Article

Journal of Biomaterials Applications

28(2) 278–287
! The Author(s) 2012
Differences in proliferation, Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
differentiation, and cytokine production DOI: 10.1177/0885328212440600
jba.sagepub.com
by bone cells seeded on titanium–nitride
and cobalt–chromium–molybdenum
surfaces

Ruud P van Hove1,2, Peter A Nolte2, Cornelis M Semeins1 and Jenneke Klein-Nulend1

Abstract
Titanium–nitride coating is used to improve cobalt–chromium–molybdenum implant survival in total knee arthroplasty,
but its effect on osteoconduction is unknown. Chromium and cobalt ions negatively affect the growth and metabolism of
cultured osteoblasts while enhancing osteoclastogenic cytokine production. Therefore, it was hypothesized that a titan-
ium–nitride surface would enhance osteoblast proliferation and/or differentiation and reduce osteoclastogenic cytokine
production compared with a cobalt–chromium–molybdenum surface. MC3T3-E1 osteoblasts showed increased prolif-
eration and decreased differentiation on titanium–nitride, while cytokine interleukin-6 production was higher on porous
cobalt–chromium–molybdenum (p < 0.05), though interleukin-1b was occasionally detected on both surfaces. These
findings suggest improved osteoconduction on titanium–nitride compared with cobalt–chromium–molybdenum surface.

Keywords
MC3T3-E1 osteoblasts, titanium–nitride, cobalt–chromium–molybdenum, proliferation, differentiation, cytokine
production

direct anchorage of the implant.6 Long-term durability

Introduction
Osteoarthritis of the knee is a common cause of pain
and disability. In severe cases, total knee arthroplasty is
recommended to relieve pain and disability with high
rate of success.1,2 However, failure of such surgery may
result in postoperative pain, poor function, and occa-
sionally revision surgery.3–5 The mechanisms underly-
ing failure include wear of the polyethylene insert,
aseptic loosening, instability, and infection.3,5 Early
failure, i.e. revision surgery within 2 years, is mainly
due to infection, instability, and loosening.5 Early
loosening has been shown to be associated with unce-
mented rather than cemented total knee arthroplasty.4,5
In total knee arthroplasty, the femoral and tibial
prosthesis’ components can be fixated by polymethyl-
methacrylate (PMMA or cement) between bone and the
components (cemented total knee arthroplasty) or by
bone ingrowth into the porous surface of the compo-
nents (uncemented total knee arthroplasty). Bone
ingrowth starts with osteoconduction, i.e. bone
growth on a surface, leading to osseointegration, and
of cemented total knee arthroplasty is questionable due
to signs of osteolytic activity at the cement–bone inter-
face.7 Uncemented total knee arthroplasty was
designed to improve the longevity of implants in
younger patients with osteoarthritis.8 However, the
rate of aseptic loosening in uncemented total knee
arthroplasty with need for revision is higher than that
in cemented total knee arthroplasty.1,4
Cobalt, chromium, and molybdenum (CoCrMo)
alloy exhibits high wear resistance and is therefore
used to manufacture the tibial and femoral components
1
Department of Oral Cell Biology, Research Institute MOVE, ACTA-
University of Amsterdam and VU University Amsterdam, The
Netherlands
2
Department of Orthopedics, Spaarne Hospital, The Netherlands
Corresponding author:
Jenneke Klein-Nulend, Department of Oral Cell Biology, Research
Institute MOVE, ACTA-University of Amsterdam and VU University
Amsterdam, Gustav Mahlerlaan 3004, 1081 LA Amsterdam, The
Netherlands.
Email: j.kleinnulend@acta.nl

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

van Hove et al. 279

Table 1. Experimental design of this study indicating the number of porous and smooth CoCrMo and TiN discs used in this study for DNA, ALP, IL-1b, IL-6, and TNF-a analysis after

TNF-a (4)

IL-1b: interleukin-1b in culture medium of MC3T3-E1 osteoblasts cultured on discs; IL-6: interleukin-6 in culture medium of MC3T3-E1 osteoblasts cultured on discs; and TNF-a: tumor necrosis factor-a in
CoCrMo: cobalt–chromium–molybdenum; TiN: titanium–nitride; DNA: DNA of MC3T3-E1 osteoblasts cultured on discs; ALP: alkaline phosphatase activity of MC3T3-E1 osteoblasts cultured on discs;
of total knee prostheses.9,10 Some implants have a titan-

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
7 (12)
ium–nitride (TiN) coating of 3–4 mm applied by phys-
ical vapor deposition.10 TiN provides the CoCrMo
surface with increased hardness, a low coefficient of

TNF-a (4)
DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
friction, and higher resistance to adhesive wear com-

4 (12)
pared with CoCrMo, which are favorable characteris-
tics for long-term survival of the implant, since it

TNF-a (4)
Porous TiN discs (48)

DNA (4)

IL-1b (4)
prevents the potentially harmful effects of accumulation

ALP (4)

IL-6 (4)
2 (12)
of chromium and cobalt ions in the body.10,11
Chromium and cobalt ions negatively affect growth
and metabolism of osteoblastic cells in vitro.12,13

TNF-a (4)
DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
These ions also enhance the release of interleukin-1b

1 (12)
Day
(IL-1b) and tumor necrosis factor-a (TNF-a).14
Cobalt ions alone enhance the release of interleukin-6

TNF-a (4)
(IL-6) by human osteoblasts in vitro.14

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
7 (12)
Osteoblast-like cells produce the cytokines interleu-
kin-1 (IL-1), IL-6, and TNF-a.15–19 In bone remodel-

TNF-a (4)
ing, cytokines IL-1, IL-6, and TNF-a regulate

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
osteoclastogenesis.20 They stimulate osteoclast differen-

4 (12)
tiation in a synergistic fashion with a net increase in
receptor activator of nuclear factor kappa B ligand

TNF-a (4)
Smooth TiN discs (48)

DNA (4)
(RANKL) activity.20,21 Binding of RANKL to

IL-1b (4)
ALP (4)

IL-6 (4)
2 (12)
RANK stimulates differentiation of osteoclast precur-
sors into mature osteoclasts and activation of osteo-
clastic bone resorption.20 Moreover, IL-1, IL-6, and

TNF-a (4)
DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
TNF-a cytokines induce osteoclast differentiation in

1 (12)
an in vitro mouse model for aseptic loosening.22 Day
The aim of this study was to test whether differences

TNF-a (4)
DNA (4)

IL-1b (4)
exist in MC3T3-E1 osteoblast proliferation, differenti-
ALP (4)

IL-6 (4)
7 (12)

ation, and/or cytokine production when these cells are

cultured on TiN surface compared with CoCrMo sur-

TNF-a (4)
face. We hypothesized that both smooth and porous
DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
TiN surface would enhance MC3T3-E1 osteoblast pro-
4 (12)

liferation and/or differentiation and diminish the pro-

Porous CoCrMo discs (48)

duction of cytokines by these cells compared with

TNF-a (4)
DNA (4)

IL-1b (4)

CoCrMo surface. This could enhance osteoconduction

ALP (4)

IL-6 (4)
2 (12)

and osseointegration, which might lead to a lower rate

of early aseptic loosening of the implant in vivo.
TNF-a (4)

culture medium of MC3T3-E1 osteoblasts cultured on discs.

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
1, 2, 4, or 7 days of MC3T3-E1 osteoblast culture.

1 (12)
Day

Materials and methods

TNF-a (4)

Materials
DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
7 (12)

Hundred and ninety-two discs composed of CoCrMo

Number of discs are given in parentheses.

alloy (diameter 25.5 mm, height 3.3 mm) were tested.

TNF-a (4)

Ninety-six discs were composed of CoCrMo alloy, of

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
4 (12)

which 48 discs were uncoated (smooth CoCrMo discs)

Smooth CoCrMo discs (48)
Total number of discs (192)

and 48 discs were coated with spherical CoCrMo beads

(porous CoCrMo discs) as used at the bone–implant
TNF-a (4)
DNA (4)

IL-1b (4)
ALP (4)

interface of total knee arthroplasty for bone ingrowth

IL-6 (4)
2 (12)

(PorocoatÕ , DePuy-Johnson and Johnson, Warsaw,

IN) (Table 1). Ninety-six other discs composed of
TNF-a (4)

CoCrMo alloy (48 uncoated discs (smooth TiN discs)

DNA (4)

IL-1b (4)
ALP (4)

IL-6 (4)
1 (12)

and 48 discs with a spherical beads coating (porous

Day

TiN discs)) were coated with a thin TiN layer

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

280
(4 mm thickness) applied by physical vapor deposition
(Implantcast GmbH, Buxtehude, Germany) (Table 1).
Polystyrene tissue culture plates (CellstarÕ , Greiner
Bio-One GmbH, Frickenhausen, Germany) were used
as reference for MC3T3-E1 osteoblast cultures. Prior to
utilization, all discs were ultrasonically cleaned in soapy
water, 70% ethanol, and de-ionized water. The discs
were packed in sterilization pouches, autoclaved
(Prestige Medical, Blackburn, UK) for 22 min at
121 C, and allowed to dry overnight at 60 C. The
discs tested were scanned using a scanning electron
microscope (SEM, Philips XL20, Philips Electronics
NV, Eindhoven, the Netherlands). SEM was used to
identify possible surface differences in samples of the
tested materials.
Bone cell culture
MC3T3-E1 osteoblasts were cultured in a-minimum
essential medium (a-MEM, Gibco, Paisley, UK) sup-
plemented with 10 mM ß-glycerophosphate (Sigma–
Aldrich, St. Louis, MO, USA), 300 mg/mL L-glutamine
(Sigma–Aldrich), 50 mg/mL ascorbic acid (Merck,
Darmstadt, Germany), 100 U/mL penicillin (Sigma–
Aldrich), 50 mg/mL streptomycin (Sigma–Aldrich),
1.25 mg/mL fungizone (Gibco), and 10% fetal bovine
serum (Hyclone, Logan, UT) in a CO2 incubator
(Hera Cell, Heraeus, Hanau, Germany) at 37 C in a
99% humidity atmosphere and 5% CO2 in air. After
reaching 90–100% confluency, MC3T3-E1 osteoblasts
were trypsinized with 0.25% trypsin (Difco
Laboratories, Detroit, MI, USA) and 0.1% EDTA
(Sigma–Aldrich) in phosphate-buffered saline (PBS) as
described before.23
MC3T3-E1 osteoblasts at passage 16 were seeded at
1  105 cells/disc in 0.5 mL culture medium consisting
of a-MEM (Gibco) with 10 mM ß-glycerophosphate
(Sigma–Aldrich), 300 mg/mL L-glutamine (Sigma–
Aldrich), 50 mg/mL ascorbic acid (Merck), 100 U/mL
penicillin (Sigma–Aldrich), 50 mg/mL streptomycin
(Sigma–Aldrich), 1.25 mg/mL fungizone (Gibco), and
10% fetal bovine serum (Hyclone) to study cell prolif-
eration, differentiation, and cytokine production.
MC3T3-E1 osteoblasts were placed on top of the
discs to allow cell adhesion to the discs for 45 min in
a CO2 incubator (Hera Cell) at 37 C in a 99% humidity
atmosphere and 5% CO2 in air, at which time an add-
itional 4.5 mL culture medium/well was added. Control
cultures without discs were cultured in 2.5 mL of cul-
ture medium to reach the same height of medium on the
MC3T3-E1 osteoblasts during culture on discs. After 1,
2, 4, and 7 days of culture at 37 C and 5% CO2 in air at
99% humidity, medium was collected to measure
IL-1b, IL-6, and TNF-a by enzyme-linked immuno-
sorbent assay (ELISA).
Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016
Journal of Biomaterials Applications 28(2)
Table 1 shows the experimental design of this study
indicating the number of porous and smooth CoCrMo
and TiN discs used in this study for DNA, alkaline
phosphatase (ALP), IL-1b, IL-6, and TNF-a analysis
after 1, 2, 4, or 7 days of MC3T3-E1 osteoblast culture.
The cell-containing discs and control cell cultures on
polystyrene were rinsed with PBS, and adhered cells
were lysed with 0.5 mL 0.05% TritonÕ X-100 and
stored at 80 C until further analysis. The porous-
coated discs were put in 50 mL polyethylene canisters
(Greiner Bio-One) and centrifuged at 600 relative cen-
trifugal force for 1 min to collect the total amount of
cell lysate.
On days 1, 2, 4, and 7, a disc of each material with
MC3T3-E1 osteoblasts was used for SEM. After rin-
sing with PBS, MC3T3-E1 osteoblasts were fixed with
4% paraformaldehyde (Merck), 1% glutaraldehyde
(Merck), and 0.1 M sodium cacodylate (SERVA
Electrophoresis GmbH, Heidelberg, Germany) in PBS
for 30 min. MC3T3-E1 osteoblasts were rinsed with
PBS, dehydrated in 70%, 80%, 90%, and 96% alcohol
and air-dried. Discs were sputtered with gold using a
sputter coater (Edwards S150B, Crawley, UK) to
visualize cells by SEM.
Proliferation
Adhered MC3T3-E1 osteoblasts were lysed using a cell
lysis buffer as described under bone cell culture. Cell
lysate was processed according to the protocol of the
CyQuant cell proliferation assay (Molecular Probes,
Carlsbad, CA). The cell proliferation assay measures
DNA concentration, which is proportionally related
to cell number. Fluorescence intensity was measured
using a fluorescence microplate reader (Wallac 1420
Victor 2, Perkin Elmer, Boston, MA, USA) with exci-
tation at 485 nm and emission detection at 530 nm.
A standard curve for fluorescence intensity versus
DNA concentration in ng/mL was generated.
Differentiation
ALP activity was measured in the cell lysate using
p-nitrophenylphosphate (Merck) as a substrate at pH
10.3, according to the method described by Bessey
et al.24 Plates were read at 405 nm. ALP activity was
expressed as nmol p-nitrophenylphosphate/mg protein.
Total cellular protein content was measured with a
bicinchoninic acid protein assay (Pierce, Rockford,
IL). Plates were read at 540–590 nm.
Cytokine production
Cytokines IL-1b, IL-6, and TNF-a were quantitatively
determined in culture medium using ELISA
van Hove et al.
(Quantikine, R&D Systems, Minneapolis, MN, USA).
Cytokine concentrations were measured according to
the manufacturers’ instructions. The minimum detect-
able dose was 3.0 pg/mL for IL-1b, 1.6 pg/mL for IL-6,
and 5.1 pg/mL for TNF-a.
Statistical analysis
Statistical analysis was performed using Kyplot
(KyensLab Inc., Tokyo, Japan). Data obtained from
separate experiments were pooled and expressed as
mean  standard error of the mean. Data were ana-
lyzed using a paired t-test. Differences were considered
statistically significant if p < 0.050.
Results
Scanning electron microscopy
SEM analysis of CoCrMo discs showed a smooth sur-
face with a spherical pattern due to polishing of irregu-
larities (Figure 1(A)). TiN discs showed a smooth
surface with circular tracks due to the production pro-
cess (Figure 1(B)). Porous-coated CoCrMo discs
showed a multilayered surface of smooth, reflectant
beads with a diameter of 200–250 mm (Figure 1(C)).
Porous-coated TiN discs showed a multilayered surface
of beads with a diameter of 300 mm, which were less
Figure 1. SEM images of discs with (A) CoCrMo surface, (B) TiN surface, (C) porous-coated CoCrMo surface, and
(D) porous-coated TiN surface.
Scale bar, 200 mm. Magnification, 100. SEM: scanning electron microscope; CoCrMo: cobalt–chromium–molybdenum; and TiN:
titanium–nitride.
Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016
281
smooth and reflectant indicating the presence of TiN
coating (Figure 1(D)).
On day 1 of culture, MC3T3-E1 osteoblasts were
present in similar density and orientation on both
smooth CoCrMo and TiN surfaces. On day 2,
MC3T3-E1 osteoblasts were present in a low density
and were connected to each other with dendrites on
smooth CoCrMo (Figure 2(A)). On day 4, osteoblast
density on smooth CoCrMo seemed to increase (Figure
2(C)). On day 2 of culture on smooth TiN, MC3T3-E1
osteoblasts aligned with the circular surface pattern
(Figure 2(B)). On day 4, osteoblasts on smooth TiN
showed similar density and alignment as on day 2
(Figure 2(D)). On day 7, osteoblasts on smooth TiN
showed increased density with similar alignment com-
pared with day 4. Unfortunately, cells were lost on
smooth CoCrMo on day 7 and on porous surfaces at
each time point.
Proliferation
Proliferation of MC3T3-E1 osteoblasts on porous
CoCrMo increased from day 1 to 4 (p ¼ 0.01), but
decreased from day 4 to 7 (p ¼ 0.03). Increased prolif-
eration was seen on porous TiN from day 1 to 7
(p ¼ 0.003). Proliferation on porous TiN was higher
than that on porous CoCrMo on day 2 (p < 0.01) and
day 7 (p < 0.01) (Figure 3(A), Table 2).
282
Proliferation increased on both smooth CoCrMo
(p ¼ 0.03) and TiN (p ¼ 0.005) from day 1 to 2.
Proliferation from day 1 to 7 was 1.6–2.0-fold higher
on smooth TiN than that on smooth CoCrMo.
Proliferation was similar on smooth TiN and polystyr-
ene from day 1 to 4 (Figure 3(B), Table 3).
Differentiation
On porous CoCrMo, MC3T3-E1 osteoblasts showed
increased ALP activity from day 1 to 2 (p ¼ 0.04),
which decreased thereafter (p ¼ 0.002). On porous
TiN, there was no change in ALP activity from day 1
to 2, but it decreased by 1.8-fold from day 2 to 7
(p < 0.001). MC3T3-E1 osteoblasts on porous TiN
showed lower ALP activity on day 4 (p < 0.01) and 7
(p ¼ 0.04) compared with porous CoCrMo
(Figure 4(A), Table 2).
On smooth CoCrMo, MC3T3-E1 osteoblasts
showed constant ALP activity levels on day 1 to 7,
while on smooth TiN, ALP activity remained constant
on day 1 to 4, but decreased thereafter by 1.8-fold on
day 7 (p ¼ 0.04). Only on day 4, MC3T3-E1 osteoblasts
on TiN showed higher ALP activity compared with
CoCrMo (Figure 4(B), Table 3).
MC3T3-E1 osteoblasts showed higher ALP activity
on smooth CoCrMo on day 1 to 7 and on smooth TiN
on day 1 to 4 (p < 0.02) as compared with polystyrene.
Figure 2. SEM images of MC3T3-E1 osteoblasts on CoCrMo and TiN after 2 and 4 days of culture. Osteoblasts on (A) CoCrMo, day
2; (B) TiN, day 2; (C) CoCrMo, day 4; and (D) TiN, day 4.
Scale bar, 100 mm. Magnification, 250. SEM: scanning electron microscope; CoCrMo: cobalt–chromium–molybdenum; and TiN:
titanium–nitride.
Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016
Journal of Biomaterials Applications 28(2)
On day 7, ALP activity was marginally higher on
smooth TiN compared with polystyrene (p ¼ 0.06)
(Figure 4(B)).
Both porous and smooth CoCrMo and TiN showed
an increase in ALP activity on day 1 to 2 which
decreased thereafter, except for smooth TiN where
ALP activity decreased after day 4 (Figure 4).
Cytokine production
On smooth CoCrMo and TiN, the cytokine IL-1b was
detected in medium of three out of four cultures on day
1. From day 2 to 7, IL-1b was detected in medium of
one out of four cultures on smooth CoCrMo on day 4
and on smooth TiN on day 2 (Table 4). IL-1b was only
detected in one out of four cultures on polystyrene on
day 2 (Table 4). TNF-a was not detected in any culture.
On porous CoCrMo, IL-1b was detected in medium
of two out of four cultures on day 1 and in one out of
four cultures on day 7 (Table 4). On porous TiN, the
cytokine IL-1b was detected in medium of one out of
four cultures on day 1 and 2, and in two out of four
cultures on day 7 (Table 4).
IL-6 production by MC3T3-E1 osteoblasts cultured
on porous CoCrMo and TiN remained constant from
day 1 to 7, but was higher on porous CoCrMo
compared with porous TiN (p < 0.05) (Figure 5(A),
Table 2).
van Hove et al. 283

(A) Porous differentiation, and IL-6 production were found on

4000 CoCrMo smooth surfaces compared with porous surfaces.
*
TiN
§
*
DNA, ng/ml

3000
Discussion
2000 Total knee arthroplasty can relieve pain and disability
resulting from severe osteoarthritis of the knee with a
1000 high rate of success.1,2 Questions have been raised
about the long-term durability of cemented fixation in
0 younger patients, leading to the design of uncemented
1 2 4 7
Days total knee arthroplasty to improve implant longevity.7,8
Smooth
The uncemented prosthesis fixates to bone by osteocon-
(B) duction and osseointegration resulting in bone
4000 * * #* CoCrMo ingrowth into the porous implant surface, thereby pro-
* *
* TiN
Polystyrene
viding a solid implant fixation.6 However, uncemented
3000 total knee arthroplasty shows higher rates of early fail-
DNA, ng/ml

* * ure than cemented total knee arthroplasty due to asep-

2000
1000
0 growth and metabolism.12–14 Sometimes implants are
1 2 4 7
Days
Figure 3. Proliferation of MC3T3-E1 osteoblasts cultured on
smooth or porous CoCrMo and TiN as well as polystyrene after
1, 2, 4, and 7 days. (A) Porous-coated CoCrMo and TiN and
(B) smooth CoCrMo, smooth TiN, and polystyrene.
CoCrMo: cobalt–chromium–molybdenum and TiN: titanium–
nitride. Data are presented as mean  SEM. *Significantly dif-
ferent from CoCrMo, p < 0.05. #Significantly different from
TiN, p < 0.05. §Difference approaching significance from
CoCrMo, p < 0.07.
IL-6 production by MC3T3-E1 osteoblasts on
smooth CoCrMo first decreased from day 1 to 4
(p ¼ 0.01), but increased thereafter (p ¼ 0.02). MC3T3-
E1 osteoblasts on smooth TiN showed constant IL-6
production from day 1 to 7. IL-6 production on smooth
TiN was higher than that on smooth CoCrMo on day 2
(p < 0.01) and day 4 (p ¼ 0.04) (Figure 5(B), Table 3).
IL-6 production by osteoblasts expressed per amount
of DNA did not show a difference between smooth TiN
and CoCrMo (data not shown). Compared with poly-
styrene, IL-6 production was lower on day 1 to 7 by
cells grown on smooth CoCrMo and TiN (p < 0.001)
(Figure 5(B)).
In summary, higher MC3T3-E1 osteoblast prolifer-
ation on day 2 and 7, lower differentiation on day 2 and
7, and lower IL-6 production were found on porous
TiN compared with porous CoCrMo (Table 2). On
smooth surfaces, higher proliferation, similar differen-
tiation, and similar IL-6 production by MC3T3-E1
osteoblasts on TiN compared with CoCrMo
were found (Table 3). Increased proliferation,
tic loosening.4,5 In an in vitro mouse model for aseptic
loosening, cytokines IL-1b, IL-6, and TNF-a induce
osteoclast differentiation.22 Additionally, chromium
and cobalt ions have been shown to diminish osteoblast
coated with a TiN layer to enhance implant hardness
and wear resistance and lower the friction coefficient of
the articular surface, in order to improve long-term
implant survival.10 Whether TiN also affects osteocon-
duction by osteoblasts is unknown. We hypothesized
that MC3T3-E1 osteoblasts on porous or smooth TiN
surface show increased proliferation and differentiation
and decreased cytokine production compared with
osteoblasts on CoCrMo surface.
This study reveals that cultured MC3T3-E1 osteo-
blasts show increased proliferation on both porous and
smooth TiN compared with CoCrMo. Furthermore,
osteoblast proliferation on smooth TiN was similar to
that on polystyrene, suggesting favorable properties of
TiN over CoCrMo with regard to cell proliferation.
Osteoblast proliferation did not rise exponentially
from day 4 to 7, which is unexpected given the rise in
proliferation from day 1 to 2 and the lack of a rise in
differentiation in all groups. It is likely that MC3T3-E1
osteoblasts reached confluency on smooth surfaces
which inhibits further proliferation.
Differentiation of MC3T3-E1 osteoblasts cultured
on porous TiN, as indicated by ALP activity, was
lower compared with porous CoCrMo. ALP activity
was decreased on both types of surfaces on day 2 to
7. Cultured MC3T3-E1 osteoblasts in the growing state
are known to have very low ALP enzyme activity, but
ALP enzyme activity increases after cells reach con-
fluency.25 In addition, ALP activity increases signifi-
cantly with the onset of growth arrest of MC3T3-E1
osteoblasts.26 On porous surfaces, MC3T3-E1 osteo-
blasts likely did not reach growth arrest, which could
explain the low ALP activity of these cells on both

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

284 Journal of Biomaterials Applications 28(2)

Table 2. Proliferation, differentiation, and IL-6 production of MC3T3-E1 osteoblasts cultured on porous CoCrMo and TiN for 1, 2,
4, or 7 days.

Porous

DNA (103 ng/mL) ALP (nmol/mg) IL-6 (pg/mL)

Day CoCrMo TiN p CoCrMo TiN p CoCrMo TiN p

1 1.29 (0.09) 1.11 (0.10) 0.31 0.21 (0.03) 0.20 (0.03) 0.40 100.1 (19.7) 39.1 (8.7) 0.04
2 1.63 (0.25) 2.74 (0.08) <0.01 0.28 (0.02) 0.24 (0.02) <0.01 74.1 (16.4) 31.9 (5.3) 0.03
4 1.97 (0.23) 2.63 (0.42) 0.07 0.23 (0.02) 0.20 (0.01) 0.16 74.1 (10.8) 18.5 (5.1) <0.01
7 1.45 (0.27) 3.19 (0.37) <0.01 0.17 (0.03) 0.13 (0.02) 0.04 129.0 (31.7) 30.2 (13.3) 0.01
CoCrMo: cobalt–chromium–molybdenum; TiN: titanium–nitride; ALP: alkaline phosphatase; and IL-6: interleukin-6.
Data are presented as mean  SEM.

Table 3. Proliferation, differentiation, and IL-6 production of MC3T3-E1 osteoblasts cultured on smooth CoCrMo and TiN for 1, 2,
4, or 7 days.

Smooth

DNA (103 ng/mL) ALP (nmol/mg) IL-6 (pg/mL)

Day CoCrMo TiN p CoCrMo TiN p CoCrMo TiN p

1 1.22 (0.08) 2.12 (0.37) 0.04 0.32 (0.06) 0.35 (0.08) 0.60 189.9 (23.3) 156.0 (32.0) 0.37
2 2.21 (0.36) 3.70 (0.14) <0.01 0.38 (0.02) 0.40 (0.02) 0.46 143.1 (11.3) 208.0 (7.9) <0.01
4 1.89 (0.45) 3.04 (0.35) 0.04 0.33 (0.03) 0.40 (0.03) 0.03 83.7 (13.5) 154.5 (30.3) 0.04
7 1.81 (0.41) 3.62 (0.17) <0.01 0.30 (0.04) 0.23 (0.05) 0.03 175.3 (19.3) 151.0 (39.5) 0.50
CoCrMo: cobalt–chromium–molybdenum; TiN: titanium–nitride; ALP: alkaline phosphatase; and IL-6: interleukin-6.
Data are presented as mean  SEM.

porous surfaces. Compared with polystyrene, MC3T3-
E1 osteoblasts cultured on smooth TiN and CoCrMo
showed higher ALP activity, suggesting increased
MC3T3-E1 osteoblast differentiation on these implant
surfaces.
In bone, the cytokine IL-6 is mainly produced by
osteoblasts and stromal cells.27 IL-6 stimulates osteo-
clast differentiation in a synergistic fashion with TNF-a
and IL-1b, which results in a net increase of RANKL
activity, leading to stimulation of osteoclast precursor
differentiation in mature osteoclasts.20,22 We found that
IL-6 production by MC3T3-E1 osteoblasts was lower
on porous TiN than that on porous CoCrMo, which
might suggest an inhibitory effect of TiN on osteoclas-
togenesis or an absence of stimulation of osteoclasto-
genesis by TiN. The cytokines IL-1, IL-6, and TNF-a
synergistically stimulate osteoclast differentiation
in vitro, although genetic knock out of these cytokines
does not alter osteolysis in mice.22 Therefore, the dif-
ferences found in ALP activity and cytokine production
by MC3T3-E1 osteoblasts cultured on the TiN and
CoCrMo surfaces might not be physiologically relevant
and caution is needed when extrapolating our conclu-
sions to the in vivo situation.
IL-6 production was higher on smooth TiN com-
pared with smooth CoCrMo at day 2 and 4, which
contradicts the lower IL-6 production on porous TiN
compared with CoCrMo. IL-6 production was
expressed in pg/ng DNA per disc to correct for possible
influences of differences in cell number on IL-6 produc-
tion. IL-6 production, expressed in pg/mL culture
medium as well as in pg/ng DNA, was lower on
porous TiN compared with porous CoCrMo. In con-
trast, IL-6 production expressed in pg/mL culture
medium, but not in pg/ng DNA, was higher on day 2
and 4 on smooth TiN compared with smooth CoCrMo.
Recently, it has been proposed that IL-6 and
IL-6-type cytokines have a dual role, i.e. besides their
stimulatory effect on bone resorption, they also exert a
positive effect on bone formation.28,29 The stimulatory
effect of IL-6-type cytokines on bone formation has
been suggested to depend on the differentiation stage
of osteoblasts.30 IL-6-type cytokines would stimulate
the early stages of osteogenic differentiation of precur-
sor cells, but prevent further stimulation of more
mature cells.30 Since MC3T3-E1 osteoblasts are rather
mature cells, high IL-6 levels would prevent these cells
from further proliferation and differentiation.

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

van Hove et al. 285

The cytokines TNF-a and IL-1b play an important
role in osteoclastogenesis in a synergistic fashion
with IL-6.20 In addition, cytokines TNF-a and
IL-1b are upregulated in aseptic loosening
in vivo.31–33 IL-1b was only detected in medium of a
(A)
0.5
ALP activity nmol/mg protein
few MC3T3-E1 osteoblast cultures on porous and
smooth CoCrMo and TiN surfaces, while TNF-a was
not detected, indicating that IL-1b and TNF-a produc-
tion by osteoblasts on these surfaces is limited and unli-
kely to lead to dramatic changes in osteoclastic bone
resorption.
SEM showed differences between the smooth
Porous CoCrMo discs (DePuy-Johnson and Johnson) and
CoCrMo
smooth TiN discs (Implantcast) provided by the two
TiN manufacturers. The smooth CoCrMo surface was

0.4 metallurgically polished, whereas the smooth TiN sur-

*
face showed a circular pattern with approximately
0.3
100 mm distance between the tracks. On day 2 and 4,
0.2 * the MC3T3-E1 osteoblasts aligned the circular pattern
of the smooth TiN surface. Topographic features with
0.1 dimensions approximating the cell size have been
reported to have strong effects on cell guidance and
0
1 2 4 7 shape regulation.34 A comparative study on micro-
Days grooved hydroxyapatite and titanium did reveal a dif-
(B) Smooth ference in shape regulation, but not contact guidance of
0.5
osteoblast-like cells.34 This suggests that the surface
CoCrMo
chemistry affects osteoblast behavior.34 Since our
ALP activity nmol/mg protein

* TiN
0.4 Polystyrene SEM images of the porous material surfaces suggest
similar topography, and because proliferation, differen-
0.3 * tiation, and cytokine production were similar to find-
*# *# ings on smooth surfaces, we suggest that the observed
0.2 *#
*¥ differences in proliferation, differentiation, and cyto-
0.1 kine production on smooth surfaces were due to surface
chemistry.
0
1 2 4 7
Kubies et al.35 reported on human osteoblast prolif-
Days eration and the expression of TNF-a and bone ALP
amongst other tissue mediators by these cells after cul-
Figure 4. ALP activity in MC3T3-E1 osteoblasts cultured on ture on commercially available implant materials such
smooth or porous CoCrMo or TiN as well as polystyrene after
as titanium and titanium–aluminum–vanadium alloy
1, 2, 4, and 7 days. (A) Porous-coated CoCrMo and TiN and
with various surface treatments, CoCrMo alloy, zirco-
(B) smooth CoCrMo, smooth TiN, and polystyrene.
CoCrMo: cobalt–chromium–molybdenum and TiN: titanium– nium oxide ceramics, stainless steel, polyethylene, and
nitride. Data are presented as mean  SEM. *Significantly dif- carbon/carbon composite. After 4 days of culture, the
ferent from CoCrMo, p < 0.05. #Significantly different from number of proliferating cells was determined and pol-
TiN, p < 0.05. ¥Difference approaching significance from TiN, ished CoCrMo was amongst the least suitable surfaces
p < 0.06. for osteoblast culture.35 Our results showed higher

Table 4. Cytokine IL-1b production by MC3T3-E1 osteoblasts cultured on porous and smooth CoCrMo and TiN surfaces and
polystyrene for 1, 2, 4, or 7 days.

IL-1b (pg/mL)

CoCrMo TiN Polystyrene

Day smooth porous smooth Porous

1 3.1 – 2.0 1.1 – – 2.4 3.4 1.6 – 7.6 4.8 2.0 – – – – – – –

2 – – – – – – – – 6.8 – – – 0.3 – – – 3.4 – – –
4 – – – 4.6 – – – – – – – – – – – – – – – –
7 – – – – – – – 1.2 – – – – – 2.7 – 1.2 – – – –
CoCrMo: cobalt–chromium–molybdenum; TiN: titanium–nitride; and IL-1b: interleukin-1b.

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

286 Journal of Biomaterials Applications 28(2)

proliferation on smooth TiN than that on CoCrMo.
This suggests that a TiN coating makes a CoCrMo
surface more suitable for osteoblast culture.
We cultured MC3T3-E1 osteoblasts on TiN and
CoCrMo surfaces as used in total knee arthroplasty
and found differences in proliferation, differentiation,
and cytokine production. Bone formation around the
implant, is dependent on biophysical stimuli and soft-
tissue formation.36 Osteoblasts arrive in a later stage of
interface tissue formation,36 but for osseointegration,
osteoblast proliferation and differentiation along the
implant surface is needed. Therefore, our results on
in vitro proliferation, differentiation, and cytokine pro-
duction of osteoblasts on implant surfaces might have
implications in vivo.
In conclusion, MC3T3-E1 osteoblasts cultured on
porous or smooth TiN surfaces showed similar differ-
entiation but increased proliferation compared with
CoCrMo surfaces. IL-6 production by osteoblasts on
TiN surfaces was decreased on porous surfaces and on
smooth surfaces. The clinical implications of these
Porous
(A)
500
400
IL-6, pg/ml
findings might be decreased bone resorption and
enhanced osteoconduction leading to improved
osseointegration. Further research on osteoconduction
and clinical research on osseointegration may reveal a
lower risk of early aseptic loosening of TiN-coated
implants as compared with CoCrMo implants.
Funding
This study received financial assistance from the Foundation
of Orthopaedic Research Spaarne Hospital, Hoofddorp,
The Netherlands.
Acknowledgments
The authors thank A. Werner for excellent technical support
with SEM and D. Holmes for critically reading the manu-
script. All authors have no conflict of interest.
References
1. Gandhi R, Tsvetkov D, Davey JR, et al. Survival and
clinical function of cemented and uncemented prostheses
in total knee replacement: a meta-analysis. J Bone Joint
Surg (Br) 2009; 91: 889–895.
2. Norman-Taylor FH, Palmer CR and Villar RN. Quality-
CoCrMo
TiN
of-life improvement compared after hip and knee replace-
ment. J Bone Joint Surg (Br) 1996; 78–B: 74–77.

3. Barrack RL, Nakamura SJ, Hopkins SG, et al. Early

300
failure of cementless mobile-bearing total knee arthro-
plasty. J Arthroplasty 2004; 19: 101–106.
200 4. Fehring TK, Odum S, Griffin WL, et al. Early failures in
*
* total knee arthroplasty. Clin Orthop Relat Res 2001; 392:
100 * *
315–318.
5. Sharkey PF, Hozack WJ, Rothman RH, et al. Why are
0
1 2 4 7 total knee arthroplasties failing today? Clin Orthop Relat
Days Res 2002; 404: 7–13.
Smooth
6. Albrektsson T and Johansson C. Osteoinduction, osteo-
(B) conduction and osseointegration. Eur Spine J 2001; 10:
*#
S96–S101.
500 CoCrMo
*# *# TiN
7. Naudie DD, Ammeen DJ, Engh GA, et al. Wear and
400 *# Polystyrene osteolysis around total knee arthroplasty. J Am Acad
IL-6, pg/ml

Orthop Surg 2007; 15: 53–64.

300 8. Bassett RW. Results of 1000 performance knees: cement-
* less versus cemented fixation. J Arthroplasty 1998; 13:
200 *
409–413.
100 9. Marti A. Cobalt-base alloys used in bone surgery. Injury
2000; 31(Suppl. 4): 18–21.
0 10. Wisbey A, Gregson PJ and Tuke M. Application of PVD
1 2 4 7
Days TiN coating to Co-Cr-Mo based surgical implants.
Biomaterials 1987; 8: 477–480.
Figure 5. IL-6 production by MC3T3-E1 osteoblasts cultured 11. Mezger PR and Creugers NHJ. Titanium nitride coatings
on smooth or porous CoCrMo and TiN as well as polystyrene in clinical dentistry. J Dent 1992; 20: 342–344.
after 1, 2, 4, and 7 days. (A) porous-coated CoCrMo and TiN and 12. Allen MJ, Myer BJ, Millett PJ, et al. The effects of par-
(B) smooth CoCrMo, smooth TiN, and polystyrene. ticulate cobalt, chromium and cobalt-chromium alloy on
IL-6: interleukin-6; CoCrMo: cobalt–chromium–molybdenum; human osteoblast-like cells in vitro. J Bone Joint Surg
and TiN: titanium–nitride. Data are presented as mean  (Br) 1997; 79: 475–482.
SEM. *Significantly different from CoCrMo, 13. Fleury C, Petit A, Mwale F, et al. Effect of cobalt
p < 0.05. #Significantly different from TiN, p < 0.05. and chromium ions on human MG-63 osteoblasts

Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016

van Hove et al.
in vitro: morphology, cytotoxicity, and oxidative stress.
Biomaterials 2006; 27: 3351–3360.
14. Wang JY, Wicklund BH, Gustilo RB, et al. Prosthetic
metals interfere with the functions of human osteo-
blast cells in vitro. Clin Orthop Relat Res 1997; 339:
216–226.
15. Bilbe G, Roberts E, Birch M, et al. PCR Phenotyping of
cytokines, growth factors and their receptors and bone
matrix proteins in human osteoblast-like cell lines. Bone
1996; 19: 437–445.
16. Gowen M, Chapman K, Littlewood A, et al. Production
of tumor necrosis factor by human osteoblasts is modu-
lated by other cytokines, but not by osteotropic hor-
mones. Endocrinology 1990; 126: 1250–1255.
17. Hanazawa S, Ohmori Y, Amano S, et al. Spontaneous
production of interleukin-1-like cytokine from a mouse
osteoblastic cell line (MC3T3-E1). Biochem Biophys Res
Commun 1985; 131: 774–779.
18. Ishimi Y, Miyaura C, Jin CH, et al. IL-6 is produced by
osteoblasts and induces bone resorption. J Immun 1990;
145: 3297–3303.
19. Passeri G, Girasole G, Manolagas SC, et al. Endogenous
production of tumor necrosis factor by primary cultures
of murine calvarial cells: influence on IL-6 production
and osteoclast development. Bone Miner 1994; 24:
109–126.
20. Kwan Tat S, Padrines M, Théoleyre S, et al. IL-6,
RANKL, TNF-alpha/IL-1: interrelations in bone resorp-
tion pathophysiology. Cytokine Growth Factor Rev 2004;
15: 49–60.
21. Ragab AA, Nalepka JL, Bi Y, et al. Cytokines synergis-
tically induce osteoclast differentiation: support by
immortalized or normal calvarial cells. Am J Physiol
Cell Physiol 2002; 283: C679–C687.
22. Taki N, Tatro JM, Lowe R, et al. Comparison of the
roles of IL-1, IL-6, and TNF-a in cell culture and
murine models of aseptic loosening. Bone 2007; 40:
1276–1283.
23. Mullender MG, Dijcks SJ, Bacabac RG, et al. Release of
nitric oxide, but not prostaglandin E2, by bone cells
depends on fluid flow frequency. J Orthop Res 2006; 24:
1170–1177.
24. Bessey OA, Lowry OH and Brock MJ. A method for the
rapid determination of alkaline phosphatase with five
Downloaded from jba.sagepub.com at UNIVERSITAT INTERNACIONAL DE CATALUNYA on March 10, 2016
287
cubic millimeters of serum. J Biol Chem 1946; 164:
331–339.
25. Sudo H, Kodama HA, Amagai Y, et al. In vitro differ-
entiation and calcification in a new clonal osteogenic cell
line derived from newborn mouse calvaria. J Cell Biol
1983; 96: 191–198.
26. Quarles LD, Yohay DA, Lever LW, et al. Distinct pro-
liferative and differentiated stages of murine MC3T3-E1
cells in culture: an in vitro model of osteoblast develop-
ment. J Bone Miner Res 1992; 6: 683–692.
27. Holt I, Davie WJ and Marshall MJ. Osteoclasts are not
the major source of interleukin-6 in mouse parietal bones.
Bone 1996; 18: 221–226.
28. Blanchard F, Duplomb L, Baud’huin M, et al. The dual
role of IL-6-type cytokines on bone remodeling and bone
tumors. Cytokine Growth Factor Rev 2009; 20: 19–28.
29. Franchimont N, Wertz S and Malaise M. Interleukin-6:
an osteotropic factor influencing bone formation? Bone
2005; 37: 601–606.
30. Malaval L and Aubin JE. Biphasic effects of leukemia
inhibitory factor on osteoblastic differentiation. J Cell
Biochem Suppl 2001; 36: 63–70.
31. Goodman SB, Huie P, Song Y, et al. Cellular profile and
cytokine production at prosthetic interfaces. Study of tis-
sues retrieved from revised hip and knee replacements.
J Bone Joint Surg (Br) 1998; 80: 531–539.
32. Hicks DG, Judkins AR, Sickel JZ, et al. Granular histio-
cytosis of pelvic lymph nodes following total hip arthro-
plasty. The presence of wear debris, cytokine production,
and immunologically activated macrophages. J Bone
Joint Surg (Am) 1996; 78: 482–496.
33. Stea S, Visentin M, Granchi D, et al. Cytokines and oste-
olysis around total hip prostheses. Cytokine 2000; 12:
1575–1579.
34. Lu X and Leng Y. Quantitative analysis of osteoblast
behavior on microgrooved hydroxyapatite and titanium
substrata. J Biomed Mater Res Part A 2003; 66: 677–687.
35. Kubies D, Himmlová L, Riedel T, et al. The interaction
of osteoblasts with bone-implant materials: 1. The effect
of physicochemical surface properties of implant mater-
ials. Physiol Res 2011; 60: 95–111.
36. Prendergast PJ, Huiskes R and Søballe K. Biophysical
stimuli on cells during tissue differentiation at implant
surfaces. J Biomechanics 1997; 30: 539–548.