Thickness of perichondrial layer
after 2 (A) and 4 (B) weeks of culturing (hematoxylin-eosin, original magnification×
Cell migration into dermal type
I collagen matrix after 2 (A) and 4 (B) weeks of culturing (hematoxylin-eosin,
original magnification ×20).
Immunolocalization after 4 weeks
of culturing (fluorescein isothiocyanate–conjugated rabbit antigoat
antibody, original magnification ×80).
incorporation of cells on the bottom of the wells.
incorporation of cells in the explants.
van Lange JWL, de Roo K, Middelkoop E, van den Bos T, Everts V, Nolst TGJ. Perichondrium-Wrapped Collagenous Matrices to Induce ChondroneogenesisAn In Vitro Study. Arch Facial Plast Surg. 2001;3(2):122-126. doi:
From the Departments of Otorhinolaryngology (Drs van Lange, de Roo,
and Nolst Trenité), Dermatology (Dr Middelkoop), Periodontology, Academical
Center of Dentistry–Amsterdam (Drs van den Bos and Everts), and Cell
Biology and Histology (Dr Everts), Academical Medical Center, University of
Amsterdam, Amsterdam, the Netherlands.
Copyright 2001 American Medical Association. All Rights Reserved.
Applicable FARS/DFARS Restrictions Apply to Government Use.2001
Objective To develop a model for cartilage regeneration in vitro, to be used for
cartilage reconstruction in vivo.
Methods Collagenous matrices were wrapped in a perichondrium layer. The matrices
served as carriers to allow migration of cells from the perichondrium into
the matrix. Culture conditions stimulated cell growth and proliferation.
Results After 4 weeks of culturing, microscopic examination showed an increase
of cell layers around the matrices but also of cells migrated into porous
matrices. Immunohistochemical staining of these cultured cells expressed type
II collagen intracellularly.
Conclusions This model seems appropriate to culture mucoperichondrial explants in
combination with collagenous matrices. Cells migrate into the pores of the
matrix, survive, and synthesize matrix components. Actual formation of cartilage
has not been shown to occur. Adding growth factors to this model may influence
induction of this activity.
IN FUNCTIONAL-AESTHETIC surgery of the nose, a variety of different
implant materials are in use. These implants are essential in closing cartilaginous
defects (eg, septum perforation), augmentation of the nose (eg, onlay graft
of the nose dorsum), and reconstruction of the contour of the nose (eg, cleft
palate syndrome). Alloplastic materials have been developed, such as silicone,
porous polyethylene, porous polytetrafluoroethylene-carbon, and porous polytetrafluoroethylene,
of which the last one is most frequently used currently. These materials are
biocompatible but not completely without reaction of the host.1-3
Allogeneic or xenogeneic implants, such as irradiated bone or cartilage,
are readily accepted by the host and show minimal tissue reaction. On the
other hand, long-term follow-up shows a tendency for their gradual resorption.2-5 Also
frequently used is autogenous cartilage. It can be obtained from the nose
(nasal septum), and in elaborate reconstruction it can be harvested from the
ear (conchae cartilage) or rib (costal cartilage). It is the implant material
of first choice, despite the morbidity of the donor site.3
This morbidity is the reason that numerous research groups are searching for
alternative implant materials. Bean and coworkers6-7
have shown that demineralized bovine bone matrix can be transformed into a
suitable cartilage substitute by exposing this material to a vascularized
perichondrial flap.6-7 In vitro
studies by Bujía et al8-9
showed that it is possible to obtain a 3-dimensional cartilaginous structure
from polylactic acid fleeces and human chondrocytes in agarose gel. Kim et
al10 engineered nasoseptal cartilage replacements
through a process of neomorphogenesis of cartilage by culturing bovine chondrocytes
in close relation to biodegradable polymers composed of polyglycolic acid
and poly-L-lactic acid. However, an implant that is easily obtained with all
the benefits of autogenous cartilage has yet to be found.
The goal of the present study was to find a model for cartilage regeneration
in vitro.8-9,11 For
this purpose, different collagenous matrices were wrapped in a perichondrium
layer.7, 12-13 The
matrices serve as a carrier for migration of cells from the perichondrium
into the matrix. Optimal culture conditions should stimulate cell growth and
proliferation. After some time, the obtained matrix in vitro will be used
for reconstruction in vivo. The implanted carrier will resorb gradually2-5,14-15
and subsequently be replaced by the newly formed matrix.
Biomaterials used for this purpose need to meet a number of criteria,
such as having no cytotoxicity, causing no graft-vs-host reaction, undergoing
gradual biodegration, and allowing ingrowth of cells. The finally replaced
tissue should be as good as or even better than autogenous grafts. The reason
for culturing perichondrium instead of chondrocytes is that chondrocytes are
fully differentiated and have probably lost the proliferation and differentiation
capacity that is characteristic of perichondrial cells.11-13,16-19
The following materials were purchased for this study: Dulbecco modified
Eagle medium, fetal calf serum, glutamine, and phosphate-buffered saline (Gibco,
Merelbeke, Belgium); gentamicin sulfate (Centrafarm, Etten-Leur, the Netherlands);
penicillin G sodium (Yamanouchi, Leiderdorp, the Netherlands); ascorbic acid
and 1% bovine serum albumin (BSA) (Sigma-Aldrich Corp, St Louis, Mo); petri
dishes (Greiner, Frickenhausen, Germany); 12- and 24-well plates and low-affinity
well plates (Costar, Cambridge, Mass); goat anti–bovine type II collagen
and fluorescein isothiocyanate–conjugated rabbit antigoat antibody (Southern
Biotechnology Associates Inc, Birmingham, Ala); mounting medium (Vectashield;
Vector Laboratories Inc, Burlingame, Calif); and tritiated proline (937 mBq/mL)
(New England Nuclear, Boston, Mass).
Mucoperichondrium was obtained from the nasal septum of adult rabbits
(female New Zealand white, 24 weeks; Charles River, Someren, the Netherlands).17-18 Fentanyl citrate, 0.315 mg/mL, with
fluanisone, 10 mg/mL (Hypnorm; Janssen Pharmaceutica, Beerse, Belgium), was
used to anesthetize the rabbit before killing; the nasal septum was excised
in toto and divided into 3 equal parts (7 × 10 mm). From each part,
the mucoperichondrial layer was removed from both sides with a perichondrium
elevator. The mucoperichondrium was rinsed 4 times with phosphate-buffered
saline (PBS). In a petri dish, a matrix was presoaked with medium and the
mucoperichondrial layer was wrapped around this matrix with the perichondrial
part of the tissue facing the matrix. The mucoperichondrium was attached to
the selected matrix with a small hemoclip (Ligaclip; Johnson & Johnson,
Amersfoort, the Netherlands) and transferred to culture plates. For each experiment
the number in 1 group was 4.
The medium consisted of Dulbecco modified Eagle medium supplemented
with 10% fetal calf serum, 1% glutamine, gentamicin sulfate (50 µg/mL),
penicillin G sodium (50 µg/mL), and ascorbic acid (50 µg/mL).
After fixation of the perichondrium to the matrix, the medium was refreshed
every half hour for 2 hours. After this, the medium was refreshed 3 times
a week. The well plate was stored in a humidified incubator at 37°C in
an atmosphere of 5% carbon dioxide and 95% air (Stericult 200; Brouwer Scientific
Inc, De Meern, the Netherlands).
The solid collagen matrices used in this study are listed in Table 1. Most of them consisted of type
I collagen matrices with different porosity. One matrix was of type II collagen.
The matrices (5 × 5 × 1 mm) were sterilized overnight by gamma
irradiation. Before use, the matrices were soaked in culture medium for 30
minutes. In addition to the dentine matrix, demineralized bovine bone matrix,
and commercially available demineralized bovine bone matrix (Osteovit; B.
Braun Surgical GmbH, Melsungen, Germany), at least 3 samples from each group
of different carriers (n = 4) were analyzed.
Tissue was fixed in a solution of 4% paraformaldehyde in 0.1-mol/L PBS
(pH 7.4), dehydrated, and embedded in glycol methacrylate. Sections were made
of the mucoperichondrium-wrapped matrices and stained with hematoxylin-eosin.
The cells on the bottom of the well plate were fixed as indicated above and
stained directly in the wells.
After the culture period, the tissue was rinsed 4 times with PBS and
fixed for 5 minutes in 75% ethanol. The tissue was air-dried and frozen at −20°C.
Cryosections were made for immunohistochemical staining with an antibody
against type II collagen. For this purpose, sections were incubated first
with 1% BSA in PBS (BSA/PBS) to block nonspecific binding (room temperature,
30 minutes). The sections were washed with 0.05% PBS-Tween (3 times, 15 minutes
each time). The anti–type II collagen was diluted 1:250 in BSA/PBS,
added to the cryosections, and incubated for 90 minutes at room temperature.
The sections were rinsed with PBS-Tween (3 times, 15 minutes each time). Fluorescein
isothiocyanate–conjugated rabbit antigoat antibody was diluted 1:500
in BSA/PBS and the sections were incubated for 60 minutes (room temperature).
The nuclei were stained with propidium iodide diluted in 1:100 PBS for 10
minutes. Finally, the sections were rinsed with PBS-Tween (3 times, 15 minutes
each time). The sections were covered with mounting medium (Vectashield).
Tritiated proline (0.74 MBq per well) was added to explants 24 hour
before the tissue was collected. The tissue samples and wells were washed
3 times with PBS, then 300 µL of 1N sodium hydroxide was added and kept
for 20 minutes at room temperature. The cells were homogenized and 100 µL
was taken out and analyzed in the scintillation counter (Wallac 1450 Microbeta
Plus; Wallac, Turku, Finland).
Autoradiography of the tissue explants was performed by adding tritiated
proline in the last 24 hours before finishing the culture (as described by
Everts and Beertsen20).
After preparation of the mucoperichondrium, the perichondrial layer
was very dense and 5 to 6 layers thick. In culture this layer lost its dense
scaffold, and in the first days of culturing a number of cells appeared to
die. This was assumed because of the turbid yellowish appearance of the medium.
After the first refreshment of the medium, the turbid yellowish appearance
was no longer seen. After the second refreshment of the medium, cells were
seen to migrate to the bottom of the well plate. During the next weeks of
culturing, the numbers of these cells increased considerably, finally occupying
almost the entire bottom of the well. Microscopic examination of the explants
showed that, compared with 2 weeks of culturing, at 4 weeks the thickness
of the perichondrial layer around the matrix had increased from 6 to 9 cell
layers (Figure 1). The space between
the matrix and the mucoperichondrium was filled up with cells, which originated
from the perichondrium.
Cells from the perichondrial layer migrated into some of the matrices.
This was only apparent in matrices with pores; in the nonporous dental and
cartilaginous matrices, such a migration of cells was absent. This migration
was more pronounced after 4 weeks of culturing than that after 2 weeks (Figure 2). The number of cells that had migrated
into matrices containing elastin was similar to the number found in cross-linked
No differences in cellular morphologic characteristics were found between
mucoperichondria in contact with type II or type I collagenous matrices. In
an attempt to increase migration of cells in collagen type II matrices, holes
were burred in these matrices. Despite this treatment, the porosity of the
matrices remained very low compared with that of dermal bovine type I collagen
matrix, and an increase in migration was not seen. In addition, epiphyseal
plate of bovine bone was tested because of the presence of naturally formed
canals in this matrix. Since a tight fixation of the mucoperichondrium to
the matrix was not possible because of the irregularity of this matrix, migration
of cells into this matrix could not be evaluated well. Up to 4 weeks of culturing,
in this in vitro model, no resorption or disintegration (except for commercially
available demineralized bovine bone matrix [Osteovit]) of all the used matrices
could be identified. Taken together, the dermal bovine type I collagen matrix
proved to best suit our goals and was selected for use in subsequent experiments.
Since numerous cells appeared to migrate from the mucoperichondrium
to the bottom of the wells and not into the matrices, stimulation of migration
into the matrix was attempted by (1) placing the explants on a grid or (2)
using low-affinity well plates. Under these conditions, however, an increased
invasion into the matrix was not found. So far, no statistical analysis has
been performed on these attempts.
After culturing for up to 4 weeks, the morphologic structure of the
cells attached to the bottom of the wells had changed from an elongated fibroblastlike
shape to a more rounded one.21 Immunohistochemical
staining of these cells with anti–type II collagen showed some expression
of this protein intracellularly. In the extracellular space, type II collagen
could not be detected (Figure 3).
Analysis of the incorporation of tritiated proline showed a time-dependent
increase for cells on the bottom of the wells (Figure 4). An increased incorporation was also found in the explants,
but this was only apparent at the 1- and 2-week intervals (Figure 5).
Autoradiography of these explants showed that cells present in the mucoperichondrium
had incorporated the labeled amino acid. In addition, labeled cells were found
in the matrix. Some of the label proved to be incorporated in the extracellular
matrix (Figure 3).
The observations presented in this study indicate that mucoperichondrium
tissue explants can be used to wrap a collagenous matrix and that the cells
of this tissue survive. Our data show that a fraction of the cells migrate
into the matrix, whereas other cells migrate out of the tissue onto the bottom
of the wells. During the first few days of culturing, some cells die, probably
because of manipulation of the tissue. At later intervals, an increased thickness
of the perichondrial tissue was observed, indicating growth of the tissue.
This finding was in line with the incorporation data, which demonstrated that
protein synthesis continued during culturing. In none of the experiments were
cytotoxic effects of the matrices observed.
Migration of cells into matrices occurred only in those containing pores.
In the nonporous dental and cartilaginous matrices, no ingrowth of cells was
found. Our data suggest that a matrix with pores is essential to finally achieve
our goal: remodeling of an existing matrix into an implantable mold.
To prevent outgrowth of cells on the bottom of the wells and thereby,
perhaps, stimulate these cells to migrate into the pores of the matrix, experiments
were performed by culturing the mucoperichondrium-wrapped matrix on a grid
or on a low-affinity well plate. Under these conditions, however, an increased
migration into the matrix was not found.
The morphologic structure of the cells on the bottom of the wells changed
during prolonged culturing from an elongated fibroblastlike shape to a more
rounded one. The rounded shape of the cells may suggest that mucoperichondrial
cells differentiated into a chondrocyte phenotype. Staining of these cells
with an antibody against type II collagen supported this option. Type II collagen
could not be visualized in the extracellular space, however. We assume that
the collagen was secreted in the medium and not deposited on the bottom of
In conclusion, this model seems appropriate to culture mucoperichondrial
explants in combination with collagenous matrices. Cells migrate into the
pores of the matrix, survive, and synthesize matrix components. Actual formation
of cartilage has not been shown to occur, but adding growth factors to this
model may induce such an activity. With the use of mucoperichondrial cells
instead of matured chondrocytes, these cultured cells could probably be more
easily induced by growth factors to stimulate growth and differentiation to
produce cells closely related to chondrocytes.22-27
Accepted for publication November 1, 2000.
Corresponding author and reprints: Jeroen W. L. van Lange, MD, Department
of Otorhinolaryngology, Academical Medical Center, University of Amsterdam,
PO Box 22660, 1100 DD Amsterdam, the Netherlands (e-mail: firstname.lastname@example.org).