Bovine osteoblasts cultured on polyanionic collagen scaffolds: an

ultrastructural and immunocytochemical study

Patrcia da Luz Moreira,
1
Selma Candel aria Genari,
1,2
Gilberto Goissis,
3
Fernando Galembeck,
4
Yuehuei H. An,
5
Arnaldo Rodrigues Santos, Jr.
6
1
Departamento de Biologia Celular, Instituto de Biologia, UNICAMP, Campinas, Sa o Paulo, Brazil
2
Centro Estadual de Educac a o Tecnol ogica Paula Souza, Faculdade de Tecnologia de Bauru, Bauru, Sa o Paulo, Brazil
3
Biotech Biom edica Produtos M edicos e Odontol ogico Ltda ME (Produc a o), Sa o Carlos, Sa o Paulo, Brazil
4
Departamento de Fsico-Qu mica, Instituto de Qu mica, UNICAMP, Campinas, Sa o Paulo, Brazil
5
Orthopaedic Research Laboratoy, Medical University of South Carolina, Charleston, South Carolina
6
Centro de Cie ncias Naturais e Humanas, Universidade Federal do ABC, Santo Andr e, Sa o Paulo, Brazil
Received 25 January 2011; revised 16 May 2012; accepted 28 June 2012
Published online 15 September 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.32804
Abstract: Collagen is the most abundant protein in the body
and is also the most important component of the extracellular
matrix. Collagen has several advantages as a biomaterial such
as lack of toxicity, biocompatibility, biodegradability, and easy
reabsorption. In this study, we examined bovine osteoblasts
cultured on native or anionic collagen scaffolds prepared from
bovine pericardium after selective hydrolysis of glutamine and
asparagine side chain amides for periods from 24 (BP24) and
48 h (BP48). The cells were cultured in control and mineraliza-
tion medium at 37

C in the presence of 5% CO
2
. Transmission
and scanning electron microscopy, energy dispersive spectros-
copy, and an immunocytochemical marker were used for anal-
ysis. Cells with an irregular morphology forming a conuent
multilayer were observed on matrices kept in control medium.
Most of these cells presented a polygonal or elongated at-
tened morphology. Several spherical deposits of calcium crys-
tal associated with phosphorus were observed on the native
and BP48 matrices. Similar results were observed in samples
kept in control medium except with lower calcium/phosphorus
ratio. Vesicles actively expelled from the cell membrane were
also seen (do this vesicles corresponds to calcium/phosphorus
deposits). Osteocalcin was clearly visible on matrices kept in
mineralization medium and was more expression on the sur-
face of BP48 matrices. The results showed that anionic colla-
gen is able to support osteoblastic differentiation, regardless
of the medium used. Finally, the BP48 matrix promoted better
osteoblast differentiation than the native matrix. VC
2012 Wiley
Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater 101B: 18
27, 2013.
Key Words: anionic collagen, cell culture, osteoblasts, bioma-
terials, tissue engineering, cell differentiation
How to cite this article: Moreira PL, Genari SC,Goissis G, Galembeck F, An YH, Santos AR, Jr. 2013. Bovine osteoblasts cultured
on polyanionic collagen scaffolds: An ultrastructural and immunocytochemical study. J Biomed Mater Res Part B 2013:101B:18
27.
INTRODUCTION
Bone fractures result in the loss of the mechanical stability
of bone, bone tissue discontinuation, and partial destruction
of blood supply. Bone repair is a complex process that con-
sists of the stabilization of bone fragments, bone consolida-
tion, reconstruction of avascular and seminecrotic frag-
ments, and nally, internal and external remodeling of the
newly formed tissue.
1,2
External factors can markedly affect
the regeneration process, although tissues act according to
biological rules that control cell proliferation and differen-
tiation, as well as extracellular matrix production, which can
independently occur despite of external interference,
although being inuenced by them.
3
In contrast, fractures
accompanied by bone mass loss require grafts or implants.
Natural polymers such as collagen
46
and elastin
7,8
are bio-
materials that offer advantages for tissue engineering, for
example, the presence of the cellular recognition sequence
Arg-Gly-Asp (RGD) that inuences cell adhesion.
9
However,
in their pure form these materials are unable to re-establish
the initial tissue characteristics.
Collagen is the most abundant protein in the body and
is also the most important component of the extracellular
matrix. Collagen molecules determine the size, tensile
strength, and cellular arrangement of all structures and
organs.
911
Collagen has several advantages as a biomaterial
such as lack of toxicity, biocompatibility, biodegradability,
and easy reabsorption by the body,
1215
as well as low anti-
genicity, high tensile strength, and high afnity for water.
Some modications are necessary to overcome these
drawbacks, including an increase in the number of crosslinks
to improve tensile strength
1618
and the addition of other
materials,
1920
growth factors,
6,21,22
glycosaminoglycans,
23
Correspondence to: A. R. Santos, Jr. (arnaldo.santos@ufabc.edu.br)
18 VC 2012 WILEY PERIODICALS, INC.
and other molecules
8,24
to the collagen scaffold. These modi-
cations also increase the mechanical stability of collagen
while, at the same time, decreasing its biodegradability and
protecting the collagen structure from in vivo reabsorption. In
addition to increasing the negative charge and carboxylic
groups in the collagen molecule,
16,25
these modications
improve the piezoelectrical properties of collagen, which pro-
mote osteogenesis.
26
In vivo tests performed with these anionic matrices have
demonstrated bone formation and a low inammatory
response even in bones with established osteoporosis, with
promising characteristics for bone defect repair.
27,28
In con-
trast, the native collagen matrix caused a mild to intense
inammatory reaction associated with reabsorption cen-
ters.
28,29
Cell interactions with biomaterials, as well as the
quality of these interactions, are known to inuence the
ability of cells to proliferate and differentiate when in con-
tact with the implant.
21,22
These interactions are inuenced
by the topography of the material, supercial energy, and
roughness, among others.
3,6
Among many reports relating
negative charges with osteogenesis, there is a few in vitro
data that could help explain the effect of negative charges
on differentiation of bone cells. Some reports have shown
variations on osteoblast metabolism when in contact with
negative charges on biomaterial surface. The cellular inter-
actions that occur on these anionic collagen matrices are
not completely understood.
In the present study, we examined bovine osteoblasts
cultured on anionic collagen scaffolds prepared from bovine
pericardium submitted to alkaline hydrolysis for 24 (BP24)
or 48 h (BP48). Analysis of the mineralization process by
transmission electron microscopy (TEM), scanning electron
microscopy (SEM), and energy dispersive spectroscopy
(EDS), together with the evaluation of the morphological
and functional differentiation of osteoblastic cells on these
anionic collagen matrices as compared to native collagen
using an immunocytochemical marker, may provide useful
information for the understanding of the biological proc-
esses observed in in vivo experiments.
25,2729
MATERIALS AND METHODS
Matrix preparation
Bovine pericardium (BP; Braile Biomedica S/A, Sa o Jose do
Rio Preto, Brazil) was used for preparation of the 3D matrix.
The samples were prepared at the Sa o Carlos Institute of
Chemistry, Sa o Paulo University, as described by Lacerda
et al.
30
and Goissis et al.
31
Briey, alkaline hydrolysis was
performed in aqueous dimethylsulfoxide solution containing
chloride and sulfate salts of alkaline and alkaline earth met-
als. The matrices were hydrolyzed for 24 (BP24) and 48 h
(BP48) and the conditions were such that only amide from
asparagine and glutamine residues was hydrolyzed. Excess
salts were removed by extensive washing with 3% boric
acid solution, 0.3% EDTA, pH 11.0, and deionized water, and
nally, the matrices were equilibrated in 0.13 mol/L phos-
phate buffer, pH 7.4. The anionic collagen matrices produced
had been characterized by thermal analysis, infrared spec-
troscopy, titration, and determination of dielectric properties
as described elsewhere.
30,31
The following matrices were
analyzed: (1) 3D native anionic collagen/elastin matrices
(not submitted to alkaline hydrolysis treatment), and (2) 3D
anionic collagen matrices submitted to alkaline hydrolysis,
BP24 and BP48. After analysis, the matrices were cut into
small circles (6 mm diameter 3 mm thick) and used in
the experiments described below.
Cell culture
A noncontinuous fetal bovine osteoblast line kindly provided
by Dr. William Whitson
32
and maintained at the Orthopedic
Research Laboratory of the Medical University of South Caro-
lina was used in this study. The cells were used in all experi-
ments until passage 4. The osteoblasts were cultured in Whit-
sons DMEM medium containing 15% fetal calf serum (both
from Sigma Chemical, St. Louis, MO) at 37

C in a 5% CO
2
atmosphere.
32
For experimental analysis, Whitsons mineraliza-
tion medium
32
containing 15% fetal calf serum (both from
Sigma) was also used for culture of the noncontinuous cell line.
This medium differs from Whitsons DMEM by the addition of
b-glycerophosphate and calcium chloride. The cells were kept
in Whitsons DMEM medium until the time of the experiment,
when they were cultured in the two media as described earlier.
The medium was changed daily until the cells reached conu-
ence at about 57 days, when they were subcultured. The cells
were cultured on the native, BP24 and BP48 scaffolds in Whit-
sons DMEM medium (called control medium in this study) and
Whitsons mineralization medium (called mineralization me-
dium), both supplemented with 15% FCS, for 21 days at 37

C
in the presence of 5% CO
2
. Suspensions containing an initial
concentration of 1.0 10
5
cells/mL were added to a 96-well
plate (200 lL/well) (Corning/Costar Corporation, Cambridge,
MA) with the matrices. After 4 h of incubation for the purpose
of adhesion, the plate to be analyzed with the mineralization
medium had the DMEM medium replaced by the mineralization
one at certain time points, the matrices were processed for
analysis by the techniques described below.
Scanning electron microscopy and energy dispersive
spectroscopy (EDS)
After xation in 4% paraformaldehyde/2.5% glutaraldehyde
(Sigma) in 0.1M phosphate buffer (Merck), pH 7.4, for 30
min, the matrices (n 5 for each culture conditions used)
were washed with phosphate buffer, postxed in 1% osmium
tetroxide (Sigma), and dehydrated in an ethanol series. The
material was critical point dried (Balzers CPD030, Balzers,
Elgin, IL) and sputtered with gold (Balzers SCD 050). The
specimens were examined under a JEOL JSM-5800 LV (Jeol,
Tokyo, Japan) SEM equipped with an EDS analysis system.
Transmission electron microscopy
The specimens (n 5 for each culture conditions used)
were xed in 4% paraformaldehyde/2.5% glutaraldehyde
(Sigma) in 0.1M phosphate buffer (Merck), pH 7.4, for 30
min, postxed in 1% osmium tetroxide in the same buffer
at 4

C, dehydrated in acetone, and embedded in epoxy resin

(Embed-812, Electron Microscopy Sciences, EMS, WA). Ultra-
thin sections (2060 nm) were stained with 2% uranyl
ORIGINAL RESEARCH REPORT
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | JAN 2013 VOL 101B, ISSUE 1 19
acetate (EMS) for 20 min and 2% lead citrate (EMS) for 7
min prior to observation under a Leo 906 TEM (Carl Zeiss,
Oberkochen, Baden-Wu rttemberg, Germany).
Immunocytochemical detection of osteocalcin
After xation in 4% paraformaldehyde (Sigma) in 0.1M phos-
phate buffer (Merck), pH 7.4, for 30 min, the matrices (n 4
for each culture conditions used) were washed with phos-
phate buffer and embedded in paraplast (Fisher Scientic, Vet-
erans Memorial Drive, Houston, TX). The blocks were cut into
5 lm were made 3 or 4 sections by slide and four slides have
been produced by sample studied. Thick sections and the
specimens were cleared using routine procedures. The sec-
tions were then washed in cold PBS (Nutricell Nutrientes
Celulares, Campinas, Brazil) and nonspecic binding sites
were inactivated by incubation in 1% BSA (Sigma)/PBS
(Nutricell) for 60 min in a humid chamber. After washing in
cold PBS, the primary rabbit antihuman osteocalcin antibody
(BT-593, Biomedical Technologies, Stoughton, MA) diluted
1:30 in PBS and 1% BSA was added and the sections were
incubated overnight in the dark in a humid chamber. After
BSA gentle removal with cold PBS, the secondary antibody
(FITC-conjugated goat antirabbit IgG, BT-557, Biomedical
Technologies, Stoughtou, MA) diluted 1:50 in PBS was added
and the sections were incubated for 1 h in the dark. Negative
controls consisted of omission of the primary antibody were
analyzed all elds in each slide. The same observer analyzed
all samples. Analysis and photographic documentation were
performed with an Olympus IX50 microscope (Olympus,
Tokyo, Japan) equipped with a FITC lter.
RESULTS
Scanning electron microscopy and energy dispersive
spectroscopy
Cells with an irregular morphology forming a conuent mul-
tilayer were observed on matrices kept in control medium
FIGURE 1. (A) Scanning electron photomicrographs of osteoblasts cultured on the native, (B) BP24, and (C) BP48 matrices in control (A1C1)
and mineralization medium (A2C2). In detail, spots of spherical mineral deposits on BP48 matrices. Scale bar 2.5 lm.
20 MOREIRA ET AL. BOVINE OSTEOBLASTS CULTURED ON NATIVE OR ANIONIC COLLAGEN SCAFFOLDS
[Figure 1 (A1C1)]. Most of these cells presented a polygo-
nal or elongated attened morphology maintained through-
out the culture period. A small number of vesicles and/or
microvilli were noted on the cell surface. The cells also pre-
sented evident lopodia and lamellipodia. Multilayered po-
lygonal/elongated conuent cells rich in vesicles and/or mi-
crovilli were observed on matrices kept in mineralization
medium [Figure (1A2C2)]. The morphological results
observed were representative and reproductive through the
different groups studied. Several spots of spherical calcium
crystal deposits were noted on the native and BP48 matri-
ces [Figure (1A2C2)]. Identication and mapping of the
elements by EDS analysis suggested the concomitant pres-
ence of P and Ca in the same regions in all matrices
(Figures 2 and 3). Samples kept in control medium pre-
sented different amounts of Ca and P [Table I and Figure
(2A1C1)]. Mapping revealed superposition of these ele-
ments [Figure (2A2C2)]. A Ca/P ratio of 0.27, 0.10, and
0.05 was obtained for native collagen, BP24, and BP48,
respectively (Figure 2). On the other hand, differences in Ca
and P deposition were observed for matrices kept in miner-
alization medium (Table I and Figure (3A1C1)], but super-
position of the elements was the same [Figure (3A2C2)].
The Ca/P ratio was 1.46, 0.10, and 2.26 for BP48, BP24,
and native collagen, respectively (Figure 3). The atomic con-
centration (%) of some elements in the different experimen-
tal conditions can be seen in Table I. Using the medium of
mineralization, the P concentration in samples BP24 and
BP48 were always higher than in native collagen. In the
same culture condition, the Ca concentration was always
greater in BP48 than other samples.
Transmission electron microscopy
All matrices presented a similar ultrastructure when cul-
tured in control [Figure 4(A)] and mineralization [Figure
4(B)] medium. The collagen matrices consisted of bers
arranged in layers that tended to run perpendicular to the
section [Figure 4(A)]. At higher magnications, when a par-
allel plane of the section was obtained, collagen showed a
regular banding pattern. When mineralization nutritional
FIGURE 2. EDS analysis of matrices cultured in control medium: (A) native, (B) BP24, and (C) BP48. A1 to C1 show the EDS spectra and calcium/
phosphorous ratio. Note the absence of Ca
2
peaks. A2 to C2 show the element mapping of calcium (Ca), phosphorous (P), and oxygen (O).
[Color gure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
ORIGINAL RESEARCH REPORT
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | JAN 2013 VOL 101B, ISSUE 1 21
conditions were supplied, accumulation of electron dense
material was observed on the matrices [Figure 4(B), aster-
isk]. Cells were mainly seen on the surface of the matrices
[Figure 4(C,D)] and presented several lopodia penetrating
the matrices [Figure 4(F)] perpendicular to the plane of the
section. No chromatin condensation was observed [Figure
4(C)], a nding suggesting the presence of an active nucleus
with vesicles actively expelled from the plasma membrane
FIGURE 3. EDS analysis of matrices cultured in mineralization medium: (A) native, (B) BP24, and (C) BP48. A1 to C1 show the EDS spectra and
calcium/phosphorous ratio. Note the Ca
2
peaks on BP48. A2 to C2 show the element mapping of calcium (Ca), phosphorous (P), and oxygen
(O). Observe the same position of Ca and P on BP48. [Color gure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
TABLE I. Atomic Concentration (%) of Some Elements in the Different Samples
Elements
Samples
Native BP24 BP48
C M C M C M
Carbon (C) 54.94 20.39 41.80 46.98 74.91 34.15
Oxygen (O) 15.20 4.90 5.00 8.70 12.07 15.86
Phosphorus (P) 0.40 2.16 1.17 2.94 2.10 12.47
Calcium (Ca) 0.11 4.90 0.12 0.32 0.12 18.26
Gold (Au) 7.19 50.76 44.80 33.65 8.08 17.89
C: control medium; M: mineralization medium.
22 MOREIRA ET AL. BOVINE OSTEOBLASTS CULTURED ON NATIVE OR ANIONIC COLLAGEN SCAFFOLDS
[Figure 4(C,E,F) arrow], and indication of secretory activity
of the cells.
Immunocytochemical detection of osteocalcin
For matrices kept in control medium [Figure 5(A,C,E)],
osteocalcin was clearly visible on the surface of the BP48
matrix [Figure 5(E)], although no signicant Ca or P deposi-
tion was detected by SEM or EDS (Figures 13, and Table
I). For matrices kept in mineralization medium, osteocalcin
expression was observed near the surface of all matrices
[Figure 5(B,D,F,G)], being more evident on the BP48 matrix
[Figure 5(F,G)]. It was made only qualitative analysis of the
results obtained in this experiment.
DISCUSSION
In the present study, osteoblasts proliferated on all matrices,
forming multilayers that could be seen by SEM and light mi-
croscopy as previously reported by our group.
33
On day 5,
osteoblasts formed multilayers on the native and BP48 mat-
rices (data not shown). However, on day 10, these charac-
teristics were no longer visible and all matrices promoted
osteoblast proliferation into multilayers. Bet et al.
34
FIGURE 4. Transmission electron photomicrographs of the matrices. (A) BP48 matrices cultured in control medium. Note that almost all collagen
bers are arranged in the same direction. Scale bar 5.6 lm. (B) Native matrices cultured in mineralization medium. Note the electron-dense
deposits (*) throughout the matrices. Scale bar 4 lm. (C) Osteoblasts cultured on BP48 matrices in mineralization medium. Note the nucleus
(n) with uncondensed chromatin and the formation of vesicles (!) within the cytoplasmic membrane. Scale bar 2 lm. (D) Cross-section of an
osteoblast cultured on the BP24 matrix in control medium. Scale bar 3 lm. (E) Detail of the cytoplasmic membrane of an osteoblast cultured
on native matrices in control medium. Note the formation of vesicles (!) close to some actin (ac) laments. Scale bar 435 nm. (F) Pseudopo-
dium of an osteoblast cultured on the BP48 matrix in common medium and formation of vesicles (!). Scale bar 1.2 lm.
ORIGINAL RESEARCH REPORT
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | JAN 2013 VOL 101B, ISSUE 1 23
proposed that new motifs described as the minimal active
recognition sequence for integrins binding of broblasts and
other cells with type I collagen would arise as a result of
Asn and Gln carboxyamide side-chain hydrolysis. Other
reports have shown that osteoblasts cultured on substrates
that presented both negative and positive charges may dem-
onstrate characteristics compatible with an active metabo-
lism when in contact with negative charges, while they are
inactive when in contact with positive charges.
35
There are many reports correlating negative charges
with osteogenesis. Krukowski et al.
36
demonstrated that
negatively charged beads stimulated bone formation at one
month after implantation in rats. Bone formation was
observed on negatively charged sephadex beads, whereas
defects receiving positively charged beads were lled with
dense connective tissue. Lately, the osteogenic response to
subperiosteal injection of negatively charged ion exchange
resins was compared in the tibiae of one month and 1622
month old rats. The resins were administered either in the
form of beads (CM Sephadex) or as particles (CM cellulose,
carboxymethylcellulose). Periosteal bone formation was
observed wherever resin was in contact with bone, and in
the resin bed the connective tissues that invested the
charged materials ossied within the rst month. Marrow
spaces commonly formed where periosteal growth was
most rapid. The osteogenic effect was independent of resin
conformation, and it was more pronounced in the younger
rats.
37
In other article, it was tested the hypothesis that posi-
tively charged dextran resin (PCDR) could inhibit bone
repair in vivo. Central physeal defects were created. The
defects in left tibiae were packed with neutral resin (con-
trol); those in right tibiae were lled with PCDR. The PCDR-
lled defects showed a signicant decrease in trabecular
bone formation as early as the 2nd week. By the 10th post-
operative week, formation of trabeculae had been reduced
by nearly 40%.
38
Chierico et al.
39
studied the effects on neutral, nega-
tively, and positively charged titanium reinforced gore tex
augmentation material (GTAM) membranes. Those authors
found a more rapid and increased bone neogenesis with the
negatively charged domes. Over time studied, negatively
charge membranes supported more new bone formation
than neutral membranes while positively charged mem-
branes showed a lest new bone. Kobayashi et al.
40
showed
that negatively charged surface of the electrically polarized
FIGURE 5. Immunocytochemical detection of osteocalcin on native (A, B), BP24 (C, D) and BP48 (E, F, and G) matrices cultured in control (A, C,
and E) and mineralization (B, D, F, and G) medium. Scale bar 50 lm. [Color gure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
24 MOREIRA ET AL. BOVINE OSTEOBLASTS CULTURED ON NATIVE OR ANIONIC COLLAGEN SCAFFOLDS
HA signicantly promoted direct bone bonding as well as
new bone formation. The difference in osteobonding and
cell response is attributed to the magnitude of the charge
and the atomic arrangement in the surface between the
polarized HA and the charged polymers. Sautier et al.
41
sug-
gested that organic matrices such as polyanionic proteins,
possibly adsorbed on bone-conductive materials, could have
a Ca-binding property and become the foothold for minerali-
zation. These data indicates that polyanionic collagen might
be useful in implants aimed at bone regeneration.
On the other hand, there is a few in vitro data that could
help explain the effect of negative charges on differentiation
of bone cells. During osteoblast differentiation, cell activity
seems to be more involved in metabolism related to prolif-
eration in order to create a three-dimensional microenviron-
ment, secondarily allowing cell differentiation expression.
42
At the end of this stage when a permissive microenviron-
ment was created, cells characterized by morphological
modications and alkaline phosphatase (ALP) activity as
previously reported by our group
33
start to express noncol-
lagen proteins such as osteocalcin. Osteocalcin, also known
as bone Gla protein, is a vitamin K-dependent calcium-bind-
ing protein and a major component of the noncollagenous
bone matrix.
43
This protein is synthesized by active mature
osteoblasts in bone, with a small amount being produced by
odontoblasts in dentin and by hypertrophic chondrocytes.
43
The function of osteocalcin is unclear, but it is thought to be
related to bone mineralization because the protein speci-
cally interacts with calcium from HA, probably regulating
the growth of HA crystals. In the present study, cells grown
near the surface of the matrices were positive for osteocal-
cin production despite diffuse staining on the native and
BP24 matrices.
Expression of osteocalcin was observed on all matrices
in the same region as the mineralization front previously
identied by von Kossa staining.
33
This observation might
be explained by the afnity of osteocalcin for HA and the
presence of the mineral as a stabilizing element of this non-
collagen protein. The relatively faint staining observed on
the BP24 matrix may be related to the fact that part of the
protein released by the cells is degraded in the extracellular
compartment unless it is stabilized in a dened conforma-
tion with other proteins or mineral elements in the mineral-
ized matrix. Morphological analysis by SEM and TEM
revealed the presence of vesicles on the osteoblast surface
and uncondensed chromatin, ndings suggesting secretory
activity of these osteoblasts. Although these vesicles were
also seen in osteoblasts of matrices kept in control medium,
they were much more evident and abundant in matrices
kept in mineralization medium.
In vivo, osteocalcin expression is normally associated
with the nal stages of osteoblast differentiation and partic-
ularly identies condensed groups of cells actively laying
down and mineralizing matrix. But, in a recent review, it
was demonstrated marked differences in osteoinduction
mediated by natural conditions, such as BMP and biomateri-
als.
44
In vitro, low porosity stimulates osteogenesis, sup-
pressing cell proliferation by forcing cell aggregation. On the
other hand, in vivo higher porosity with larger pores pro-
motes bone growth. However, these factors results in low
mechanical properties and, therefore, functional limit exists
for porosity and pore size.
45
On the basis of these studies, a
minimum pore size of 100 lm was established, consider-
ing cell size, migration requirements, and transport. In vitro
and in vivo results suggest pore size >300 lm to facilitate
vascularization of grafts. Rocha et al.
25
reported that BP24
presents pore sizes from 12 to 105 lm, while BP48 has
pores between 14 and 134 lm. Thus, it is possible that the
scaffolds could stimulate an accelerated osteogenesis with
nodules in the advanced stages of differentiation near
regions where differentiation is still in the earliest phase.
Anionic collagen can be mineralized in vitro when incu-
bated with a saturated calcium solution,
46
but this experi-
mental condition differs quite from that used in this study.
In vivo, implanted polyanionic collagen matrices are directly
biomineralized by osteoblasts as shown by ultrastructural,
histoenzymological, and morphological analysis.
47
Polya-
nionic collagen may act as a cofactor for ALP secreted by
osteoblasts, facilitating mineralization of the scaffold. More-
over, the presence of ALP excludes the possibility that min-
eralization occurred by dystrophic calcication.
48
Evidence indicates that different forms of calcium phos-
phate are related to osteogenic cell differentiation. Shu
et al.
49
observed a 2.3-fold increase of ALP activity in
MC3T3-E1 preosteoblasts grown on HA as compared to the
culture plate. A signicant increase of specic ALP activity
was demonstrated for osteogenically induced human bone
marrow stem cells (hBMSC) cultured on calcium phosphate
cement.
50
Investigating hBMSC cultured on collagen type I
with nanocrystalline HA, Bernhardt et al.
51
reported attenu-
ated proliferation rates when compared to cells grown on
tissue culture polystyrene, but markedly increased activity
of ALP on the mineralized collagen, indicating its promoting
effect on the osteogenic differentiation of hBMSC. Thus, the
calcium deposition on anionic collagen scaffolds observed in
this study may restimulate the osteogenic differentiation of
cells through an autocrine signaling pathway.
EDS analysis of the mineralized matrices showed a Ca/P
ratio of 1.46 for the BP48 matrix. Although the HA charac-
terization could not be made only by EDS, a possible inter-
pretation of our results is that, on anionic collagen matrix,
the osteoblasts have formed a structure similar to poorly
crystallized form of HA similar to that found in mineralized
fetal bone.
52
A Ca/P ratio higher than 1.67, characteristic of
bovine bone HA,
53
was only observed for the native matrix
kept in mineralization medium. Some studies represented
the overall range of bioapatite chemistry as (Ca, Na, Mg, K,
Sr, Pb, . . .)
10
(PO
4
, CO
3
, SO
4
, . . .)
6
(OH, F, Cl, CO
3
)
2
. The multi-
ple substitutions shown above also explain why/how typical
bioapatites have lower Ca:P atomic ratios than the 1.67
value for stoichiometric hydroxylapatite.
54,55
The anionic collagen has been tested in vitro and in vivo.
In vitro, it was found that anionic collagen matrices are ca-
pable of maintaining the osteoblasts differentiation pattern.
In vivo, in experiments conducted with mice, the data in the
literature show a good osseointegration. So far we have no
ORIGINAL RESEARCH REPORT
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | JAN 2013 VOL 101B, ISSUE 1 25
data on other experimental animals such as rabbits or dogs.
There are also no clinical data on the use of anionic collagen
matrices. However, the use of such material seems
promising.
In summary, the present study showed the sequential
differentiation of bovine osteoblasts on polyanionic collagen
scaffolds. The results showed that these anionic collagen
matrices are able to support osteoblastic proliferation until
conuence and stratication are reached, regardless of the
medium used. However, SEM and EDS analysis demon-
strated that the native and BP48 matrices kept in minerali-
zation medium increased calcium phosphate deposition, a
fact that may explain the higher expression of osteocalcin
on these matrices, especially BP48. Finally, the BP48 matrix
seems to be more adequate for bone repair since it pro-
moted better osteoblast differentiation and bone nodule for-
mation as compared to the native matrix. The latter is char-
acterized by poor tensile strength and piezoelectrical
properties.
REFERENCES
1. Chao EYS, Inoue N. Biophysical stimulation of bone fracture
repair, regeneration and remodeling. Eur Cells Mater 2003;6:
7285.
2. Shapiro F. Bone development and its relation to fracture repair.
The role of mesenchymal osteoblasts and surface osteoblasts.
Eur Cells Mater 2008;15:5376.
3. Santos AR Jr, Lombello CB, Genari SC. Technologies applied to
stimulate bone regeneration. In: Davies J, editor. Tissue Regener-
ationFrom Basic Biology to Clinical Application. Rijeka: InTech;
2012. pp. 339366.
4. Chevallay B, Herbage D. Collagen-based biomaterials as 3D scaf-
fold for cell cultures: Applications for tissue engineering and gene
therapy. Med Biol Eng Comput 2000;38:211218.
5. Lee CH, Singla A, Lee Y. Biomedical applications of collagen. Int J
Pharm 2001;221:122.
6. Reddi AH. Morphogenesis and tissue engineering of bone and
cartilage: Inductive signals, stem cells, and biomimetic biomateri-
als. Tissue Eng 2000;6:351359.
7. Hafemann B, Ensslen S, Erdmann C, Niedballa R, Zuhlke A, Gho-
frani K, Kirkipatrick CJ. Use of a collagen/elastin membrane for
the tissue engineering of dermis. Burns 1999;25:373384.
8. Lefebvre F, Gorecki S, Bareille R, Amedee J, Bordenave L, Rabaud
M. New articial connective matrix-like structure made of elastin
solubilized peptides and collagens: Elaboration, biochemical and
structural properties. Biomaterials 1992;13:2833.
9. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Mo-
lecular Biology of the Cell, 5th ed. New York: Garland Science;
2008.
10. Burgeson RE, Nimni ME. Collagen types. Molecular structure and
tissue distribution. Clin Orthop 1992;282:250272.
11. Haralson MA, Hassel JR. The extracellular matrixAn overview.
In: Haralson MA, Hassel JR, editors. Extracellular Matrix: A practi-
cal approach. The Practical Approach Series. New York: Oxford
University Press; 1995. pp. 130.
12. Friess W. CollagenBiomaterial for drug delivery. Eur J Pharm
Biopharm 1998;45:113136.
13. Maeda M, Tani S, Sano A, Fujioka K. Microstructure and release
characteristics of the minipellet, a collagen-based drug delivery
system for controlled release of protein drugs. J Control Release
1999;62:313324.
14. Sano A, Hojo T, Maeda M, Fujioka K. Protein release from colla-
gen matrices. Adv Drug Deliv Rev 1998;31:247266.
15. Werkmeister JA, Ramshaw JA. Collagen-based biomaterials. Clin
Mater 1992;9:137138.
16. Goissis G, Marcantonio E Jr, Marcantonio RA, Lia RC, Cancian
DC, de Carvalho WM. Biocompatibility studies of anionic collagen
membranes with different degree of glutaraldehyde cross-linking.
Biomaterials 1999;20:2734.
17. Singh MP, Stefko J, Lumpkin JA, Rosenblatt J. The effect of elec-
trostatic charge interactions on release rates of gentamicin from
collagen matrices. Pharm Res 1995;12:12051210.
18. Suh H, Park JC, Han DW, Lee DH, Han CD. A bone replaceable ar-
ticial bone substitute: Cytotoxicity, cell adhesion, proliferation,
and alkaline phosphatase activity. Artif Organs 2001;25:1421.
19. Takaoka K, Nakahara H, Yoshikawa H, Masuhara K, Tsuda T, Ono
K. Ectopic bone induction on and in porous hydroxyapatite com-
bined with collagen and bone morphogenetic protein. Clin Orthop
1988;234:250254.
20. Tan W, Krishnaraj R, Desai TA. Evaluation of nanostructured com-
posite collagen-chitosan matrices for tissue engineering. Tissue
Eng 2001;7:203210.
21. Kimura M, Zhao M, Zellin G, Linde A. Bone-inductive efcacy of
recombinant human bone morphogenetic protein-2 expressed in
Escherichia coli: An experimental study in rat mandibular defects.
Scand J Plast Reconstr Surg Hand Surg 2000;34:289299.
22. Zellin G, Hedner E, Linde A. Bone regeneration by a combination
of osteopromotive membranes with different BMP preparations:
A review. Connect Tissue Res 1996;35:279284.
23. van Susante JLC, Pieper J, Buma P, van Kuppevelt TH, van Beu-
ningen H, van Der Kraan PM, Veerkamp JH, van den Berg WB,
Veth RPH. Linkage of chondroitin-sulfate to type I collagen scaf-
folds stimulates the bioactivity of seeded chondrocytes in vitro.
Biomaterials 2001;22:23592369.
24. Doillon CJ, Silver FH. Collagen-based wound dressing: Effects of hyal-
uronic acid and bronectin on wound healing. Biomaterials 1986;7:38.
25. Rocha LB, Goissis G, Rossi MA. Biocompatibility of anionic collagen
matrix as scaffold for bone healing. Biomaterials 2002;23:449456.
26. Kryszewski M. Fifty years of study of the piezoelectric properties
of macromolecular structured biological materials. Acta Physica
Polonica A 2004;4:389408.
27. Rosa FP, Lia RCC, Souza KOF, Goissis G, Marcantonio E Jr. Tissue
response to polyanionic collagen:elastin matrices implanted in rat
calvaria. Biomaterials 2003;24:207212.
28. Rocha LB, Brochi MA, Bellucci AD, Rossi MA. Efcacy of polya-
nionic collagen matrices for bone defect healing. J Biomed Mater
Res B Appl Biomater 2004;71:355359.
29. Cunha MR, Santos AR Jr, Goissis G, Genari SC. Implants of poly-
anionic collagen matrix in bone defects of ovariectomized rat. J
Mater Sci Mater Med 2008;19:13411348.
30. Lacerda C, Plepis AMG, Goissis G. Selective hydrolysis of carbox-
amides of asparagine and glutamine residues of collagen: prepa-
ration and characterization of anionic collagen matrices for
biomaterial applications. Quim Nova 1998;21:267271.
31. Goissis G, Piccirili L, Goes J, Plepis A, Das-Gupta D. Anionic colla-
gen: Polymer composites with improved dielectric and rheologi-
cal properties. Artif Organs 1998;22:203209.
32. Whitson SW, Whitson MA, Bowers DE Jr, Falk MC. Factors inu-
encing synthesis and mineralization of bone matrix from fetal bo-
vine cells grown in vitro. J Bone Mineral Res 1992;7:727741.
33. Moreira PL, An YH, Santos AR Jr, Genari SC. In Vitro analysis of
anionic collagen scaffolds for bone repair. J Biomed Mater Res B
Appl Biomater 2004;71B:229237.
34. Bet M, Goissis G, Vargas S, Selistre-de-Araujo H. Cell adhesion and
cytotoxicity studies over polyanionic collagen surfaces with variable
negative charge and wettability. Biomaterials 2003;24:131137.
35. Davies JE, Causton B, Bovell Y, Davy K, Sturt CS. The migration
of osteoblasts over substrata of discrete surface charge. Biomate-
rials 1986;7:231233.
36. Krukowski M, Shively RA, Osdoby P, Eppley BL. Stimulation of
craniofacial and intramedullary bone formation by negatively
charged beads. J Oral Maxillofac Surg 1990;48:468.
37. Krukowski M, Snyders RV Jr, Eppley BL, Simmons DJ. Negatively
charged resins stimulate bone formation in subperiosteal site in
rats. Clin Orthop Relat Res 1994;298:266271.
38. Cobos JA, Yang J, Zhang R, Krukowski M, Simmons DJ. Posi-
tively charged dextran resin inhibits trabecular bone repair in the
rabbit tibial physis. J Biomed Mater Res 1998;39:458461.
39. Chierico A, Valentini R, Majzoub Z, Piattelli A, Scarano A, Okun L,
Cordioli G. Electrically charged GTAM membranes stimulate
26 MOREIRA ET AL. BOVINE OSTEOBLASTS CULTURED ON NATIVE OR ANIONIC COLLAGEN SCAFFOLDS
osteogenesis in rabbit calvarial defects. Clin Oral Impl Res 1999;
10:415424.
40. Kobayashi T, Nakamura S, Yamashita K. Enhanced osteobonding
by negative surface charges of electrically polarized hydroxyapa-
tite. J Biomed Mater Res 2001;57:477484.
41. Sautier JM, Nefussi JR, Forest N. Mineralization and bone forma-
tion on microcarrier beads with isolated rat calvaria cell popula-
tion. Calcif Tiss Int 1992;50:527532.
42. Nefussi JR, Brami G, Modrowski D, Obuf M, Forest N. Sequen-
tial expression of bone matrix proteins during rat calvaria osteo-
blast differentiation and bone nodule formation in vitro. J
Histochem Cytochem 1997;45:493503.
43. Gundberg CM. Matrix proteins. Osteoporos Int 2003;14:3742.
44. Habibovic P, de Groot K. Osteoinductive biomaterialsProperties
and relevance in bone repair. J Tissue Eng Regen Med 2007;1:2532.
45. Hutmacher DW, Schantz JT, Lam CXF, Tan KC, Lim TC. State of the
art and future directions of scaffold-based bone engineering from a
biomaterials perspective. J Tissue Eng Regen Med 2007;1:245260.
46. Goissis G, Maginador SVS, Martins VCA. Biomimetic mineraliza-
tion of charged collagen matrices: In vitro and in vivo study. Artif
Organs 2003;27:437443.
47. Rocha LB, Adam RL, Leite NJ, Metze K, Rossi MA. Biomineraliza-
tion of polyanionic collagen-elastin matrices during calvarial bone
repair. J Biomed Mater Res 2006;79A:237245.
48. Nimni ME, Bernick S, Cheung DT, Ertl DC, Nishimoto SK, Paule
WJ, Salka C, Strates BS. Biochemical differences between dystro-
phic calcication of cross-linked collagen implants and mineraliza-
tion during bone induction. Calcif Tissue Int 1988;42:313320.
49. Shu R, McMullen R, Baumann MJ, Mccabe LR. Hydroxyapatite
accelerates differentiation and suppresses growth of MC3T3-E1
osteoblasts. J Biomed Mater Res A 2003;67:11961204.
50. Oreffo RO, Driessens FC, Planell, JA, Triftt JT. Growth and differ-
entiation of human bone marrow osteoprogenitors on novel cal-
cium phosphate cements. Biomaterials 1998;19:18451854.
51. Bernhardt A, Lode A, Boxberger S, Pompe W, Gelinsky M. Miner-
alised collagen-an articial, extracellular bone matrix-improves
osteogenic differentiation of bone marrow stromal cells. J Mater
Sci Mater Med 2008;19:269275.
52. Glimcher MJ. Recent studies of the mineral phase in bone and its
possible linkage to the organic matrix by protein-bound phos-
phate bonds. Philos Trans R Soc Lond (Biol) 1984;304:479508.
53. Joschek S, Nies B, Krotz R, Gopferich A. Chemical and physico-
chemical characterization of porous hydroxyapatite ceramics
made of natural bone. Biomaterials 2000;21:16451658.
54. Skinner HCW. Biominerals. Mineralogical magazine 2005;69:
621664.
55. Pasteris JD, Wopenka B, Valsami-Jones E. Bone and Tooth Miner-
alization: Why apatite? Elements 2008;4:97104.
ORIGINAL RESEARCH REPORT
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | JAN 2013 VOL 101B, ISSUE 1 27