the recombinant protein following cell
encapsulation in alginate in vitro and whether such
overexpression modifies articular cartilage repair when
transplanted into sites of articular cartilage damage
in vivo. In this study, we tested the hypothesis that
overexpression of human IGF-I by articular chondrocytes
transplanted in a hydrogel into osteochondral cartilage
defects enhances cartilage repair in vivo.
Results
IGF-I production by transfected chondrocytes
encapsulated in alginate in vitro
Chondrocyte-alginate implants were characterized and
the time course of recombinant IGF-I production was
assessed in vitro. Lapine articular chondrocytes were
transfected in monolayer culture with expression plasmid vectors carrying either the Escherichia coli lacZ gene
or a human IGF-I cDNA using the nonliposomal lipid
formulation FuGENE 6. At 1 day after transfection, the
modified chondrocytes were encapsulated in the hydrogel, alginate. At the time of encapsulation, implants
constructed from lacZ- or IGF-I transfected chondrocytes
and alginate (termed lacZ or IGF-I implants) had a mean
diameter of 3.270.3 and 3.270.5 mm, respectively, and a
cell viability of 84.4–90.7%. Each implant contained from
3.570.1 104 to 5.070.2 104 viable cells (Table 1, Figure
1a and b). The total number of chondrocytes per implant
in the lacZ-transfected group declined 14% over the 36
days of in vitro cultivation. In contrast, the total number
of chondrocytes in IGF-I implants increased 24% over
this time period (Table 1, Figure 1c and d).
Mean transfection efficiency as determined by X-gal
staining of lacZ-transfected chondrocytes 2 days following transfection was 35.777.9%. At all time points
tested, conditioned medium from lacZ implants contained p10.876.7 ng IGF-I per 1 107 viable cells/24 h.
In comparison, IGF-I implants produced up to 123.27
Table 1
Figure 1 Representative histologic sections of lacZ implants (left) and
IGF-I implants (right) stained with hematoxylin and eosin (a–d) or a
polyclonal anti-caspase-3 IgG (e, f). The chondrocyte-alginate implants
were cultured in vitro for 2 (a, b) or 36 days (c, d) in serum-free medium.
Over the time of in vitro cultivation, cellularity was preserved and matrix
staining increased. (e, f) Histologic sections of a lacZ implant (e) and an
IGF-I implant (f) transplanted into an osteochondral defect after 3 weeks in
vivo. Cells that underwent apoptosis are distinguishable by their brown
color. Photomicrographs were obtained using standardized photographic
parameters, including light intensity. Original magnifications 200
(a–d) and 250 (e, f).
22.3 ng IGF-I per 1 107 viable cells/24 h. This maximum
was observed on day 4 with a decrease to 28.474.2 at 36
days (Table 1).
Macroscopic findings
We next tested the hypothesis that chondrocyte-alginate
implants overexpressing IGF-I augment the repair of
In vitro chondrocyte-alginate implant analysis
Day post transfection
2
4
7
12
24
36
lacZ implants
Total cells ( 104)/implant
Total viable cells ( 104)/implant
Viability (%)
IGF-I (ng/107 viable cells/24 h)
IGF-I (pg/implant/24 h)
4.570
3.570.1
78.271.7
1.871.0
6.573.8
4.470.2
3.570.2
80.170.1
7.772.0
26.978.4
4.270.4
3.470.4
79.870.3
5.772.6
18.876.7
4.070.6
3.370.5
80.570.7
5.373.0
16.377.0
4.070.8
3.170.7
79.770.4
10.876.7
31.573.6
3.871.1
3.070.8
80.570.7
9.978.4
26.0717.0
IGF-I implants
Total cells ( 104)/implant
Total viable cells ( 104)/implant
Viability (%)
IGF-I (ng/107 viable cells/24 h)
IGF-I (pg/implant/24 h)
6.270.3
5.070.2
81.072.1
118.576.5
596.0759.7
6.570.1
5.170.1
80.373.1
123.2722.3
631.87101.5
6.870.6
5.670.5
80.071.3
96.679.0
534.271.8
7.171.0
5.770.9
80.072.8
26.170.3
147.6719.9
7.471.4
5.871.0
80.070.4
50.474.4
291.9725.8
7.771.8
6.271.6
79.071.4
28.474.2
121.87153.7
Lapine articular chondrocytes transfected with pCMVlacZ or pCMVhIGF-I were encapsulated in alginate on day 1 post transfection and kept
in basal medium containing 2% FBS for 36 days. Chondrocyte-alginate implants were collected at the denoted times. Released chondrocytes
were counted and their viability was assessed using a Neubauer chamber and trypan blue exclusion staining, respectively. IGF-I protein was
assessed as described in Materials and methods. Data are expressed as mean7standard deviations per time point of six samples per
condition.
Gene Therapy
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
articular cartilage defects in vivo. At 1 day after
encapsulation (2 days after transfection), lacZ or IGF-I
implants were press-fit into osteochondral defects in
each patellar groove of 12 rabbits. Animals were
euthanized and defects were analyzed three and 14
weeks following surgery. At 3 weeks after transplantation, the margins of the defects were visible and the new
tissue in all knees was whiter than the surrounding host
cartilage. At 14 weeks after transfection, the color of the
defects was similar to the surrounding articular cartilage
and the margins of the defects were difficult to discern.
No joint effusion, osteophytes or adhesions were
observed at either time point. Macroscopically, there
were no descriptive differences between knee joints
receiving lacZ and IGF-I implants at either time point.
(more normal) for defects treated with IGF-I implants
than for defects treated with lacZ implants (Po0.01 to
Po0.0001) (Table 4). Similarly, the mean total score at 3
weeks was significantly lower for defects treated with
IGF-I implants than for defects treated with lacZ
implants (Po0.0001; n ¼ 6) (Table 4, Figure 2).
At 14 weeks, the mean score for each individual
parameter continued to be significantly better in defects
that were treated with IGF-I implants than those that
1173
Microscopic evaluation of synovium
There was no significant difference in the thickness or
architecture of synovial villi or in the presence of
inflammatory cell infiltrates in knees receiving lacZ
implants compared to IGF-I implants at either 3 weeks
(P40.05) or 14 weeks (P40.05) (Table 2).
Immunohistochemical analysis
At 3 weeks after transfection, immunoreactivity to type II
collagen was negative in the repair tissue of both groups
(Figure 2). At 14 weeks after transfection, immunoreactivity to type II collagen in the repair tissue was
subjectively greater in both groups than at 3 weeks
following transfection (Figure 3). At both 3 and 14 weeks,
immunoreactivity to type II collagen was subjectively not
different in the matrix of defects receiving IGF-I implants
than in defects treated with lacZ implants.
Histological grading
Articular cartilage repair was evaluated using a previously published histological grading system developed
to quantitate repair of articular cartilage defects.4,5 This
grading system is based on eight parameters (Table 3)
that are scored individually. The individual scores are
then combined. The resulting total score ranges from 0
points (normal articular cartilage) to 31 points (no repair
tissue). At 3 weeks following surgery, the mean score for
each of the individual histological parameters was lower
Table 2 Histological grading of the synovium at 3 and 14 weeks
Category
IGF-I implants LacZ implants
Mean7s.d.
P-value
Mean7s.d.
3 weeks
Villus thickening
Villus architecture
Inflammatory cell infiltrate
Average total score
0.970.1
0.970.1
0.270.2
2.070.6
0.870.1
0.570.2
0.270.2
1.470.4
40.05
40.05
40.05
40.05
14 weeks
Villus thickening
Villus architecture
Inflammatory cell infiltrate
Average total score
0.370.5
0.870.5
0.370.4
1.570.5
0.370.5
0.570.6
0
0.870.4
40.05
40.05
40.05
40.05
Effects of IGF-I gene transfer at 3 and 14 weeks on histological
grading of the synovium.
Figure 2 Representative histologic sections of defects 3 weeks following
treatment with a lacZ implant (left; a, c, e, g) or IGF-I implant (right; b, d,
f, h) stained with safranin O (a–d), a monoclonal mouse anti-human typeII collagen IgG (e, f) or hematoxylin and eosin (g, h). Images (c) (d) are
magnified views of images (a) and (b) illustrating the area of integration
between the repair tissue (left side of each picture) with the adjacent normal
articular cartilage (right side of each picture). The implants are visible at
the bottom of images (a) and (b). Immunoreactivity to type II collagen is
present in the native articular cartilage and absent in the repair tissue of
both groups (e, f). The sections illustrated were taken from defects having a
histological rating equal to the mean score for its respective treatment
group. Photomicrographs were obtained using standardized photographic
parameters, including light intensity. Original magnifications 20 (a, b)
or 100 (c–h).
Gene Therapy
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1174
Table 3 Histological grading system
Category
1. Filling of the defect relative to surface of normal
adjacent cartilage
2. Integration of repair tissue with surrounding
articular cartilage
3. Matrix staining with safranin O-fast green
4. Cellular morphology
Normal
Mostly round cells with the morphology of
chondrocytes
50% round cells with the morphology of
chondrocytes
Mostly spindle-shape (fibroblast-like cells)
5. Architecture within entire defect (not including
margins)
6. Architecture of surface
7. Percentage of new subchondral bone
If new bone is below original tidemark
If new bone is above original tidemark
8. Formation of tidemark
Point value
0–4
0–3
0–3
0
0–2
2–4
5
0–4
0–3
0–4
0–4
0–4
Articular cartilage repair was quantitated using a histological
grading system previously published by Sellers et al.4,5 This grading
system is based on eight individual parameters. The individual
scores are then added. The resulting total score ranges from 0 points
(normal articular cartilage) to 31 points (empty defect with no repair
tissue). Adapted from Sellers et al.4,5
Figure 3 Representative histologic sections of defects 14 weeks following
treatment with a lacZ implant (left; a, c, e, g) or IGF-I implant (right; b, d,
f, h) stained with safranin O (a–d), a monoclonal mouse anti-human typeII collagen IgG (e, f) or hematoxylin and eosin (g, h). Images (c) and (d) are
magnified views of images (a) and (b) illustrating the area of integration
between the repair tissue (left side of each picture) with the adjacent normal
articular cartilage (right side of each picture). A residual lacZ implant can
be identified at the bottom of image (a). Defects receiving lacZ implants (e)
or IGF-I implants (f) are characterized by a positive immunoreactivity to
type II collagen. The repair tissue in the defect receiving a lacZ implant
(c, g) is hypocellular and characterized by a minimal matrix staining by
safranin O (c). The repair tissue of the defect receiving an IGF-I implant
(d, h) contains cells with the morphology of chondrocytes and safranin O
staining indicating matrix proteoglycans. The subchondral bone is nearly
completely restored. A tidemark is present in both groups (c, d).
Photomicrographs were obtained using standardized photographic parameters, including light intensity. Original magnifications 20 (a, b) or
100 (c–h).
observed for safranin O staining (lacZ and IGF-I
implants; Po0.01), cell morphology (IGF-I implants;
Po0.01), formation of subchondral bone (lacZ and IGFI implants; Po0.01) and tidemark (lacZ and IGF-I
implants; Po0.01). All other individual parameters
included in the histological scale were not significantly
different at these time points. The mean total score for
defects treated with lacZ and IGF-I implants improved
significantly during this time period (Po0.01).
Over the course of the in vivo experiment, all
chondrocyte-alginate implants were found below the
newly formed tidemark. To study the survival of
the chondrocytes within the alginate, we examined the
activity of caspase-3, a key enzyme activated in cells
undergoing apoptosis. Several chondrocytes in lacZ and
IGF-I implants were positive for activated caspase-3 after
3 weeks (Figure 1e and f). To evaluate the fate of the
implants, the area occupied by lacZ and IGF-I implants
excluding any ingrown tissue was measured. The area
occupied by lacZ and IGF-I implants after 3 weeks was
1.370.4 and 1.070.2 mm2, respectively. At 14 weeks
in vivo, the area occupied by lacZ and IGF-I implants
was 0.570.2 and 0.370.1 mm2, respectively (P40.05 for
both time points). These data suggest that the alginate
is progressively resorbed following placement in this
subchondral location.
Discussion
received lacZ implants (Po0.01 to Po0.0001) (Table 5).
The mean total score at 14 weeks also remained
significantly better in the IGF-I implant group
(Po0.0001; n ¼ 5) (Table 5, Figure 3).
Between 3 weeks and 14 weeks in vivo, improvement
in the score of individual histological parameters was
Gene Therapy
A major challenge in articular cartilage regeneration is
to identify both an effective therapeutic agent and a
responsive target cell population. An additional challenge is the identification of delivery systems for the
therapeutic agent and/or cells. Nearly all animal experiments for articular cartilage repair have required
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1175
Table 4 Histological grading of the repair tissue at 3 weeks
Category
IGF-I implants
lacZ implants
Mean (95% CI)
Filling of defect
Integration
Matrix staining
Cell morphology
Architecture within defect
Architecture of the surface
Subchondral bone
Tidemark
Average total score
0.32
0.42
2.75
2.17
0.62
1.10
2.18
2.48
12.0
(0.01–0.63)
(0.02–0.75)
(2.50–3.00)
(1.78–2.55)
(0.20–1.03)
(0.60–1.60)
(1.78–2.57)
(2.13–2.82)
(10.7–13.2)
F-test
P-value
7.01
9.56
11.86
15.56
12.58
70.17
169.88
153.40
258.48
o0.01
o0.01
o0.01
o0.001
o0.001
o0.0001
o0.0001
o0.0001
o0.0001
Mean (95% CI)
1.13
1.38
3.71
4.28
1.82
1.98
3.84
3.86
22.0
(0.72–1.54)
(1.05–1.72)
(3.46–3.96)
(3.89–4.67)
(1.39–2.24)
(1.48–2.46)
(3.44–4.23)
(3.51–4.20)
(20.7–23.2)
Effects of IGF-I gene transfer at 3 weeks on histological grading of repair tissue. Each category and total score is based on the average of two
independent evaluators. Points for each category and total score were compared between IGF-I and lacZ groups using a mixed general linear
model with repeated measures (knees nested within the same animals; CI ¼ confidence interval). Means indicate the estimated scores in
points for each category. Highly significant treatment effects were observed for each of the variables.
Table 5 Histological grading of the repair tissue at 14 weeks
Category
IGF-I implants
lacZ implants
Mean (95% CI)
Filling of defect
Integration
Matrix staining
Cell morphology
Architecture within defect
Architecture of the surface
Subchondral bone
Tidemark
Average total score
0.30
0.63
1.60
1.25
0.45
1.06
0.48
0.65
6.3
(0.15–0.45)
(0.39–0.86)
(1.38–1.82)
(0.90–1.60)
(0.13–0.77)
(0.85–1.28)
(0.24–0.72)
(0.37–0.93)
(5.3–7.3)
F-test
P-value
12.67
22.76
35.55
38.04
75.49
49.82
92.05
74.21
220.08
o0.01
o0.001
o0.0001
o0.0001
o0.0001
o0.0001
o0.0001
o0.0001
o0.0001
Mean (95% CI)
0.62
1.38
2.33
2.38
2.40
2.31
1.80
1.60
15.0
(0.50–0.74)
(1.15–1.60)
(2.12–2.54)
(2.05–2.73)
(2.08–2.72)
(2.10–2.52)
(1.57–2.03)
(1.34–1.86)
(14.0–16.1)
Effects of IGF-I gene transfer at 14 weeks on histological grading of repair tissue. Each category and total score is based on the average of two
independent evaluators. Points for each category and total score were compared between IGF-I and lacZ groups using a mixed general linear
model with repeated measures (knees nested within the same animals; CI ¼ confidence interval). Means indicate the estimated scores in
points for each category. Highly significant treatment effects were observed for each of the variables.
relatively large, often repeated, doses of growth factors.2–6,10
This is consistent with a relatively rapid clearance of
otherwise potent therapeutic molecules and suggests a
value for local growth factor production. Gene therapy at
the site of articular cartilage repair is a clinically relevant
application of gene transfer technology and represents a
potential solution to these problems. In this study, we
tested the hypothesis that overexpression of a human
IGF-I cDNA by articular chondrocytes transplanted in a
hydrogel enhances the repair of full-thickness cartilage
defects in vivo. The data indicate that chondrocytes
transfected by a lipid-based method with a human IGF-I
expression plasmid vector and encapsulated in alginate
in vitro secrete relevant amounts of recombinant human
IGF-I protein. The data further demonstrate that transplantation of these cells in alginate enhances the repair of
full-thickness osteochondral cartilage defects in vivo.
We selected IGF-I for these studies based on previous
work demonstrating that this factor augments chondrocyte mitotic activity, proteoglycan and type-II collagen
synthesis,7–9 new tissue formation by transplanted
chondrocytes,11 and the structural and functional properties of tissue-engineered cartilage.12 Several in vivo
studies further support the selection of IGF-I as a
candidate for articular cartilage repair.3,6 In a canine
model of osteoarthritis, intra-articular delivery of 2 mg
IGF-I three times per week for 3 weeks led to improvement in indices of cartilage damage.10 Treatment of
chondral (partial thickness) defects in adult rabbit and
Yucatan minipig models with 50 ng/ml IGF-I applied in
a fibrin clot improved the cellularity of the repair tissue.2
In a rat chondral (partial thickness) model, rib perichondrial cells containing an adenoviral IGF-I vector improved short-term repair when delivered in a fibrin
clot.14 Treatment of extensive osteochondral (full-thickness) defects with 25 mg IGF-I, applied in a fibrin
composite, enhanced repair in a horse model,3 as did
IGF-I coupled with cell-based treatment.6 The present
approach circumvents the high initial peak that occurs
after the delivery of a peptide4 and bypasses the
hindrance of the rapid clearance of IGF-I. In the present
study, 6.372.6 104 viable chondrocytes overexpressing
IGF-I embedded in alginate were transplanted directly
into an osteochondral articular cartilage defect. Such
local administration of these modified cells to the site of
the injury may be important since no morphologic
changes in articular cartilage were seen when 1 106
fibroblasts transduced ex vivo with a recombinant
adenovirus carrying a human IGF-I cDNA were injected
into the knee joint in a mouse model.15 In such cell-based
Gene Therapy
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1176
therapies, the overexpressed gene product is unlikely to
act in isolation. Potential interaction between the overexpressed IGF-I and factors produced by the transplanted and host cells in this model remain to be
elucidated.
The alginate system16,17 was selected because of its
well-characterized ability to maintain the differentiated
phenotype of chondrocytes,18,19 the comparability of its
negative charge density to that of native cartilage
matrix,20 its stability over short time periods16 and its
potential to allow allogeneic cell transplantation.21
Alginate has been successfully used to deliver bone
marrow-derived cells22 and chondrocytes23 to osteochondral articular cartilage defects in a rabbit model. In
addition, alginate has been shown not to influence
osteoarticular defect healing in the rabbit knee24 and
may protect transplanted allogeneic cells from immune
rejection.25 Furthermore, chondrocytes embedded in
alginate are capable of expressing a transgene for at
least 6 weeks in vitro26 and 32 days in vivo even
when nonviral, nonintegrating gene delivery is used.13
Finally, alginate has already been used in human
studies.27 The present data indicate that transfection
with an IGF-I expression plasmid vector does not impair
chondrocyte viability when encapsulated in alginate.
The data further indicate that transfection with an
IGF-I expression plasmid vector is sufficient to induce
cell proliferation in this in vitro model system as
determined by an increased cell number after 36 days
of cultivation.
Chondrocytes embedded in alginate have been reported not to significantly participate in the generation of
the repair tissue in an osteochondral articular cartilage
defect.23 The data from the present study suggest that the
IGF-I released by the cells within the alginate achieved
biologically relevant concentrations in vitro and was
sufficient to stimulate osteochondral defect repair in vivo.
Chondrocytes were selected for these studies because
they are adapted to the avascular environment of the
joint and normally exist in suspension in a matrix. The
fate of the transplanted cells in these full-thickness
defects is uncertain. The data from the present study
indicate that apoptotic cell death affects a part of the
transplanted cells. Other cells may have undergone
necrosis, migration or other fate following transplantation. It is possible that chondrocyte survival within the
implants was mediated by the overexpression of IGF-I.28
Transgene expression in vivo13 suggests that some cells
remain alive, while implant resorption suggests eventual
cell removal. These results are consistent with the
observation in prior studies that chondrocytes placed in
these full-thickness defects do not appear to remain for
long time periods.23,29
We employed full-thickness, osteochondral defects
because these are analogous to the common clinical
circumstance in which defects penetrate, or are treated
by methods that penetrate, the subchondral bone.1 These
methods permit marrow-derived cells to participate in
the repair process. IGF-I has been shown to promote
the differentiation of marrow-derived3 and synovialderived2 cells into cartilage. Thus, the augmented repair
may reflect the action of IGF-I produced by the
transplanted cells on marrow-derived cells, synovialderived cells, the transplanted chondrocytes or all of
these cell types.
Gene Therapy
The model system of an osteochondral articular
cartilage defect in the rabbit knee joint was chosen based
on its widespread use as a tool to study articular cartilage
repair.2,4,5,30–33 It is important to note that there is
insufficient intrinsic healing of the defects in this model
system, as evidenced by the high (abnormal) score values
in the lacZ control group at both time points. We have
previously demonstrated that recombinant genes can be
introduced into osteochondral defects in vivo by transfected articular chondrocytes in alginate and that
transgene expression remains present in vivo for at least
32 days post transfection.13 Overexpression of human
IGF-I by articular chondrocytes has been shown to
increase cell proliferation and matrix synthesis as early
as 5 days in an in vitro model of cartilage repair11 and to
enhance the functional properties of tissue-engineered
cartilage as early as 28 days.12 A mean residence time
of 8 days was estimated for human bone morphogenic
protein-2 (BMP-2) delivered in a collagen sponge to
osteochondral cartilage defects in rabbits in vivo.4 This
relatively short time period was sufficient to elicit
significant improvements in the histological appearance
and composition of the repair tissue.4 In the present
study, improvements in articular cartilage repair were
already apparent at 3 weeks. It is interesting to note that
the magnitude of improvement in articular cartilage in
the present study is similar to the enhancement obtained
when recombinant human BMP-2 protein was applied in
a collagen sponge.5
We used a grading system developed specifically for
the quantitative assessment of articular cartilage defect
repair4,5 (Table 3) rather than a measure designed to
assess arthritic changes. The scale is sensitive to change
over time and has been used in this model to evaluate the
effect of other cell-signaling molecules on articular
cartilage repair.4,5 Indeed, significant improvements in
articular cartilage repair were seen for both groups at
the late time point of 14 weeks compared to 3 weeks. The
improvements in cellular morphology and architecture
observed in vivo within these defects are in agreement
with our previous findings that IGF-I overexpression
enhances cellularity and matrix formation in an in vitro
model of articular chondrocyte transplantation.11
Efforts to achieve effective articular cartilage repair in
vivo have been hindered in part by difficulty in achieving
integration of the newly formed tissue with the
surrounding cartilage at the margins of the defect.32
Articular cartilage defects treated with chondrocytes
overexpressing IGF-I showed significantly better integration of the repair tissue at the interface with the
surrounding normal articular cartilage than did control
defects. Such integration appears to be important in
establishing the biomechanical environment of both the
repair and surrounding cartilage.34
In summary, the data indicate that alginate is a
suitable three-dimensional carrier for genetically modified allogeneic articular chondrocytes that allows for a
prolonged secretion and release of bioactive recombinant
human IGF-I in vitro. The data further demonstrate that a
single application of such implants containing chondrocytes secreting recombinant human IGF-I to osteochondral articular cartilage defects is sufficient to augment
articular cartilage repair for a prolonged period of time.
These results suggest that therapeutic growth factor gene
delivery via encapsulated and transplanted genetically
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
modified cells may be applicable to sites of focal,
traumatic articular cartilage damage. Further studies
will be needed to evaluate the long-term properties of
this repair tissue.
Materials and methods
Materials
Reagents were obtained from Invitrogen/Gibco (Karlsruhe, Germany) unless otherwise indicated. Collagenase
type I (activity: 232 U/mg) was from Biochrom (Berlin,
Germany). Bovine testicular hyaluronidase and alginate
were from Sigma (Munich, Germany). Plasticware was
from Falcon (Becton Dickinson, Pont de Claix, France).
Cells, expression plasmid vectors and transfection
Articular cartilage was obtained from the knee and hip
joints of male Chinchilla bastard rabbits (mean weight:
2.870.4 kg; Charles River, Sulzfeld, Germany). The
animals were in their late juvenile stage according to
post-mortem analysis of their distal femoral growth
plate. The cartilage was washed, diced into 2.0 2.0
0.5 mm pieces and transferred to DMEM with 50 mg/
ml ascorbic acid, 100 U/ml penicillin G and 100 ml/ml
streptomycin (basal medium) containing 0.02% collagenase at 371C in a humidified atmosphere with 10% CO2
for 16 h. Isolated cells were filtered through a 125 mm
mesh to remove undigested matrix. Cell number was
determined by hemocytometry. Viability, as determined
by trypan blue exclusion, always exceeded 90%.
In vitro studies were based on optimized transfection
of primary cultures of lapine articular chondrocytes
using the nonliposomal lipid formulation FuGENE 6
(Roche, Mannheim, Germany) essentially as previously
described.13 Briefly, chondrocytes were transfected with
endotoxin-free expression plasmid vectors carrying
either the E. coli lacZ gene (pCMVlacZ; lacZ-transfected),35 or a human IGF-I cDNA (pCMVhIGF-I; IGF-Itransfected)11,12 under the control of the human cytomegalovirus (CMV) immediate-early promoter/enhancer. Briefly, plasmid DNA was complexed with FuGENE6 in Opti-MEM and transferred to subconfluent
chondrocyte monolayers. Cells were treated with 4 U/ml
bovine testicular hyaluronidase 12 h before and during
transfection. Transfections of chondrocytes for encapsulation in alginate were performed with chondrocytes at
passage 2, 10–14 days after cell isolation, seeded at a
density of 0.8 106 cells/dish in basal medium containing 10% fetal bovine serum (growth medium) in 100 mm
dishes.13
Construction of chondrocyte-alginate implants
To facilitate the transplantation of chondrocytes into
articular cartilage defects, implants were made by using
chondrocytes in the hydrogel, alginate. Transfected
chondrocytes were encapsulated in alginate essentially
as previously described.13 Briefly, 1 day after transfection, chondrocytes were trypsinized, washed and suspended in 1.2% alginate in 0.15 M NaCl at 2 106 cells/
ml. The cell suspension was then extruded into a 102 mM
CaCl2 solution at room temperature under constant
shaking and the chondrocytes-alginate composite was
allowed to polymerize for 10 min. The resulting implants
were then washed twice in 0.15 M NaCl followed by two
consecutive washes in basal medium and placed in
growth medium (five implants/ml medium) that was
changed three times per week and kept at 371C in a
humidified atmosphere of 10% CO2. Implant diameters
were measured 2 h after encapsulation using a microscope fitted with a micrometer scale. Cultured chondrocyte-alginate implants were assessed for cell number and
viability at 2, 4, 7, 12, 24 and 36 days post transfection.
Individual implants were solubilized by incubation in
100 ml 55 mM sodium citrate, 90 mM NaCl, pH 6.8 for
20 min at room temperature. Released chondrocytes
were counted and their viability assessed using a
Neubauer chamber and trypan blue exclusion staining
based on 4 counts per sample. To determine IGF-I
production by chondrocytes encapsulated in alginate,
chondrocyte-alginate implants were individually cultivated in 96-well plates in growth medium. At the
indicated time points, implants were washed twice in
basal medium and the medium was replaced with 200 ml
basal medium. After 24 h, the conditioned medium was
collected, centrifuged to remove cell debris and stored
at 801C.
1177
Transplantation to articular cartilage defects in vivo
The Saarland Governmental Animal Care Committee
approved all animal procedures. Rabbits (Charles River)
were kept in air-conditioned rooms with constant
temperatures and a regular light/dark scheme. They
were fed a standard diet and received water ad libitum. A
total of 12 male Chinchilla bastard rabbits (six animals
per group) were anesthetized by intramuscular injection
of Ketavet (0.75 mg/kg of body weight; Pharmacia &
Upjohn, Erlangen, Germany) and Rompun (0.2 ml/kg of
body weight; Bayer, Leverkusen, Germany). Ampicillin
(Pfizer, Karlsruhe, Germany) was administered intramuscularly 2 h preoperatively at a dosage of 25 mg/kg of
body weight. The knee joint was entered through a
medial parapatellar approach. The patella was dislocated
laterally and the knee flexed to 901. A cylindrical
osteochondral cartilage defect was created in each
patellar groove (n ¼ 24 defects) with a manual cannulated burr (3.2 mm in diameter; Synthes, Umkirch,
Germany). Each defect was washed with saline and
blotted dry. Chondrocyte-alginate implants were pressfit into the defects 1 day after encapsulation (2 days post
transfection). Implants from one preparation were
employed in all defects. Implants constructed from
lacZ-transfected chondrocytes and alginate contained
6.970.9 104 chondrocytes with a viability of
90.772.5%. Implants constructed from IGF-I-transfected
chondrocytes and alginate contained 6.372.6 104 chondrocytes with a viability of 87.171.9%.
The right and left knees alternately received lacZ- or
IGF-I implants. The patella was reduced and the knee
was put through a range of motion to assure the stability
of the implants. The incisions were closed in layers. No
postoperative immobilization was used and animals
were allowed immediate full weight bearing. After 6
weeks postoperation, animals were allowed to climb and
jump.
Gene transfer analyses
Mean transfection efficiency was assessed by X-gal
staining of lacZ-transfected chondrocytes in monolayer
culture as previously described.35 To determine IGF-I
Gene Therapy
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1178
protein production, conditioned medium was analyzed
by ELISA (R&D Systems, Minneapolis, MN, USA) with a
detection limit of 26 pg/ml.
Histological and immunohistochemical analysis
Alginate implants cultured in vitro were fixed in 10%
phosphate-buffered formalin, embedded in paraffin,
sectioned and submitted to histological analysis as
described below. One rabbit was removed from the
protocol because of death following a gastrointestinal
infection at day 22 post operation. Three (n ¼ 6) and 14
weeks (n ¼ 5) after transfection, animals were euthanatized and the knee joints were exposed and examined
grossly for synovitis, adhesions or other adverse reactions from the implants. The appearance of the repair
tissue (color, integrity, contour) and articular surfaces
was documented. Distal femurs were retrieved, fixed
in 10% phosphate-buffered formalin, trimmed to
2.0 1.5 1.0 cm and decalcified.
Paraffin-embedded frontal sections (5 mm) were
stained with safranin O and hematoxylin and eosin
according to routine histological protocols. To study
activated caspase-3, sections were deparaffinized in
xylene, passed through decreasing concentrations of
ethanol, washed in PBS and submerged in 0.3% hydrogen peroxide for 30 min. After washing with PBS,
sections were incubated in 0.1% Trypsin for 30 min,
washed with PBS and blocked with 3% bovine serum
albumin in PBS (blocking buffer) for 30 min. Sections
then were incubated overnight at room temperature
with a 1:50 dilution of a cleaved caspase-3 polyclonal
antibody (Asp175; Cell Signaling Technology, Frankfurt,
Germany). This antibody detects endogenous levels
of only the short fragment (17/19 kDa) of activated
caspase-3, but not of the full-length caspase-3. A
biotinylated goat anti-mouse antibody was used as a
secondary antibody for streptavidin–biotin complex
peroxidase staining (1:200; LSAB 2 System HRP; DakoCytomation, Hamburg, Germany). Diaminobenzidine
(DakoCytomation) was used as chromogen. Sections
were counterstained with hemalaun and examined
by light microscopy. For type-II collagen immunostaining, deparaffinized sections were submerged in 0.3%
hydrogen peroxide for 30 min. After washing with
PBS, sections were incubated in 0.1% Trypsin for
30 min, washed with PBS and blocked with blocking
buffer for 30 min. Sections then were incubated with a
1:10 dilution of a monoclonal mouse anti-human
type-II collagen IgG (Medicorp, Montréal, Canada) in
blocking buffer for 24 h at 41C, washed and exposed
to a 1:50 dilution of a biotinylated anti-mouse antibody (Vector Laboratories, Burlingame, CA, USA)
for 1 h at room temperature. Sections were washed,
incubated for 30 min with avidin–biotin–peroxidase
reagent (Vectastain Elite ABC kit; Vector Laboratories),
washed and exposed to diaminobenzidine (Vector
Laboratories). Control tissues for the primary antibodies
included lapine skin and lapine articular cartilage. To
control for secondary immunoglobulins, sections were
processed as above with omission of the primary
antibody.
Evaluation of the sections
Serial histological sections of the distal femora were
taken at 200 mm intervals. All sections were taken within
Gene Therapy
approximately 1.2 mm from the center of the defects
(n ¼ 5–10 per defect). Two individuals with no knowledge of the treatment groups independently graded all
sections using an articular cartilage repair scoring
system.5 Specific parameters evaluated in this system
include defect filling, integration, safranin O staining,
cell morphology, defects architecture, surface architecture, subchondral bone and tidemark formation (Table
3). Each section was scored and all scores for each
treatment group were combined to determine the mean
score for each group. A total of 181 sections were scored.
In each joint, the synovium was evaluated using a
separate scoring system.6 The three categories in this
scoring system include villus thickening (fibrosis), villus
architecture (blunting) and the presence of inflammatory
cell infiltrates.
The area occupied by lacZ and IGF-I implants was
measured from hematoxylin and eosin-stained serial
histological sections of the distal femora that were taken
at 400 mm intervals. Sections were taken within approximately 1.2 mm from the center of the defects (n ¼ 4–6
per defect). Low-magnification images of the cartilage
defects were acquired by a solid-state CCD camera
(Olympus, Hamburg, Germany) mounted on a microscope (BX 45; Olympus, Hamburg, Germany) and
analyzed with the analySIS program (Soft Imaging
System Corp., Münster, Germany).
Statistical analysis
On the basis of literature values for selected cartilage
repair procedures, a standard deviation of 25% for the
mean total score was estimated to determine the sample
size. For a power of 80% and a two-tailed a level of 0.05, a
sample size of six animals per group would be required
to detect a mean difference of 5 points between the
groups assuming a pooled standard deviation of 2.5
points (effect size ¼ 5/2.5 ¼ 2.0, using the two-sample
Student’s t-test (version 5.0, nQuery Advisor, Statistical
Solutions, Saugus, MA, USA).
Each test condition was performed in quadruplicate
for in vitro characterization experiments and with six
defects per group and time point for in vivo experiments.
To evaluate the in vivo experiments, points for each
category and total score were compared between the two
groups using a mixed general linear model with
repeated-measures (knees nested within the same animals). Data are expressed as mean 795% confidence
interval. A two-tailed Po0.05 was considered statistically significant.
Acknowledgements
We thank J Becker, E Gluding and T Thurn for expert
technical assistance and B Vollmar for valuable discussions. This study is supported by the Deutsche Forschungsgemeinschaft (DFG MA 2363/1-1, H.M.), the AO
ASIF Foundation and NIH Grants AR 31068 and AR
45749 (SBT).
References
1 Buckwalter JA, Mankin HJ. Articular cartilage repair and
transplantation. Arthritis Rheum 1998; 41: 1331–1342.
Enhanced cartilage repair by IGF-I overexpression
H Madry et al
1179
2 Hunziker EB, Rosenberg LC. Repair of partial-thickness defects
in articular cartilage: cell recruitment from the synovial
membrane. J Bone Joint Surg Am 1996; 78: 721–733.
3 Nixon AJ, Fortier LA, Williams J, Mohammed H. Enhanced
repair of extensive articular defects by insulin-like growth
factor-I-laden fibrin composites. J Orthop Res 1999; 17: 475–487.
4 Sellers RS et al. Repair of articular cartilage defects one year
after treatment with recombinant human bone morphogenetic
protein-2 (rhBMP-2). J Bone Joint Surg Am 2000; 82: 151–160.
5 Sellers RS, Peluso D, Morris EA. The effect of recombinant
human bone morphogenetic protein-2 (rhBMP-2) on the healing
of full-thickness defects of articular cartilage. J Bone Joint Surg
Am 1997; 79: 1452–1463.
6 Fortier LA, Mohammed HO, Lust G, Nixon AJ. Insulin-like
growth factor-I enhances cell-based repair of articular cartilage.
J Bone Joint Surg Br 2002; 84: 276–288.
7 Trippel SB, Van Wyk JJ, Foster MB, Svoboda ME. Characterization of a specific somatomedin-c receptor on isolated bovine
growth plate chondrocytes. Endocrinology 1983; 112: 2128–2136.
8 Trippel SB. Growth factor actions on articular cartilage.
J Rheumatol Suppl 1995; 43: 129–132.
9 Trippel SB et al. Effect of somatomedin-C/insulin-like growth
factor I and growth hormone on cultured growth plate and
articular chondrocytes. Pediatr Res 1989; 25: 76–82.
10 Rogachefsky RA, Dean DD, Howell DS, Altman RD. Treatment of
canine osteoarthritis with insulin-like growth factor-1 (IGF-1) and
sodium pentosan polysulfate. Osteoarthr Cartilage 1993; 1: 105–114.
11 Madry H, Zurakowski D, Trippel SB. Overexpression of human
insulin-like growth factor-I promotes new tissue formation in an
ex vivo model of articular chondrocyte transplantation. Gene Ther
2001; 8: 1443–1449.
12 Madry H et al. Gene transfer of a human insulin-like growth
factor I cDNA enhances tissue engineering of cartilage. Hum
Gene Ther 2002; 13: 1621–1630.
13 Madry H et al. Sustained transgene expression in cartilage
defects in vivo after transplantation of articular chondrocytes
modified by lipid-mediated gene transfer in a gel suspension
delivery system. J Gene Med 2003; 5: 502–509.
14 Gelse K et al. Articular cartilage repair by gene therapy using
growth factor-producing mesenchymal cells. Arthritis Rheum
2003; 48: 430–441.
15 Gelse K et al. Fibroblast-mediated delivery of growth factor
complementary DNA into mouse joints induces chondrogenesis
but avoids the disadvantages of direct viral gene transfer.
Arthritis Rheum 2001; 44: 1943–1953.
16 Bonaventure J et al. Reexpression of cartilage-specific genes by
dedifferentiated human articular chondrocytes cultured in
alginate beads. Exp Cell Res 1994; 212: 97–104.
17 Lemare F et al. Dedifferentiated chondrocytes cultured in
alginate beads: restoration of the differentiated phenotype and
of the metabolic responses to interleukin-1beta. J Cell Physiol
1998; 176: 303–313.
18 von der Mark K, Gauss V, von der Mark H, Mueller P.
Relationship between cell shape and type of collagen synthesis
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
as chondrocytes lose their cartilage phenotype in culture. Nature
1977; 267: 531–532.
Benya PD, Padilla SR, Nimni ME. Independent regulation of
collagen types by chondrocytes during the loss of differentiated
function in culture. Cell 1978; 15: 1313–1321.
You JO et al. Preparation of regular sized Ca-alginate microspheres using membrane emulsification method. J Microencapsul
2001; 18: 521–532.
Zimmermann U et al. Hydrogel-based non-autologous cell and
tissue therapy. Biotechniques 2000; 29: 564–572.
Diduch DR, Jordan LC, Mierisch CM, Balian G. Marrow stromal
cells embedded in alginate for repair of osteochondral defects.
Arthroscopy 2000; 16: 571–577.
Mierisch CM et al. Chondrocyte transplantation into articular
cartilage defects with use of calcium alginate: the fate of the cells.
J Bone Joint Surg Am 2003; 85-A: 1757–1767.
Mierisch CM et al. Transforming growth factor-beta in calcium
alginate beads for the treatment of articular cartilage defects in
the rabbit. Arthroscopy 2002; 18: 892–900.
Rokstad AM et al. Transplantation of alginate microcapsules
with proliferating cells in mice: capsular overgrowth and
survival of encapsulated cells of mice and human origin. Ann
NY Acad Sci 2001; 944: 216–225.
Dinser R et al. Comparison of long-term transgene expression after non-viral and adenoviral gene transfer into
primary articular chondrocytes. Histochem Cell Biol 2001; 116:
69–77.
Soon-Shiong P et al. Insulin independence in a type 1 diabetic
patient after encapsulated islet transplantation. Lancet 1994; 343:
950–951.
Loeser RF, Shanker G. Autocrine stimulation by insulinlike growth factor 1 and insulin-like growth factor 2 mediates
chondrocyte survival in vitro. Arthritis Rheum 2000; 43:
1552–1559.
Grande DA et al. The repair of experimentally produced defects
in rabbit artiular cartilage by autologous chondrocyte transplantation. J Orthop Res 1989; 7: 208–218.
Bentley G, Greer III RB. Homotransplantation of isolated
epiphyseal and articular cartilage chondrocytes into joint
surfaces of rabbits. Nature 1971; 230: 385–388.
Furukawa T, Eyre DR, Koide S, Glimcher MJ. Biochemical
studies on repair cartilage resurfacing experimental defects in
the rabbit knee. J Bone Joint Surg Am 1980; 62: 79–89.
Shapiro F, Koide S, Glimcher MJ. Cell origin and differentiation
in the repair of full-thickness defects of articular cartilage. J Bone
Joint Surg Am 1993; 75: 532–553.
Hunziker EB. Biologic repair of articular cartilage. Defect models
in experimental animals and matrix requirements. Clin Orthop
1999; 367S: S135–S146.
Bae WC, Rivard KL, Law AW, Sah RL. Indentation testing is
sensitive to cartilage laceration and repair. Transactions 48th Ann
Mtg Orthop Res Soc 2002; 27: 399.
Madry H, Trippel SB. Efficient lipid-mediated gene transfer to
articular chondrocytes. Gene Ther 2000; 7: 286–291.
Gene Therapy