Effect of blasting treatment and Fn coating on MG63 adhesion
and differentiation on titanium: a gene expression study using
M. Pegueroles•A. Aguirre•E. Engel•
G. Pavon•F. J. Gil•J. A. Planell•
V. Migonney•C. Aparicio
Received: 30 June 2009/Accepted: 4 January 2011/Published online: 22 January 2011
? Springer Science+Business Media, LLC 2011
in the adsorbed protein layer, and the presence of specific
functional groups can influence integrin binding specificity,
thereby modulating cell adhesion and differentiation pro-
cesses. The adsorption of fibronectin, a protein directly
involved in osteoblast adhesion to the extracellular matrix,
has been related to different physical and chemical prop-
erties of biomaterial surfaces. This study used blasting
particles of different sizes and chemical compositions to
evaluate the response of MG63 osteoblast-like cells on
smooth and blasted titanium surfaces, with and without
fibronectin coatings, by means of real-time reverse
Biomaterial surface properties, via alterations
transcription-polymerase chain reaction (qRT-PCR) assays.
This response included (a) expression of the a5, avand a3
integrin subunits, which can bind to fibronectin through the
RGD binding site, and (b) expression of alkaline phos-
phatase (ALP) and osteocalcin (OC) as cell-differentiation
markers. ALP activity and synthesis of OC were also tes-
ted. Cells on SiC-blasted Ti surfaces expressed higher
amounts of the a5mRNA gene than cells on Al2O3-blasted
Ti surfaces. This may be related to the fact that SiC-blasted
surfaces adsorbed higher amounts of fibronectin due to
their higher surface free energy and therefore provided a
higher number of specific cell-binding sites. Fn-coated Ti
M. Pegueroles ? E. Engel ? F. J. Gil ? J. A. Planell ? C. Aparicio
Biomaterials, Biomechanics, and Tissue Engineering Group,
Department of Material Science and Metallurgy, Universitat
Polite `cnica de Catalunya, Av. Diagonal 647,
08028 Barcelona, Spain
F. J. Gil
J. A. Planell
e-mail: email@example.com; firstname.lastname@example.org
A. Aguirre ? E. Engel ? J. A. Planell
Institute for Bioengineering of Catalonia (IBEC), Baldiri Reixac
13, 08028 Barcelona, Spain
G. Pavon ? V. Migonney
Laboratoire des Biomate ´riaux et Polyme `res de Spe ´cialite ´—UMR
CNRS 7052, Institut Galile ´e, 99 Avenue J.B. Cle ´ment,
93430 Villetaneuse, France
Laboratoire de Bio-inge ´nierie de Polyme `res Cardiovasculaires—
INSERM U698, Institut Galile ´e, Ba ˆtiment E, Universite ´ Paris 13
Nord, 99 Avenue J.B. Cle ´ment, 93430 Villetaneuse, France
C. Aparicio (&)
Department of Restorative Sciences, Minnesota Dental Research
Center for Biomaterials and Biomechanics (MDRCBB), School
of Dentistry, University of Minnesota, 16-212 Moos Tower,
515 Delaware St. S.E., Minneapolis, MN 55455, USA
J Mater Sci: Mater Med (2011) 22:617–627
surfaces decreased a5mRNA gene expression, by favoring
the formation of other integrins involved in adhesion over
a5b1. The changes in a5mRNA expression induced by the
presence of fibronectin coatings may moreover influence
the osteoblast differentiation pathway, as fibronectin coat-
ings on Ti surfaces also decreased both ALP mRNA
expression and ALP activity after 14 and 21 days of cell
The biological processes that take place when cells come
into contact with a material surface include initial focal
adhesion formation, cell adhesion, and extracellular matrix
formation and reorganization. The quality of the cell–
material interaction influences cells’ capacity to proliferate
and differentiate . Cells indirectly bind to surfaces
producing cytoskeletal associated transmembrane receptors
that attach to specific extracellular proteins , such as
fibronectin and vitronectin . Integrins are the most
important transmembrane receptors in cell binding pro-
cesses [4, 5]. Biomaterial surface properties, via alterations
in adsorbed protein structure, and the presence of specific
functional groups may influence integrin binding specific-
ity, thereby modulating signaling and expression of dif-
ferentiated phenotypes [6–10].
The use of human cell lines such as MG63 offers a
useful tool for investigating the effects of biomaterials and
for identifying the mechanisms of cell response. To
understand the mechanisms of bone formation on the
implant surface, the material’s effects on the surrounding
cells must be analysed. In vitro studies support the
hypothesis that implant surface may directly affect osteo-
blast migration, attachment, proliferation, and differentia-
tion, as increased surface roughness has been shown to
stimulate osteogenesis [11–13]. In addition to having cel-
lular characteristics associated with osteogenic differenti-
ation, surface roughness has also been shown to alter
integrin and growth factor expression .
MG63 osteoblast-like osteosarcoma cells cultured on
Ti of varying roughnesses have shown different osteo-
blastic differentiation expression depending on the degree
of roughness, and Ravalues of micro-rough implant sur-
faces ranging from 3 to 5 lm increased in vitro and in
vivo cell response [12, 15]. Cells cultured on Ti discs
with an average roughness[6 lm showed a reduced
capacity to differentiate compared with cells cultured on
smooth (\0.1 lm) surfaces. This is because the increase
in real surface area implies a lower cell confluence at the
time of measurement, resulting in reduced ALP activity
. However, the effect of surface roughness on cells
may be the result of the topographical/surface roughness
characteristics of the surface itself, as well as of the layer
of proteins and other biomolecules adsorbed on the sur-
face. The latter is also influenced by the surface charac-
teristics, which emerge as the material surface is
conditioned by the media and serum. This initial inter-
action produces a layer of macromolecules that directly
modulate cell response.
Fibronectin is an adhesive protein that mediates cell
adhesion and the only protein that activates the a5b1inte-
grins of cells [2, 17]. Thus, a5is an integrin subunit specific
for fibronectin protein . Several studies have demon-
strated that on Ti and Ti-alloy surfaces, osteoblasts express
a5integrin subunits, among others [17, 18]. The effect of
fibronectin coating has been added as a variable in this
study, since human fibronectin is a glycoprotein that pro-
motes cell attachment  to the biomaterial surface
through its central-binding domain arginine–glycine–
aspartic acid (RGD) sequence. The RGD sequence is
present in a variety of adhesive proteins (e.g. fibronectin
and vitronectin) and recognized by several integrins.
Specifically, integrins a5b1and avb3bind to and compete
for the RGD site on Fn [20, 21]. Moreover, a3b1can also
bind to fibronectin through the RGD binding site , and
a3is one of the most abundant a subunits expressed in
unstimulated MG63 cells .
Previously in our group, an optimized shot-blasted
titanium (c.p. Ti) dental implant was developed . The
optimal roughness and appropriate abrasive particles for a
better in vitro response and earlier in vivo osseointegration
were determined . However, the first biological events
leading to this successful response and the exact influence
of the surface properties are still poorly understood. To
address these questions, first, a thorough topographical and
surface free energy characterization of the optimized sur-
faces was carried out . Studies of fibronectin adsorption
on the different roughened surfaces revealed differences in
both the adsorbed fibronectin and the spatial organization
of the extracellular matrix depending on specific topo-
graphical and physico-chemical properties of the tested
As a further step, in this work we test the MG63
osteoblast-like cells response on smooth and blasted tita-
nium surfaces—different sizes and chemical composition
of the blasting particles were used with and without
fibronectin coatings using real time reverse transcription-
polymerase chain reaction (qRT-PCR). We have studied
the cell adhesion response quantifying a5, a3,and avinte-
grin-subunit gene expression; and the cell differentiation
response quantifying alkaline phosphatase (ALP) and
osteocalcin (OC) gene expression as well as ALP activity
and synthesis of OC.
618J Mater Sci: Mater Med (2011) 22:617–627
2 Materials and methods
Seven different series of commercially pure titanium grade
2 (c.p. Ti) disks, 8 mm in diameter and 2 mm thick, were
obtained depending on the surface treatments applied to
Smooth: polished c.p. Ti. The samples were subse-
quently abraded with 400, 600 and 1200 grit silicon carbide
abrasive paper and then polished with a water suspension
of 1 lm alumina powder.
A3: c.p. Ti blasted with Al2O3particles of 212–300 lm
A6: c.p. Ti blasted with Al2O3particles of 425–600 lm
A9: c.p. Ti blasted with Al2O3particles of 1,000–1,400
lm in size.
S3: c.p. Ti blasted with SiC particles of 212–300 lm in
S6: c.p. Ti blasted with SiC particles of 425–600 lm in
S9: c.p. Ti blasted with SiC particles of 1,000–1,400 lm
For the qRT-PCR experiments, only the smooth, A6 and
S6 surface finishes were tested. Tissue culture polystyrene
(TCPS) was studied as a positive control for these
Blasting was carried out with a laboratory blasting
machine at 0.25 MPa pressure for the time required for
saturation of the roughness of the samples. The particles
used for each sample series were as shown above.
After the blasting, the samples were cleaned by soni-
cation in acetone for 15 min, followed by sonication in
distilled water for 15 min. Finally, all samples were steam-
sterilized at 121?C for 30 min and kept under vacuum.
2.1.2 Fn coating
Two series of the different surface finishes were performed
in parallel for all the studies: (a) fibronectin (Fn) pre-coated
surfaces; and (b) uncoated/plain surfaces. The former were
coated with human Fn (Sigma-Aldrich) at a concentration
of 20 lg/ml for 30 min at 37?C. The latter were not coated
prior to the in vitro cell assays.
Human osteoblast-like MG63 cells (American Type Cul-
ture Collection, Rockville, MD) were grown in Dulbecco’s
modified Eagle’s medium (DMEM, Gibco) supplemented
with 10% fetal calf serum (FCS), 1% penicillin/strepto-
mycin, 1% L-glutamine, and 1% pyruvate (Sigma-Aldrich)
at 37?C in a 5% CO2/95% air atmosphere and at 100%
humidity. The MG63 cells were obtained from American
Type Culture Collection (Rockville, MD). The culture
medium was changed every 2 days. For the experiments,
cells were harvested at 70–90% confluence by trypsin/
EDTA, centrifuged and re-suspended in a serum-free
medium before being plated at 2 9 104MG63 cells per
For the differentiation studies, after 7 days of incubation,
the time course of mineralization was accelerated by chang-
ing to an osteogenic medium, consisting of a complete
2-phosphate ascorbic acid, and 10 mM b-glycerophosphate.
2.2.1 Real-time qRT-PCR technique
Gene expression of integrins a5, a3and av(30 min and 4 h
incubation) and of ALP and OC (7 and 14 days of culture)
was determined through a real-time reverse transcription-
polymerase chain reaction (qRT-PCR) assay. b2-lglobulin
was used as a housekeeping gene.
The following protocol was applied: (1) mRNA
extraction; (2) spectrophotometric quantification of RNA;
(3) reverse transcription (RT); and (4) polymerase chain
A kit (RNeasy Minikit, QIAGEN, Germany) was used to
extract the mRNA from the cell pellets, and quantification
was achieved by measuring the absorbance with a Nano-
Drop?Spectrophotometer ND-1000 (Thermo Scientific,
USA) using the ND-1000 software (v 3.5.2) (Thermo
Scientific, USA). Reverse transcription is accomplished
when the enzyme reverse transcriptase makes a double-
stranded cDNA copy of the mRNA molecules (QuantiTect
Reverse Transcription Kit, QIAGEN, Germany). The
resulting cDNA is then amplified using the SYBR green
method (QuantiTect SYBR
Germany). This requires a double-stranded DNA dye in the
PCR reaction that binds to newly synthesized double-
stranded DNA and generates fluorescence.
Green kit, QIAGEN,
2.2.2 Cell differentiation markers
Cell differentiation was also assessed by measuring alka-
line phosphatase activity (ALP) and osteocalcin production
(OC). Cells were culture for 14 and 21 days in osteogenic
media as described above. An additional control group
consisted of cells on non-coated surfaces after 3 days in
culture was tested. Both ALP and OC activity were
J Mater Sci: Mater Med (2011) 22:617–627619
normalised to the total protein concentration quantified
with a BCA Protein Assay kit (PIERCE).
ALP activity was measured with a Phosphatase, Alka-
line Acid, Prostatic Acid Assay kit (SIGMA Diagnostics).
Samples were rinsed twice in phosphate-buffered saline
(PBS) and soaked in 500 ll MPER (M-PER Mammalian
Protein Extraction Reagent, Thermo Scientific). The
MPER was meant to detach the cells from the surfaces and
lysate them. Readings were taken at 405 nm on a Power-
WaveX Bio-Tek Instruments spectrophotometer.
The osteocalcin (OC) concentration was measured by
means of a Metra?Osteocalcin kit (Roche), a competitive
immunoassay, and normalized to BCA readings.
Each group was tested in triplicate.
2.2.3 Statistical analysis
ANOVA tables with Fisher’s or Tukey’s post-hoc multiple-
comparison tests were performed using MinitabTMRelease
14 software (Minitab Inc., USA) to assess statistically
significant differences (P value B 0.05) between groups.
Tukey’s test was performed when comparing groups with
samples of different size.
3.1 Cell adhesion
3.1.1 a5, avand a3integrin subunit expression
Cells’ a5 integrin subunit (a5 mRNA) gene expression
increased with time on all surfaces except smooth ones
(Fig. 1a). After 30 min in culture, cells on smooth Ti sur-
faces showed the highest levels of a5mRNA expression.
However, after 4 h of cell culture, cells on rough Ti sur-
faces expressed higher levels of a5mRNA than cells on
smooth Ti samples. In contrast, gene expression of avand
a3integrin subunits increased with time on all surfaces
except the alumina-blasted series. The A6 surface finish
showed a decrease in a3mRNA with time (Fig. 1b), while
gene expression levels of av(Fig. 1c) remained statistically
constant with time.
The SiC-blasted series (S6) expressed higher levels of a5
and av mRNA than the Al2O3-blasted ones (A6). This
effect was more evident after 4 h of incubation (Fig. 1a, c).
The a3integrin subunit was also expressed more S6 sur-
faces than on A6, but only in the non Fn-coated series.
Osteoblastic cells on Fn pre-coated surfaces showed a
lower level of a5mRNA expression than cells on uncoated
surfaces (Fig. 1a). This effect was most noticeable on the
expression of this gene in cells in the TCPS series and least
noticeable in cells cultured on smooth surfaces. Fn delayed
the expression of a5mRNA by cells after 4 h in culture on
all studied surfaces except for smooth ones. Conversely, Fn
coating increased avintegrin subunit expression after 4 h of
relative amount of α5
relative amount of α3
relative amount of αv
Fig. 1 Gene expression of a a5, b a3and c avintegrin subunits by
cells on Ti surfaces with (Fn) or without (0) Fn coating measured by
qRT-PCR after 30 min and 4 h of cell culture. Ratios of target genes
relative to the housekeeping gene—lglobulin—were expressed as a
percentage relative to the uncoated smooth surfaces after 30 min of
cell culture. Bars with the same symbols (*, #) indicate that the
differences between them are not statistically significant when
compared in the same surface finish group
620J Mater Sci: Mater Med (2011) 22:617–627
incubation time on all surfaces except alumina ones, on
which it remained constant (Fig. 1c). Also, Fn-coating
surfaces increased gene expression of a3integrin subunit
on all surfaces after 4 h of incubation time, except for S6
on which a notable decrease of the a3gene expression was
detected (Fig. 1c).
3.2 Cell differentiation
3.2.1 Alkaline phosphatase gene expression
The ALP mRNA gene expression of MG63 osteoblast-like
cells increased with time on all the studied surfaces
(Fig. 2). Cells on rough and uncoated surfaces exhibited
higher levels of ALP mRNA expression than cells on TCPS
After 7 days of cell culture, Fn coatings did not have a
prevalent effect on ALP expression. In contrast, after
14 days in culture, cells on Fn-coated surfaces exhibited
lower levels of ALP mRNA expression than cells on
uncoated surfaces in all cases (Fig. 2).
3.2.2 Alkaline phosphatase activity
The ALP results were normalized to the BCA readings.
Figure 3 shows the results of the ALP/total protein ratio
for cells on all smooth and rough Ti series. The cells’
ALP activity increased with time on all studied surfaces.
Differences were statistically significant when the results
were compared after 3 days.
Roughness did not influence the cells’ ALP activity after
14 days in culture. In contrast, after 21 days of cell culture,
roughness was seen to affect ALP activity, which was
generally higher on rough surfaces than on smooth ones
(Fig. 4). This trend was more noticeable for uncoated
surfaces than for Fn-coated ones.
After 14 days of cell culture, Fn did not have a signif-
icant impact on ALP activity. Conversely, after 21 days in
culture, cells on Fn pre-coated series showed lower levels
of ALP activity than cells on uncoated ones (Fig. 3).
3.2.3 Osteocalcin gene expression
OC mRNA expression of MG63 osteoblast-like cells
increased with time with all studied surfaces (Fig. 5)
except the uncoated alumina-blasted Ti ones.
The effects of roughness and Fn coating on OC mRNA
gene expression were unclear. However, blasting particles
had a notable effect on it. OC mRNA expression on alu-
mina-blasted surfaces was higher than on SiC-blasted
3.2.4 Osteocalcin production
Figure 6 shows the OC production of osteoblast-like cells
cultured on all the studied surfaces. OC production
decreased with time in cells on uncoated surfaces (Fig. 7,
left graph), whereas it showed an increase with time in
cells on Fn-coated surfaces (Fig. 7, right graph).
Cells on rough surfaces had higher OC production levels
than cells on smooth surfaces. This effect was more pro-
nounced in uncoated samples than in Fn-coated ones.
TCPS Smooth S6A6
relative amount of ALP
Fig. 2 ALP gene expression of MG63 osteoblast-like cells on Ti
surfaces with (Fn) or without (0) Fn coating measured by qRT-PCR
after 7 and 14 days of cell culture. Ratios of target genes relative to
the housekeeping gene—lglobulin—were expressed as a percentage
relative to the uncoated smooth surfaces after 7 days of cell culture.
Bars with the same symbols (*, #) indicate that the differences
between them are not statistically significant when compared in the
same surface finish group
ALP (mmol/μg protein/min)
Fig. 3 ALP activity of MG63 cells after 3, 14 and 21 days in culture
on different rough Ti surfaces with (Fn) and without (0) Fn coating.
Bars with the same symbols (*, #) indicate that the differences
between them are not statistically significant when compared in the
same surface finish group
J Mater Sci: Mater Med (2011) 22:617–627621
As for the chemical composition of the blasting particles
used to roughen the surfaces, cells on Al2O3-blasted sur-
faces showed a greater increase in OC production than cells
on SiC-blasted surfaces after the same time in culture on
uncoated series (Fig. 7, left graph). No correlation was
found between OC production and the nature of the
blasting particles for cells on Fn-coated surfaces.
The influence of surface properties and fibronectin coatings
on smooth and blasted rough Ti samples on MG63 osteo-
blast-like cell adhesion and differentiation was studied to
learn more about their possible effect on cell osteogenic
processes. Expression of a5integrin subunit mRNA was
investigated, since this integrin’s specificity for fibronectin
makes it a good indicator of the quality of cell adhesion in
these experiments. avand a3integrin subunits were also
studied, since they compete with a5to bind to the Fn RDG
site; however, these integrins are not specific for this
adhesive protein. Cell differentiation was studied not only
by analyzing differentiation markers (alkaline phosphatase
activity and osteocalcin production) using traditional
techniques, but also by investigating alkaline phosphatase
and osteocalcin mRNA gene expression using the more
reliable qRT-PCR technique.
MG63 osteoblast-like cells, originally isolated from a
human osteosarcoma, are widely used and studied for in
vitro tests [12, 15, 16, 26]. They are a well-characterized
cell model to study the influence of the topography of a
surface on the osteoblast adhesion, proliferation and dif-
ferentiation. Although a tumor cell line, MG63 cells are
relatively immature osteoblasts that exhibit bone forming
osteoblastic traits. Alkaline phosphatase activity and oste-
ocalcin are examples of phenotypic markers that MG63
cells can exhibit. We focused here on assessing surface
ALP (mmol/μg protein/min)
ALP (mmol/μg protein/min)
Fig. 4 ALP activity of MG63 cells after 14 and 21 days in culture on different rough Ti surfaces. The depicted values are the same ones shown
in Fig. 3, but the uncoated (left) and Fn-coated (right) series are separated, and ALP activity after 3 days is not included
TCPS SmoothS6 A6
relative amount of OC
Fig. 5 Gene expression of OC of MG63 cells after 7 and 14 days in
culture on blasted Ti surfaces with (Fn) or without (0) Fn coating
measured by qRT-PCR. Ratios of target genes relative to the
housekeeping gene—lglobulin—were expressed as a percentage
relative to uncoated smooth surfaces after 7 days in culture. Bars with
the same symbols (*, #) indicate that differences between them are
not statistically significant when compared in the same surface finish
OC (ng/μg protein)
Fig. 6 Osteocalcin production of MG63 cells after 3, 14 and 21 days
in culture on different rough Ti surfaces with (Fn) and without (0) Fn
coating. Bars with the same symbols (*, #) indicate that differences
between them are not statistically significant when compared in the
same surface finish group
622J Mater Sci: Mater Med (2011) 22:617–627
effects on osteoblasts response using the model MG63 cell
line, but studies using human primary cells will be further
4.1 Cell adhesion
a5mRNA expression was considered a good indicator of
the adhesion response for osteoblastic cells in view of
previous knowledge of the selectivity of the integrin
receptor a5b1for fibronectin. a5b1binding to fibronectin
has been shown to be necessary for bone-like nodule for-
mation in vitro when osteoprogenitor cells are grown on
tissue culture plastic and other synthetic biomaterials .
However, a5b1is not the only integrin that binds to fibro-
nectin. Other integrins also compete to bind to the RGD
site. This is the case with the a3b1integrin, which binds to
laminin, collagen, and fibronectin , and avb3, which
binds to vitronectin, fibrinogen, fibronectin, and osteo-
pontin, among others proteins .
The quantitative results for a5mRNA gene expression
indicated that the values of a5mRNA expression increased
with the time of cell culture except in cells cultured on
smooth Ti surfaces (Fig. 1a). After 30 min in culture, gene
expression in cells on smooth Ti surfaces significantly
increased compared to cells in all other series. However,
after 4 h in culture, cells on smooth Ti surfaces had the
lowest gene expression values, whereas the cells on the rest
of the samples maximized the gene expression differences
between them. Moreover, the studies performed after
30 min in culture showed nearly no differences between
the Fn pre-coated series and the uncoated ones for any of
the integrin subunits (i.e. a5, avor a3). On the basis of these
results, it can be concluded that 30 min of culture is too
short a time for cells to develop the appropriate biological
mechanisms to adhere to surfaces. Therefore, it is not
advisable to differentiate the effects of the different surface
treatments on osteoblast adhesion response at 30 min of
Osteoblasts interact with the substrate through integrin
receptors. The type of substrate determines which integrins
and ECM proteins are expressed. After 4 h of cell culture,
cells on rough-blasted Ti surfaces expressed higher levels
of a5 mRNA than cells on smooth samples (Fig. 1a).
Others have reported similar results for MG63 cultured on
Ti, in which the expression of integrin subunits a2, a3, a5,
b1, and b3increased more on rougher surfaces than on
polished ones .
Our results confirmed the influence of the nature of the
blasting particles on a3 and av integrin subunit gene
expression (Fig. 1). Cells on SiC-blasted Ti surfaces
expressed higher amounts of the a5, avand a3mRNA gene
than cells on Al2O3-blasted Ti surfaces after 4 h in culture;
the differences were not statistically significant for the a5
integrin subunit. In particular, this behavior was most
noticeable with avmRNA expression. These differences
may be caused by the higher amount of Fn that is absorbed
on SiC-blasted surfaces, as determined previously .
The higher amount of adsorbed Fn, which is due to the
higher surface free energy of SiC-blasted surfaces, may
give the surface a higher number of specific cell-binding
sites. This, in turn, would result in increased a5, avand a3
mRNA gene expression.
We previously determined an increase in number of
adhered MG63 cells after 4 h of incubation when surface
roughness was increased on uncoated surfaces . Those
findings correlate with the increase in a5mRNA expression
found in this work (Fig. 1). Any relation was found
between cell adhesion on uncoated surfaces and the rest of
the studied integrins. We also demonstrated that Fn coat-
ings on Ti surfaces led to a decrease in number of adhered
cells in combination with enhanced cell spreading. Inter-
estingly, those results nicely correlate with decreasing and
increasing a5mRNA and aVmRNA expression, respec-
tively. Thus, the expression of a5b1integrin is the main
factor influencing cell adhesion in the type of surfaces
tested in this work.
SmoothS3 S6S9A3A6 A9
OC (ng/μg protein)
Smooth S3S6 S9A3A6 A9
OC (ng/μg protein)
Fig. 7 OC production of MG 63 cells after 14 and 21 days in culture on different rough Ti surfaces. The depicted values are the same ones
shown in Fig. 6, but the uncoated (left) and Fn-coated (right) series are separated, and OC production after 3 days is not included
J Mater Sci: Mater Med (2011) 22:617–627 623
Differences in the a5, avand a3mRNA gene expression
of osteoblasts detected between the different substrates
studied were statistically significant only after 4 h of cell
culture. This includes one of the most interesting results
obtained, namely, that cells on Fn-coated Ti surfaces
showed lower a5mRNA gene expression than cells on the
same type of surface when it was uncoated. This was true
of all surface types analyzed. This is an unexpected result.
One might speculate that the increased gene expression on
Fn-coated surfaces is due to the aforementioned specificity
of the a5b1integrin—and, thus, the associated a5expres-
sion—to bind to fibronectin.
The a5integrin subunit binds to b1to form the integrin
a5b1, and it is via a5b1that cells initiate the attachment to
fibronectin [27, 29]. One reason for cells to express less a5
mRNA on Fn-coated surfaces might be that cells express
other integrins instead of a5b1. Moreover, different inte-
grins can bind to Fn through the RDG site, including a5b1,
avb3and a3b1. The fact that Fn coating might provide
the surface with a higher number of specific cell-binding
sites may be the cause for cells to express other integrins.
In this respect, cells tend to provide the necessary inte-
grins to perform their biological activities by up- or down-
regulating the required subunits and integrins.
With respect to a2b1integrin expression, Raz et al. 
found higher b1integrin subunit expression and lower a5
subunit expression on Ti surfaces with rough microtopog-
raphies than on surfaces with smooth ones when MG63
cells were cultured up to confluence. Thus, a preferential
a2b1integrin expression by cells on rough surfaces was
assessed. It is worth noting that a2b1is known to be pref-
erentially expressed when cells reach confluence and
undergo osteoblastic differentiation , as observed by
Raz et al. . However, Anselme et al.  did not find
any a2b1expression when human osteoblasts were cultured
on Ti-alloy surfaces for 24 h, 7 days or 14 days.
Thus, in our experiments, an up-regulation of a2integrin
subunit is not expected since the shorter adhesion times we
tested, i.e., cells are far from being confluent.
However, our results showed that surfaces with Fn
coatings induced cells to up-regulate avintegrin subunit
expression on all surface finishes, except for those blasted
with Al2O3 particles. Also, surfaces with Fn coatings
induced cells to up-regulate a3integrin subunit expression
on all surface finishes, except for those blasted with SiC
particles. Consequently, either the upregulation of avb3, or
a3b1, or both integrin subunits depending on the type of
surface finishing can compensate the measured down-
regulation of the Fn-specific a5b1integrin.
Given that both a5b1and avb3integrins are known to
bind to the Fn RGD site [20, 21], it is important to note that
integrin binding specificity—a5b1versus avb3—has been
linked in the literature by Garcia and coworkers  to the
regulation of osteoblastic differentiation and the enhance-
ment of in vivo bone healing and functional implant
osseointegration . In this study, cells were grown on
rough Ti substrates, which also stimulate osteoblastic dif-
ferentiation more than smooth surfaces. Improved bone
tissue formation and functional osseointegration may be
attributed to a5b1-specific Ti implants because it increases
recruitment of osteoprogenitor cells and differentiation into
osteoblasts at the tissue-implant interface. Moreover a5b1
integrin is the main fibronectin receptor, and its expression
has been associated with increased mineralization of oste-
osarcoma and calvarial osteoblast cells . However, the
adsorption of RGD containing proteins such as fibronectin
or vitronectin in a non-specific way may support avb3-
mediated adhesion due to the conformation of the adsorbed
protein. Our results support this claim, since Fn coating,
adsorbed in a non-specific way, increased avintegrin sub-
unit expression on all studied surfaces except the Al2O3
ones (Fig. 1c). In fact, Garcia and coworkers  grafted a
specific sequence of fibronectin (9th type repeat) onto Ti,
which in the proper structural context exhibits high selec-
tivity for the integrin a5b1.They also grafted a linear RGD
oligopeptide (GRGDSPC) onto Ti, demonstrating that it
primarily supports avb3-mediated adhesion. Keselowsky
 and Cheng et al.  showed that avb3-mediated
surface adhesion suppresses osteoblastic differentiation.
This agrees with our results, further explained in the fol-
lowing section, since Fn pre-coated surfaces showed high
expression of avmRNA integrin and low expression of a5
mRNA integrin, low cell differentiation, and low expres-
sion of ALP mRNA and ALP activity compared to
uncoated surfaces (Figs. 2, 3).
While these results showed substrate-dependent differ-
ences in integrin subunit expression levels, the influence of
surface topography on integrin functions, such as binding
activity and downstream biological responses, was still
unknown. Therefore, further analyses were undertaken to
identify the mediation of a5, avand a3mRNA integrin
subunit expression in osteoblast differentiation in relation
to surface-dependent differences.
4.2 Cell differentiation
This study has shown that roughness and Fn coating
influences a5, avand a3mRNA gene expression and, thus,
cell adhesion. Therefore, it seemed reasonable to assume
that Fn coating would alter the time course of differentia-
tion. In all types of surfaces tested, both ALP mRNA
expression (Fig. 2) and ALP activity (Fig. 3) were lower
(sometimes with statistically significant differences) on Fn
pre-coated samples than on uncoated surfaces after 14 and
21 days of cell culture. However, none of these differences
between Fn-coated and uncoated series was statistically
624 J Mater Sci: Mater Med (2011) 22:617–627
significant for ALP activity after 14 days of cell culture.
Consequently, the decrease in MG63 a5as well as increase
in MG63 av mRNA expression upon cell adhesion is
related to the influence of the non-specific Fn adsorption on
the osteoblast differentiation pathway.
As expected, increased roughness led to increased ALP
activity. This response was more significant for non pre-
coated surfaces after 21 days of cell culture (Fig. 2). The
same trend was seen in ALP mRNA levels after 14 days of
cell culture (Fig. 3). At shorter cell culture times, when cell
differentiation was in earlier stages, no statistically sig-
nificant differences were found for ALP mRNA expression
(7 days) or ALP activity (14 days).
Gene expression occurs earlier than the production and
activity of the proteins tested as cell differentiation mark-
ers. Consequently, the qRT-PCR assays for ALP mRNA
and OC mRNA were performed after shorter periods than
the tests for the differentiation markers. The time of cell
culture at which differentiation markers should be mea-
sured is difficult to determine, since many parameters, such
as the cell line or the number of cells seeded, can influence
the results obtained.
It is generally accepted that smooth surfaces promote
early differentiation markers, such as ALP activity [13, 26,
36], whereas rougher surfaces promote later steps in cell
maturation, such as mineralization of the ECM [13, 16, 36,
37]. The influence of rough surfaces on ALP activity has
been extensively studied. Schwartz et al.  concluded
that rougher surfaces increased alkaline phosphatase
activity and osteocalcin production by cells compared to
smooth surfaces. This was also reported by Martin et al.
, who found that human osteoblast-like MG63 cells and
normal human osteoblasts exhibited more differentiated
phenotypes when grown on substrates with micro-rough
surfaces. Orsini et al.  observed that sandblasting and
acid-etching can improve cellular adhesion and the pro-
liferation of osteoblast-like MG63 cells. Some of the dif-
ferences in cellular response to surface roughness may
have been due to variations in the degree of cell confluence
obtained or the cell line studied.
This study found differences in ALP mRNA expression
and ALP activity depending on the nature of the grit-
blasting particle. After 14 days, cells cultured on uncoated
alumina-blasted surfaces expressed higher levels of the
ALP mRNA gene than cells on uncoated SiC-blasted sur-
faces (Fig. 2). Likewise, after 21 days of cell culture, ALP
activity was higher for the A3 and A6 surfaces than for the
S3 and S6 surfaces, respectively (Fig. 4, left). The fact that
SiC surfaces showed higher Fn adsorption than Al2O3
surfaces due to their higher SFE  led to higher avb3and
lower a5b1integrin expression and, thus, less cell differ-
entiation, as explained in the previous section. However,
this blasting-particle effect was not found in cells on Fn
pre-coated surfaces (Fig. 4, right).
The osteocalcin gene has been shown to be involved in
other ways in the process of bone mineralization. OC,
which is an osteoblast-specific protein, has been demon-
strated to stimulate bone mineral maturation . In this
study, OC mRNA gene expression increased with culture
time (Fig. 5). It is worth remembering that OC mRNA
gene expression usually occurs prior to OC production.
Following its occurrence, one might expect OC production
to increase with time. However, this sequence of events
might not occur, since OC gene expression by cells is not
directly related to the final cell protein synthesis. Different
factors, such as changes in the dynamic protein layer
adsorbed on biomaterials surface, could lead to a change in
protein synthesis by cells. No relevant differences in OC
production were obtained after 14 and 21 days of cell
culture (Fig. 6), but rather the results fluctuated. On
uncoated surfaces, OC mRNA gene expression increased
with time, while OC production decreased with time. In
spite of the fact that OC gene expression increased between
7 and 14 days of incubation, a significant down-regulation
of OC mRNA gene expression can occur after that time.
This was eventually detected as a decrease in OC pro-
duction following the longest cell culture periods.
One possible explanation for these results might be the
differences in the protein layer adsorbed on Ti surfaces,
which would affect integrin expression. With regard to the
results for ALP activity, the up-regulation of avb3expres-
sion could also decrease cell differentiation, i.e., decrease
OC activity with time.
Cell maturation stage may also explain the decrease in
OC production after 21 days of cell culture. Depending on
the state of cell confluence and the cell line, the detection
periods for cell differentiation markers can vary .
Furthermore, the use of total protein concentration to
normalize the values of cell differentiation markers may
not be the most appropriate choice. Total protein concen-
tration depends on the total number of lysed cells. These
cells may be shielded by the extracellular matrix. Zinger
et al.  suggest using the number of cells to normalize
ALP and OC results. The cell number is obtained by
trypsinizing the culture cells and then counting them with a
cytometer. Other methods, such as lactate dehydrogenase
(LDH), are not recommended since they also rely on cell
Roughness has been shown to influence cell adhesion
and differentiation. In general, cells cultured on rougher
surfaces tend to exhibit attributes of more highly differ-
entiated osteoblasts and higher OC production than cells
cultured on smoother surfaces for comparable periods.
However, Fn coating affected cells’ OC production and OC
J Mater Sci: Mater Med (2011) 22:617–627625
mRNA expression. With Fn-coated surfaces, both OC gene
expression and OC production by cells increased with time
throughout the experiment. This indicates that the cells on
those surfaces remained in an earlier stage of maturation
than the cells on the non-Fn-coated surfaces, i.e., they had
not yet reached the mineralization stage.
General relationships between ALP and OC were
described by Owen et al.  in a study of differentiation-
associated genes (ALP and OC) based on the mRNA levels
expressed by calvarial-derived osteoblasts. A model of the
relationship between proliferation and differentiation was
established. One of the main conclusions was that, as the
cultures progressed into the mineralization stage, cellular
levels of alkaline phosphatase declined. Osteocalcin
exhibited the opposite pattern of expression. In general,
there is an initial period of active proliferation in which
cells grow and related genes are actively expressed. Then,
the down-regulation of proliferation leads to a period of
matrix maturation. Finally, when the alkaline phosphatase
gene is maximally expressed, the extracellular matrix is
competent for mineralization. The results of our study
show that cells’ ALP activity is either still under or near the
maximum values at 21 days of culture. Consequently, OC
mRNa expression and the associated OC production were
measured when cells were still in an immature stage, at
which point their values may fluctuate.
Cells on SiC-blasted Ti surfaces expressed higher amounts
of the a5, a3, and avmRNA gene than cells on Al2O3-
blasted Ti surfaces. This may be related to the fact that
SiC-blasted surfaces adsorb higher amounts of fibronectin
due to their higher surface free energy and, thus, offer a
higher number of specific cell-binding sites.
Ti surfaces coated with fibronectin decreased the a5
mRNA gene expression by cells. Either the upregulation of
avb3, or a3b1, or both integrin subunits depending on the
type of surface finishing compensate the measured down-
regulation of the Fn-specific a5b1integrin.
The changes in a5and avmRNA expression induced by
the presence of fibronectin coatings can also influence the
osteoblast differentiation pathway, as fibronectin coatings
on Ti surfaces decreased both cell ALP mRNA expression
and ALP activity after 14 and 21 days of cell culture.
Spanish Interministerial Commission for Science and Technology
(CICYT) for financial support under the MAT2003-08165 project.
They would also like to thank Klockner S. L. for technical help in
sample preparation and MPA, S. L. (Materias Primas Abrasivas) for
providing the blasting particles and blasting machine. M. Pegueroles
The authors would like to acknowledge the
would like to thank the Universitat Polite `cnica de Catalunya (UPC)
for grant funding to complete her PhD thesis.
1. Anselme K. Osteoblast adhesion on biomaterials. Biomaterials.
2. Puleo DA, Nanci A. Understanding and controlling the bone-
implant interface. Biomaterials. 1999;20:2311–21.
3. Groth T, Altankov G, Klosz K. Adhesion of human peripheral-
blood lymphocytes is dependent on surface wettability and pro-
tein preadsorption. Biomaterials. 1994;15:423–8.
4. Degasne I, Basle MF, Demais V, Hure G, Lesourd M, Grolleau B,
Mercier L, Chappard D. Effects of roughness, fibronectin and
vitronectin on attachment, spreading, and proliferation of human
osteoblast-like cells (Saos-2) on titanium surfaces. Calcif Tissue
5. Lebaron RG, Athanasiou KA. Extracellular matrix cell adhesion
peptides: functional applications in orthopedic materials. Tissue
6. Faucheux N, Tzoneva R, Nagel MD, Groth T. The dependence of
fibrillar adhesions in human fibroblasts on substratum chemistry.
7. Garcia AJ. Get a grip: integrins in cell–biomaterial interactions.
8. Keselowsky BG, Collard DM, Garcia AJ. Surface chemistry
modulates focal adhesion composition and signaling through
changes in integrin binding. Biomaterials. 2004;25:5947–54.
9. Gronthos S, Stewart K, Graves SE, Hay S, Simmons PJ. Integrin
expression and function on human osteoblast-like cells. J Bone
Miner Res. 1997;12:1189–97.
10. Sousa SR, Moradas-Ferreira P, Saramago B, Melo LV, Barbosa
MA. Human serum albumin adsorption on TiO2 from single
protein solutions and from plasma. Langmuir. 2004;20:9745–54.
11. Aparicio C, Gil FJ, Thams U, Munoz F, Padros A, Planell JA.
Osseointegration of grit-blasted and bioactive titanium implants:
histomorphometry in mini pigs. Key Eng Mater. 2004;254–2:
DL, Dean DD, Schwartz Z. Response of MG63 osteoblast-like
cells to titanium and titanium alloy is dependent on surface
roughness and composition. Biomaterials. 1998;19:2219–32.
13. Groessnerschreiber B, Tuan RS. Enhanced extracellular-matrix
production and mineralization by osteoblasts cultured on titanium
surfaces in vitro. J Cell Sci. 1992;101:209–17.
14. ter Brugge PJ, Torensma R, De Ruijter JE, Figdor CG, Jansen JA.
Modulation of integrin expression on rat bone marrow cells by
substrates with different surface characteristics. Tissue Eng. 2002;
15. Kieswetter K, Schwartz Z, Hummert TW, Cochran DL, Simpson
J, Dean DD, Boyan BD. Surface roughness modulates the local
production of growth factors and cytokines by osteoblast-like
MG-63 cells. J Biomed Mater Res. 1996;32:55–63.
16. Martin JY, Schwartz Z, Hummert TW, Schraub DM, Simpson J,
Lankford J, Dean DD, Cochran DL, Boyan BD. Effect of titanium
surface-roughness on proliferation, differentiation, and protein-
synthesis of human osteoblast-like cells (Mg63). J Biomed Mater
17. Siebers MC, ter Brugge PJ, Walboomers XF, Jansen JA. Integrins
as linker proteins between osteoblasts and bone replacing mate-
rials. A critical review. Biomaterials. 2005;26:137–46.
18. Keselowsky BG, Garcia AJ. Quantitative methods for analysis of
integrin binding and focal adhesion formation on biomaterial
surfaces. Biomaterials. 2005;26:413–8.
626 J Mater Sci: Mater Med (2011) 22:617–627
19. Sousa SR, Moradas-Ferreira P, Barbosa MA. TiO2type influences
fibronectin adsorption. J Mater Sci Mater Med. 2005;16:1173–8.
20. Garcia AJ, Vega MD, Boettiger D. Modulation of cell prolifer-
ation and differentiation through substrate-dependent changes in
fibronectin conformation. Mol Biol Cell. 1999;10:785–98.
21. Hynes RO. Integrins—versatility, modulation, and signaling in
cell-adhesion. Cell. 1992;69:11–25.
22. Heino J, Massague ´ J. Transforming growth factor-b switches the
pattern of integrins expressed in MG-63 human osteosarcoma cell
and causes a selective loss of cell adhesion to laminin. J Biol
23. Aparicio C, Gil FJ, Fonseca C, Barbosa M, Planell JA. Corrosion
behaviour of commercially pure titanium grit blasted with dif-
ferent materials and sizes of grit particles for dental implant
applications. Biomaterials. 2003;24:263–73.
24. Pegueroles M, Gil FJ, Planell J, Aparicio C. The influence of
blasting and sterilization on static and time-related wettability
and surface-energy properties of titanium surfaces. Surf Coat
25. Pegueroles M, Aparicio C, Bosio M, Engel E, Gil FJ, Planell JA,
Altankov G. Spatial organization of osteoblast fibronectin-matrix
on titanium surfaces—effects of roughness, chemical heteroge-
neity and surface free energy. Acta Biomater. 2010;6:291–301.
26. Zhao G, Raines AL, Wieland M, Shwartz Z, Boyan BD.
Requirement for both micron- and submicron scale structure for
topography. Biomaterials. 2007;28:2821–9.
27. Keselowsky BG, Wang L, Schwartz Z, Garcia AJ, Boyan BD.
Integrin a5 controls osteoblastic proliferation and differentiation
responses to titanium substrates presenting different roughness
characteristics in a roughness independent manner. J Biomed
Mater Res A. 2007;80:700–10.
28. Lange R, Luthen F, Beck U, Rychly J, Baumann A, Nebe B. Cell-
extracellular matrix interaction and physico-chemical character-
istics of titanium surfaces depend on the roughness of the
material. Biomol Eng. 2002;19:255–61.
29. Luthen F, Lange R, Becker P, Rychly J, Beck U, Nebe JGB.
The influence of surface roughness of titanium on beta 1- and
beta 3-integrin adhesion and the organization of fibronectin in
human osteoblastic cells. Biomaterials. 2005;26:2423–40.
30. Raz P, Lohmann CH, Turner J, Wang L, Poythress N, Blanchard
C, Boyan BD. 1a, 25(OH)2D3 regulation of integrin expression is
substrate dependent. J Biomed Mater Res A. 2004;71:217–25.
31. Jikko A, Harris SE, Chen D, Mendrick DL, Damsky CH.
Collagen integrin receptors regulate early osteoblast differentia-
tion induced by BMP-2. J Bone Miner Res. 1999;14:1075–83.
32. Anselme K, Bigerelle M, Noel B, Dufresne E, Judas D, Iost A,
Hardouin P. Qualitative and quantitative study of human osteo-
blast adhesion on materials with various surface roughnesses.
J Biomed Mater Res. 2000;49:155–66.
33. Keselowsky BG, Collard DM, Garcia AJ. Integrin binding
specificity regulates biomaterial surface chemistry effects on cell
differentiation. Proc Natl Acad Sci USA. 2005;102:5953–7.
The effect of integrin-specific bioactive coatings on tissue healing
and implant osseointegration. Biomaterials. 2008;29:2849–57.
35. Cheng SL, Lai CF, Fausto A, Chellaiah M, Feng X, Mchugh KP,
Teitelbaum SL, Civitelli R, Hruska KA, Ross FP, Avioli LV.
Regulation of alpha v beta 3 and alpha v beta 5 integrins by
dexamethasone in normal human osteoblastic cells. J Cell Bio-
36. Aparicio C, Gil FJ, Planell JA, Engel E. Human-osteoblast pro-
liferation and differentiation on grit-blasted and bioactive tita-
nium for dental applications. J Mater Sci Mater Med. 2002;
37. Boyan BD, Lossdorfer S, Wang L, Zhao G, Lohmann CH,
Cochran DL, Schwartz Z. Osteoblasts generate an osteogenic
microenvironment when grown on surfaces with rough microto-
pographies. Eur Cells Mater. 2003;6:22–7.
38. Schwartz Z, Lohmann CH, Oefinger J, Bonewald LF, Dean DD,
Boyan BD. Implant surface characteristics modulate differentia-
tion behavior of cells in the osteoblastic lineage. Adv Dent Res.
39. Orsini G, Assenza B, Scarano A, Piatelli M, Piatelli A. Surface
analysis of machined versus sandblasted and acid-etched titanium
implants. Int J Oral Maxillofac Implants. 2000;15:779–84.
40. Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty
G. Fourier transform infrared microspectroscopic analysis of
bones of osteocalcin-deficient mice provides insight into the
function of osteocalcin. Bone. 1998;23:187–96.
41. Ku CH, Piolettli DP, Browne M, Gregson PJ. Effect of different
Ti–6Al–4V surface treatments on osteoblasts behaviour. Bioma-
42. Zinger O, Zhao G, Schwartz Z, Simpson J, Wieland M, Landolt
D, Boyan BD. Differential regulation of osteoblasts by substrate
microstructural features. Biomaterials. 2005;26:1837–47.
43. Owen TA, Aronow M, Shalhoub V, Barone LM, Wilming L,
Tassinari MS, Kennedy MB, Pockwinse S, Lian JB, Stein GS.
Progressive development of the rat osteoblast phenotype invitro—
reciprocal relationships in expression of genes associated with
osteoblast proliferation and differentiation during formation of the
bone extracellular-matrix. J Cell Physiol. 1990;143:420–30.
J Mater Sci: Mater Med (2011) 22:617–627 627