The relationship between the nanostructure of titanium surfaces and

bacterial attachment
Sabrina D. Puckett
1
, Erik Taylor
1
, Theresa Raimondo, Thomas J. Webster
*
Division of Engineering and Department of Orthopaedics, Brown University, Providence, RI 02917, USA
a r t i c l e i n f o
Article history:
Received 10 August 2009
Accepted 21 September 2009
Available online 30 October 2009
Keywords:
Titanium
Nanotopography
Bacteria
Adhesion
Fibronectin
a b s t r a c t
Infection of an orthopedic prosthesis is undesirable and causes a decrease in the success rate of an
implant. Reducing the adhesion of a broad range of bacteria could be an attractive means to decrease
infection and allow for subsequent appropriate tissue integration with the biomaterial surface. In this in
vitro study, nanometer sized topographical features of titanium (Ti) surfaces, which have been previously
shown to enhance select protein adsorption and subsequent osteoblast (bone-forming cell) functions,
were investigated as a means to also reduce bacteria adhesion. This study examined the adhesion of
Staphylococcus aureus, Staphylococcus epidermidis, and Pseudomonas aeruginosa on conventional Ti,
nanorough Ti produced by electron beam evaporation, and nanotubular and nanotextured Ti produced
by two different anodization processes. This study found that compared to conventional (nano-smooth)
Ti, the nanorough Ti surfaces produced by electron beam evaporation decreased the adherence of all of
the aforementioned bacteria the most. The conventional and nanorough Ti surfaces were found to have
crystalline TiO
2
while the nanotubular and nanotextured Ti surfaces were found to be amorphous. The
surface chemistries were similar for the conventional and nanorough Ti while the anodized Ti surfaces
contained uorine. Therefore, the results of this study in vitro study demonstrated that certain nano-
meter sized Ti topographies may be useful for reducing bacteria adhesion while promoting bone tissue
formation and, thus, should be further studied for improving the efcacy of Ti-based orthopedic
implants.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
Infection has been reported on an array of implantable devices
including central venous catheters and needleless connectors,
endotracheal tubes, intrauterine devices, mechanical heart valves,
pacemakers, peritoneal dialysis catheters, tympanostomy tubes,
voice prostheses, orthopedic joint prosthetics, and percutaneous
orthopedic devices (including external xators and bone-anchored
amputee prosthetics) [1,2]. Prosthetic joint replacements are being
used with increasing frequency to alleviate pain, to promote
mobility, and to improve the quality of life. Yet, such implantation
suffers from the added risk of infection occurring in about 1.52.5%
of all hip and knee arthroplasties resulting in failure of the device
and, thus, a high need for revision surgery [3]. Joint prosthetic
infection costs about $50,000 U.S. dollars per episode while the
associated mortality rate may be as high as 2.5% [3]. In addition, if
the infection persists into the deep tissue, amputation may also be
required.
To put these percentages into perspective, it is important to note
that in 2004, 265,441 total hip arthroplasties (THA) and 496,018
total knee arthroplasties (TKA) were performed in the U.S. alone [4].
Of these, an estimated 3352 THA (1.23%) and 5838 TKA (1.21%) were
treated for infection [4]. It was further revealed that of these total
arthroplasties performed in 2004, 38,629 THA and 36,425 TKA
were due to revision surgeries [4]. Revision surgeries are a result of
implant failure that can be caused from stressstrain imbalances,
implant migration, wear debris, lack of integration, and infection
[1,58]. Of these implant failure modes, about 8% of THA and 15% of
TKA revision surgeries were a direct result of infection [4].
In addition to orthopedic joint prosthetics, percutaneous
implant devices suffer from a lack of successful skin integration
around the biomaterial exit site due to bacteria invasion [9].
As bacteria colonize either the implant surface or adjacent
damaged tissue sites, biomaterial exit sites become the gateway to
infection [1012], possibly leading to bacteria spreading internally
and causing osteomyelitis [12,13]. According to previous studies,
the occurrence of osteomyelitis after insertion of an external xator
can be up to 4% [1012]. In addition to osteomyelitis, infection leads
to bone implant loosening [11] and fracture malunion or nonunion,
leading to early failure of the device.
* Corresponding author. Tel.: 1 401 523 3802; fax: 1 401 523 9107.
E-mail address: thomas_webster@brown.edu (T.J. Webster).
1
Contributed equally.
Contents lists available at ScienceDirect
Biomaterials
j ournal homepage: www. el sevi er. com/ l ocat e/ bi omat eri al s
0142-9612/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2009.09.081
Biomaterials 31 (2010) 706713
Reduction of microbial adhesion to an implant (without the use
of drugs) could be an attractive method for reducing infection.
Planktonic (suspended) bacteria present in the body can be cleared
by host defense mechanisms and are more susceptible to antibiotic
treatment [14]. However, once bacteria bind to the biomaterial
surface, changes in their functions occur. More specically, gene
expression changes, growth rates are altered, host defense mech-
anisms are no longer able to remove them from the body, and
formation of an antibiotic resistant biolm occurs [1,1416].
Development of this biolm is responsible for many chronic
infections [1,1416] and prevents proper integration of the implant
to the surrounding tissue. In addition, antibiotic resistant strains
cannot be treated by antibiotic therapy after adhesion to the
prosthetic surface. Multiple antibiotic resistant strains including
Staphylococcus aureus (S. aureus) and Staphylococcus epidermidis
(S. epidermidis) are well documented in the clinical orthopedic
setting [17,18]. Thus, it can be argued that the prevention of bacteria
adhesion without drugs may be one of the best ways to reduce
orthopedic implant infection [18].
Along these lines, altering surface roughness of an implant
material fromone that possesses conventional, micron size features
to one that possesses nanometer size features has been shown to
enhance certain cellular, such as osteoblasts (bone-forming cells)
[1924], adhesion and subsequent cellular functions (such as
calcium deposition) while simultaneously decreasing competitive
cell, such as broblast (cells that create the brous tissue around an
implanted material preventing proper bone integration) function
[25,26]. Research has specically demonstrated that nanorough Ti
(created through electron beam evaporation) [27] and nanotubular
and nanotextured Ti (created through anodization) can enhance
osteoblast adhesion and other functions (such as alkaline phos-
phatase synthesis, calcium deposition, and collagen secretion)
compared to their micron nano-smooth counterparts [28].
Increased select protein adsorption on such Ti surfaces containing
nanofeatures has been correlated to the improved functions of
osteoblasts [2931]. Previous studies have also shown that by
varying the surface roughness of a biomaterial, bacteria adhesion
decreases [32]. However, more research is required to understand
the underlying factors for such a phenomenon and translating such
results to metals commonly used in orthopedics.
Based on the evidence from these previous studies, this study
explored the adhesion of multiple bacteria species well-known to
lead to orthopedic implant infection on nanotubular, nanotextured,
nanorough, and conventional Ti. Moreover, initial protein adsorp-
tion events were explored that may explain such bacteria adhesion
trends. Specically, gram positive S. aureus and S. epidermidis along
with gram negative Pseudomonas aeruginosa (P. aeruginosa) were
examined since these strains have been shown to be clinically
prevalent in orthopedic prosthetic infections [18]. With selectively
improved bone cell responses, as already demonstrated, and
decreased bacteria adhesion, integration between bone and the
implant surface would be promoted, thus, improving the success
rate of orthopedic prosthetics.
2. Materials and methods
2.1. 2.1. Titanium substrates
Titanium(Ti) foils (100 100 1 mm; 99.2%pure; AlfaAesar, WardHill, MA, USA)
were cut into 10 10 mm squares using a shear cutter. All substrates were ultra-
sonically cleaned with a diluted cleaning solution (Branson, Dabury, CT, USA) for
20 min followed by sonication in acetone, 70% ethanol, and deionized water (DI) for
10 min. Substrates were then dried in an oven (VWR) at 40

C for 15 min. Some of
these substrates served as the conventional, unmodied Ti substrates used
throughout thepresent work, whileothers wereusedbelowinSection2.2for electron
beam evaporation or Section 2.3 for anodization. Prior to cell culture experiments,
these substrates were sterilized in a steamautoclave at 120

C and 17 psi for 30 min.
2.2. Electron beam evaporation
A Temescal Electron Beam Evaporator (Reston, VA, USA) was used to create
nanorough Ti substrates. Electron beam evaporation concentrates a large amount of
heat produced by high energy electron beam bombardment on the source material
to be deposited, in this case 99.995% pure Ti pellets (Kamis, Mahopac Falls, NY, USA).
The electron beam is generated by an electron gun that uses the thermoionic
emission of electrons produced by an incandescent lament. A magnet focuses and
bends the electron trajectory so that the beam is accelerated towards a graphite
crucible (Lesker, Clairton, PA, USA) containing the source material. As the beam
rotates and hits the surface of the source material, heating and vaporization occur.
The vapor ow then condenses onto the substrate surface located at the top of the
vacuum chamber. In this study, Ti was deposited onto the Ti substrates at a rate of
3.5 /s and at a thickness of 500 nm. Following deposition, the nanorough Ti
samples were rinsed thoroughly with DI, air dried, and sterilized in a steam auto-
clave at 120

C and 17 psi for 30 min.
2.3. Anodization
Prior to anodization, Ti substrates were immersed in a dilute acidic mixture of
nitric acid (HNO
3
) and hydrouoric acid (HF) for 5 min to remove the thin oxide
layer that spontaneously forms on Ti while in the presence of air. Titania nanotube
arrays were then formed into the Ti surface by anodization. Anodization is an
electrolytic passivation process used to increase the thickness of the natural oxide
layer on metal surfaces, in this case Ti. This process was conducted with a DC
powered electrochemical cell which consisted of a two electrode conguration:
a platinummesh which served as the cathode and Ti foils which served as the anode.
To fabricate nanotubular Ti surfaces, the anodization process took place in 1.5 wt%
HF for 10 min at a constant voltage of 20 V [28]. To fabricate nanotextured Ti
surfaces, the anodization process took place in 0.5 wt% HF for one minute at
a constant voltage of 20 V [28]. The nanotubular and nanotextured Ti substrates
were rinsed with large amounts of DI immediately after anodization, air dried, and
sterilized under ultraviolet light for 3 h per substrate side.
2.4. Surface characterization
The conventional, nanorough, nanotubular, and nanotextured Ti substrates were
characterized for chemistry, surface roughness, and crystallinity. For chemical
analysis, X-ray photoelectron spectroscopy (XPS) was performed using a Perkin
Elmer 5500 Multitechnique Surface Analyzer System (Waltham, MA, USA). XPS was
specically used to determine the composition of the surface oxide formed on the Ti
substrates. An aluminum K-alpha monochromatized X-ray source was used to
stimulate photoemission of the inner shell electrons on the Ti surface. The energy
from these electrons was then recorded and analyzed for identication purposed
concerning chemical composition of the nanotubular, nanorough, and unmodied Ti
substrates.
For qualitative surface roughness analysis, scanning electron microscopy (SEM)
was performed on the conventional, nanorough, nanotubular, and nanotextured Ti
substrates. Images were taken using a LEO 1530VP SEM at varying magnications.
Digital images were created using secondary electrons collected with an in-lens
detector at an accelerating voltage of 3 kV for nanorough and conventional Ti
substrates and 5 kV for nanotubular and nanotextured Ti substrates.
Phase analysis was carried out by X-ray diffraction (XRD) analysis using a D500
Siemens Diffractometer (Bruker AXS, WI). Spectra were taken using a power supply
of 30.0 mA and 40.0 kV.
2.5. Surface energy and contact angles
Material surface energy and wettability were investigated with a drop shape
analysis system (EasyDrop, Kruss, Hamburg, Germany). The contact angle of 3 mL
sessile droplets was measured at two locations on each of the four samples (the
nanotubular, nanotextured, nanorough, and unmodied Ti). To determine surface
energy, three different liquid solvents, distilled water, glycerol, and polyethylene
glycol, were used. Measurements were taken 5 sec after placing the droplet on the
sample surface under ambient conditions. Drop shape analysis software (DSA1,
Kruss, Hamburg, Germany) was used to calculate surface energy by entering surface
tension and contact angle data into the OwensWendt equation:
1 g
1
cos q 2

g
d
s
g
d
1

g
p
c
g
p
1

q

(1)
Here, g
d
s
and g
p
s
are the respective dispersion and polar terms of the solid surface
tension, g
s
; g
d
1
and g
p
1
are the respective dispersion and polar terms of the liquid
surface tension, g
l
. Other theories were investigated (Fowkes and Zisman) but
results showed the same trends of surface energy as those obtained with the
OwensWendt model.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 707
2.6. Bacteria culture
Bacteria cell lines used in this study were S. epidermidis, P. aeruginosa, and
S. aureus obtained in freeze-dried form from the American Type Culture Collection
(35984, 25668, and 25923 respectively). The dry pellet was rehydrated in 6 mL of
Luria broth (LB) consisting of 10 g tryptone, 5 g yeast extract, and 5 g NaCl per liter
double distilled water with the pH adjusted to 7.4 (all chemicals were obtained from
Sigma Aldrich, St. Louis, MO, USA). The bacteria solutionwas agitated under standard
cell conditions (5% CO2/95% humidied air at 37

C) for 24 h until the stationary
phase was reached. The second passage of bacteria was diluted at a ratio of 1:200
into fresh LB and incubated until it reached stationary phase. The second passage
was then frozen in one part LB and one part glycerol (Sigma Aldrich) and stored at
18

C. All experiments were conducted from this frozen stock. One day before
bacterial seeding for experiments, a sterile 10 ml loop was used to withdraw bacteria
from the frozen stock and to inoculate a centrifuge tube with 3 mL of fresh LB.
2.7. Bacteria adhesion
Prior to seeding, sterilized substrates were placed into a standard 24-well
culture plate and were washed twice with phosphate buffer saline (PBS). Bacteria
were seeded on the substrates at a density of 1 10
7
bacteria/mL (as estimated by
the McFarland scale) by diluting the LB bacteria cultures to an optical density of
0.52 at 562 nm and then further diluted at a ratio of 1:90 in Dulbeccos Modied
Eagles Medium (DMEM, Hyclone; Logan, UT/USA) supplemented with 10% fetal
bovine serum (FBS, Hyclone), 1% penicillinstreptomycin (P/S, Hyclone), 50 mg/mL
L-ascorbate acid (Sigma Aldrich), and 10 mM b-glycerophosphate (Sigma Aldrich).
The bacteria were allowed to adhere for one hour under standard cell conditions
(5% CO
2
/95% humidied air at 37

C) with constant shaking at 200 rpm to prevent
settling of the cell solution. At the end of the prescribed time period, the substrates
were rinsed twice with Tris-buffered saline (TBS) comprised of 42 mM TrisHCl,
8 mM Tris Base, and 0.15 M NaCl (Sigma Aldrich) and then incubated for 15 min with
the BacLight Live/Dead solution (Life Technologies Corporation, Carlsbad, CA)
dissolved in TBS at the concentration recommended by the manufacturer. Substrates
were then rinsed twice with TBS and placed into a 50% glycerol solution in TBS prior
to imaging. Bacteria were then visualized and counted in situ using a Leica
DM5500 B uorescence microscope with image analysis software captured using
a Retiga 4000R camera. Adhesion experiments were run in duplicate and repeated
three different times per substrate type. Total bacteria colonies were determined by
summing the number of live and dead bacteria colonies found using Image J.
The data was represented by the mean value with the standard error of the mean
(SEM) noted. A students t-test was used to check statistical signicance between
means and p <0.1 was considered statistically signicant.
2.8. Fibronectin adsorption (ELISA)
The enzyme-linked immunosorbent assay (ELISA) is a well established proce-
dure for measuring the amount of protein, in this study bronectin, adsorbed to the
conventional, nanorough, nanotextured, and nanotubular Ti substrates. Substrates
were placed in a standard 24-well culture plate and immersed in 1 mL of Dulbeccos
Modied Eagles Medium (DMEM, Hyclone; Logan, UT, USA) supplemented with and
without 10% FBS and 1% P/S for 24 h at 37

C in 5% CO
2
/95% humidied air. After
rinsing in PBS, areas that did not adsorb proteins were blocked and incubated for one
hour in bovine serum albumin, BSA (2 wt% in PBS, Sigma Aldrich, MO, USA).
Substrates were again rinsed twice with PBS before bronectin was directly linked
with primary rabbit anti-bovine bronectin (AB2047, Millipore, CA, USA) at
a concentration of 6 mg/mL in 1% BSA for 1 h at 37

C in 5% CO
2
/95% humidied air.
After rinsing 3 times with 0.05% Tween 20 for 5 min with each rinse, the samples
were further incubated for another 1 h with a secondary goat anti-rabbit conjugated
with horseradish peroxidase (HRP, Bio-Rad, MD, USA) at a concentration of 10 mg/mL
in 1% BSA. Followed by another 3 rinses with 0.05% Tween 20 for 5 min with each
rinse, the amount of bronectin adsorbed to the surfaces was measured with an
ABTS substrate kit (Vector Laboratories, CA, USA) that reacted only with the HRP.
Light absorbance was measured at 405 nm on a spectrophotometer and analyzed
with computer software. The average adsorbance was subtracted by the average
adsorbance obtained fromthe negative controls soaked in DMEMwith no FBS or P/S.
ELISA was performed in duplicate and repeated three different times per substrate.
3. Results
3.1. Surface characterization
The unmodied titanium (Ti) as purchased from the vendor
possessed micron rough surface features as displayed under SEM
(Fig. 1(a)). After electron beam evaporation, the Ti substrates
possessed a high degree of nanometer surface features, thus,
creating a more nanometer rough surface topography (Fig. 1(b)).
Fig. 1. SEM micrographs of Ti before and after electron beam evaporation and anodization: (a) conventional Ti as purchased from the vendor; (b) nanorough Ti after electron beam
evaporation; (c) nanotextured Ti after anodization for 1 min in 0.5% HF at 20 V; (d) nanotubular Ti after anodization for 10 min in 1.5% HF at 20 V. Scale bars 200 nm.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 708
Completion of anodization for 1 min in 0.5% hydrouoric acid (HF)
at 20 V resulted in a Ti substrate containing nanotextured surface
features (Fig. 1(c)). Increasing the anodization time (10 min) and
concentration of HF (1.5%) resulted in a Ti surface that contained
nanotubular like structures with an inner diameter from 60 to
70 nm, as estimated from the SEM images (Fig. 1(d)).
One high resolution XPS spot was taken on each sample to
examine the Ti 2 p binding energy. Data indicated that other than
TiO
2
, noother Ti species, suchas TiOandTi
2
O
3
, werepresent (Table1).
XPS results also showed that for all sample types the outermost layer
of oxide contained O and Ti (Table 2). The nanotubular and nano-
textured Ti substrates also contained a small amount of F in the
outermost level, due to the anodization process involving the use of
HF (Table 2). In particular, nanotubular Ti contained a higher
percentage of uorine compared to the nanotextured Ti surfaces
(Table 2).
XRD spectra (data not shown) conrmed the presence of
amorphous titania (no anatase or rutile phase was observed) for the
nanotextured and nanotubular Ti substrates. XRD spectra also
conrmed the presence of crystalline titania for the nanorough and
conventional Ti substrates. More specically, the conventional Ti
contained rutile TiO
2
and no anatase TiO
2
, while the nanorough Ti
contained anatase TiO
2
and no rutile TiO
2
.
3.2. Surface energy and contact angles
Surface energy calculations from contact angle data indicated
that increasing surface roughness increased surface energy.
All nanofabricated surfaces (nanorough, nanotextured, and nano-
tubular Ti) had a surface energy signicantly higher than that of
unmodied, conventional Ti surfaces (Fig. 2). In addition, the
nanofabricated surfaces had a lower contact angle for each liquid
used to determine surface energy (Table 3), indicating increased
surface energy for such samples compared to the unmodied Ti.
Interestingly, among the nanofabricated surfaces, nanotextured
and nanotubular Ti had much lower contact angles for each liquid
used to determine surface energy compared to the nanorough Ti
(Table 3), indicating that nanotextured and nanotubular Ti surfaces
had the greatest surface energy of the substrates created in this
study.
3.3. Bacterial adhesion
Fig. 3 shows the total bacteria colonies, including both live and
dead bacteria, attached after 1 h on the substrates of interest to this
study. The results of this study revealed that bacteria adhered the
least to the nanorough Ti substrates. More specically, when
normalized to the projected surface area, there was a signicantly
lower attachment of colonies for all bacteria lines (S. aureus,
S. epidermidis, and P. aeruginosa) on the nanorough Ti substrates
compared to the conventional, nanotubular, and nanotextured Ti
substrates (Fig. 3). In addition, when further examining bacteria
behavior on the anodized Ti surfaces, results indicated that bacteria
signicantly adhered more to the nanotubular Ti compared to the
nanotextured Ti (Fig. 3). Fig. 4 qualitatively highlights the decreased
attachment of S. aureus, S. epidermidis, and P. aeruginosa on the
nanorough Ti substrates. Interestingly, data also indicated that the
nanotubular and nanotextured Ti substrates had the highest
number of bacterial colonies for all cell lines compared to
conventional and nanorough Ti substrates (Fig. 3). Fig. 4 also
visually highlights the signicantly greater number of S. aureus,
S. epidermidis, and P. aeruginosa present on the nanotextured and
nanotubular Ti substrates compared to conventional and nano-
rough Ti.
Fig. 5(a) displays the number of live bacteria colonies present on
the surfaces after 1 h while Fig. 5(b) displays the number of dead
bacteria colonies present on the surfaces after 1 h. Results indicated
that nanorough Ti substrates had the least amount of living bacteria
after 1 h. In other words, when normalized to the projected surface
area, there was a signicantly lower attachment of live bacteria
colonies for all strains (S. aureus, S. epidermidis, and P. aeruginosa)
on the nanorough Ti substrates compared to conventional, nano-
tubular, and nanotextured Ti substrates (Fig. 5(a)). In addition,
results showed that the nanotubular and nanotextured Ti
substrates had more live colonies for each bacteria line compared to
conventional Ti substrates (Fig. 5(a)). Furthermore, upon examining
the amount of dead bacteria present on the surfaces, nanotubular
and nanotextured Ti substrates contained the greatest number of
dead bacteria colonies for all bacteria lines, while the nanorough Ti
Table 1
Binding energy of the high resolution Ti 2p peaks for Ti before and after electron
beam evaporation (nanorough Ti) and anodization (nanotubular and nanotextured
Ti) as examined by XPS.
Substrates Peak Binding
Energy (ev)
Conventional Ti Ti 2p3/2 458.8
Ti 2p1/2 464.5
Nanorough Ti Ti 2p3/2 458.8
Ti 2p1/2 464.5
Nanotubular Ti Ti 2p3/2 458.8
Ti 2p1/2 464.5
Nanotextured Ti Ti 2p3/2 458.8
Ti 2p1/2 464.5
Table 2
Atomic percentage of selective elements in the outermost layers of Ti before and
after electron beam evaporation (nanorough Ti) and anodization (nanotubular and
nanotextured Ti) as examined by XPS.
Substrates O Ti F
Conventional Ti 51.23 48.77 0
Nanorough Ti 49.34 50.66 0
Nanotubular Ti 57.00 33.62 9.38
Nanotextured Ti 56.27 34.95 6.03
Fig. 2. Total surface energy of the Ti before and after electron beam evaporation
(nanorough Ti) and anodization (nanotubular and nanotextured Ti). Surface energy
was calculated for each sample by measuring the contact angle of three liquids at the
sample surface and entering values into the OwensWendt equation. Values are
mean SEM; n 4; *p <0.01 compared to unmodied Ti; **p <0.01 compared to
nanorough Ti; ***p <0.05 compared to nanotextured Ti.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 709
substrates contained the least amount of dead bacteria colonies
(Fig. 5(b)).
Fig. 6 shows the percentage of live bacteria, as calculated from
the data provided in Fig. 5. The results indicated that the nanorough
Ti substrates contained the signicantly highest percentage of live
bacteria for all strains attached to the surface after 1 h (Fig. 6).
When examining this data as well as the total number of bacterial
colonies (Fig. 3), it can be concluded that the nanorough Ti
substrates appeared to be the best surface for inhibiting bacteria
compared to the conventional, nanotubular, and nanotextured Ti
substrates.
3.4. Protein adsorption
Fig. 7 shows that all nanofabricated surfaces (nanorough,
nanotextured, and nanotubular Ti) increased the adsorption of
bronectin compared to the conventional Ti surfaces (Fig. 7).
Among the nanofabricated surfaces, nanotextured and nanotubular
Ti signicantly increased bronectin adsorption compared to the
nanorough Ti (Fig. 7). Interestingly, bronectin adsorption corre-
lated to the surface energy data with the greatest surface energy
samples adsorbing the most bronectin (Fig. 2).
4. Discussion
Infection carries a signicant burden for orthopedic devices by
leading to early failure of the device as well as preventing proper
integration between the tissue and implant. Bacteria that are
surface bound versus planktonic are clinically more relevant to
biomaterial infection. Reducing bacteria adhesion has been previ-
ously explored on implant surfaces, but needs to be further
addressed. Nanotechnology has specically been investigated for
improving the success of orthopedic implants. Surfaces containing
features in the nanometer regime have been shown to increase cell
behavior, such as osteoblasts, while simultaneously reducing the
adhesion of bacteria. Yet, the role of nanotechnology as a potential
solution for decreasing bacteria adhesion warrants further expla-
nation. The purpose of this study was to explore a possible means
for the reduction of bacteria attachment to the implant surfaces.
Results from this study indicated that the presently prepared
nanorough Ti surfaces are the best surfaces for inhibiting bacterial
adhesion. Compared to conventional surfaces, nanostructured
materials have excellent biocompatibility properties due to
enhanced protein interaction (including adsorption and confor-
mation) resulting in improved cellular adhesion and tissue growth
[2931]. It has been demonstrated here that there is a linear rela-
tionship between nano-roughness, surface energy, and protein
adsorption. More specically, a surface that has more nanorough
features possesses increased surface energy which leads to greater
protein adsorption [2931]. This study also conrmed the same
correlation as it revealed that nanorough, nanotubular, and nano-
textured Ti possessed higher degrees of nanometer features, higher
surface energy, and increased bronectin adsorption compared to
conventional Ti. Furthermore, research has also shown that
increased protein adsorption, such as bronectin, results in
decreased bacteria attachment [33,34]. In the present study, this
trend was observed between the nanorough Ti, which promoted
the least amount of bacteria attachment, and conventional Ti.
Compared to conventional Ti, nanorough Ti possessed no chemical
difference, and, thus, the presence of nanometer features alone
(higher surface energy) increased bronectin adsorption which
decreased bacterial attachment.
Based on this thinking, decreased bacterial attachment would
also be expected for the nanotubular and nanotextured Ti
substrates since these surfaces had greater nanometer surface
roughness, surface energy, and bronectin adsorption. However,
increased bacteria attachment was observed on both the nano-
tubular and nanotextured Ti compared to the nanorough and
conventional Ti. It is possible that uorine present on the nano-
tubular and nanotextured on Ti surfaces (Table 2) increased
bacterial adhesion compared to conventional and nanorough Ti
surfaces. Further examining bacteria attachment on the anodized
nanotubular surfaces revealed the highest bacteria attachment
compared to the anodized nanotextured surfaces (Fig. 3), corre-
lating well to the possible role of greater bacteria attachment with
uorine concentration (Table 2). Other studies have conrmed this
trend that uorine present on a material surface increases bacterial
adhesion [3538]. Specically, Katsikogianni and colleagues
examined bacteria (S. epidermidis) function on polymers with and
without uorine [38]. They showed that the polymers containing
uorine increased bacteria attachment [38]. Li and colleagues
showed that increasing the concentration of surface ions encour-
aged the binding for both gram positive (Bacillus subtilis) and gram
negative (two P. aeruginosa strains, three Escherichia coli strains,
and two Burkholderia cepacia strains) bacteria to glass or metal
oxide surfaces [37]. This observationwas not affected by the surface
charge or hydrophobicity/hydrophilicity of the bacteria surface
[37]. These ndings may explain the results of this current study
which demonstrated that nanotubular and nanotextured Ti
surfaces containing uorine ions increased bacteria attachment
despite observed increases in bronectin adsorption. It is also
interesting to note that a previous study by Popat and colleagues
were able to decrease the adhesion of bacteria (S. epidermidis) on
Fig. 3. Decreased S. aureus, S. epidermidis, and P. aeruginosa colonies on nanorough and
conventional Ti compared to nanotubular and nanotextured Ti after 1 h. Data are
mean SEM; n 3; *p <0.01 compared to nanorough Ti; **p <0.01 compared to
conventional Ti; ***p <0.01 compared to nanotextured Ti;
#
p <0.1 compared to
nanotextured Ti;
##
p <0.05 compared to nanotextured Ti for respective bacteria lines.
Table 3
Contact angle measurements (deg) of three liquids on Ti before and after electron
beam evaporation (nanorough Ti) and anodization (nanotubular and nanotextured
Ti). Contact angle data was used to determine surface energy via the OwensWendt
equation. Values are mean SEM, n 4.
Substrates Contact angle of
DI water
Contact angle
of glycerol
Contact angle
of PEG
Conventional Ti 70.6 1.58 69.3 0.84 41.18 1.20
Nanorough Ti 59.3 1.13 57.6 0.89 28.3 1.74
Nanotubular Ti 26.5 2.40 21.9 1.32 11.1 0.80
Nanotextured Ti 29.5 1.13 25.4 1.35 17.2 1.43
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 710
nanotubular Ti (prepared by an anodization similar to the process
used in this study) compared to conventional counterparts, only
after loading antibiotics into the Ti nanotubes [39].
Although total bacteria adhered the most to the anodized
nanotubular surfaces, this study also revealed that the anodized
surfaces (nanotubular and nanotextured Ti) decreased the
percentage of living cells compared to the non-uorinated surfaces
(nanorough and conventional Ti). This could be a result of the
antibacterial effects caused by the presence of uorine, as shown by
other studies [4042]. For example, Raulio and colleagues found
that by coating stainless steel with uoropolymers, it was possible
to reduce biolm formation of several bacteria strains, including
S. epidermidis [41]. In addition, Yoshinari and colleagues found
uorine ion-implanted Ti surfaces contained fewer viable bacteria
colony forming units further suggesting antibacterial properties of
uorine. This study demonstrated that uorine ions were not
released suggesting that the formation of metal uoride bonds
were sufcient for producing antibacterial effects.
Another reason for the observation of increased total bacteria
colonies on nanotubular and nanotextured Ti surfaces compared to
nanorough and conventional Ti surfaces could be explained by the
large number of adherent dead bacteria (Fig. 5 (b)). Dead bacteria
bound to a biomaterial surface can aid in the adhesion of subse-
quent live bacteria [43,44]. Specically, dead or dying P. aeruginosa
can release intracellular lectins to promote the adhesion of living
bacteria [43]. In a similar manner, S. aureus can release an inter-
cellular protein upon death to enhance the adhesion to other
microorganisms [44]. Clearly, the release of such compounds from
adherent dead bacteria may have promoted subsequent live
bacterial adhesion in this study (Fig. 5(a)).
Furthermore, there was also a difference in the crystallinity
between the anodized Ti surfaces, nanorough, and conventional Ti
that can be linked to bacteria adherence. Nanotextured and nano-
tubular Ti contained amorphous TiO
2
while the nanorough and
conventional surfaces contained crystalline TiO
2
(anatase and rutile
phase). Research has shown that amorphous TiO
2
promoted
bacteria attachment compared to anatase TiO
2
(which is known to
possess antibacterial properties [45,46]). Thus, despite the fact that
nanotubular and nanotextured Ti surfaces increased nanometer
surface roughness, surface energy, and bronectin adsorption over
conventional Ti, the presence of amorphous TiO
2
may have also
increased bacterial attachment.
Fig. 4. Fluorescent micrographs of decreased S. aureus colonies on (b) nanorough Ti compared to all other substrates and increased bacteria colonies on the (c) nanotextured and
(d) nanotubular Ti compared to (a) conventional Ti after 1 h. These micrographs were representative of S. epidermidis and P. aeruginosa.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 711
5. Conclusions
A simple means for the reduction of bacteria on and subsequent
infection of titanium using nanometer sized Ti surface features was
explored here for orthopedic applications. In summary, results of
this in vitro study demonstrated the decreased adhesion of
S. aureus, S. epidermidis, and P. aeruginosa (bacteria that limit
orthopedic implant function and efcacy) on nanorough Ti surfaces
created through electron beam evaporation while nanotubular and
nanorough Ti created through anodization resulted in an increase
of bacteria attachment. This research demonstrated that through
careful selection of nanometer surface properties to increase
bronectin adsorption, while maintaining favorable chemistry and
crystallinity (specically, anatase TiO
2
) it was possible to decrease
bacteria adhesion. This study, thus, provided further knowledge to
the orthopedic eld on ways to reduce bacteria colonization,
a prerequisite for infection, which should be investigated as
a means to improve the longevity of orthopedic implants.
Acknowledgements
The authors wish to acknowledge the VA Pre-Doctoral Associ-
ated Health Rehabilitation Research Fellowship Program and the
Department of Veterans Affairs, RR & D, A3772C for funding along
with the National Science Foundation Graduate K-12 Teaching
Fellowship. The authors also wish to acknowledge the Microelec-
tronics Facility and the Leduc Bioimaging Facility at Brown
University as well as the University of Rhode Island Surface Char-
acterization Laboratory. Lastly, the authors would like to thank
Mr. Anthony McCormick for his help with SEM.
Appendix
The full color images can be found in the online version, at doi:
10.1016/j.biomaterials.2009.09.081.
References
[1] Donlan RM. Biolms and device-associated infections. Emerg Infect Dis
2001;7:27781.
[2] von Recum AF, Park JB. Permanent percutaneous devices. Crit Rev Bioeng
1981;5:3777.
[3] Lentino JR. Prosthetic joint infections: bane of orthopedists, challenge for
infectious disease specialists. Clin Infect Dis 2003;36:115761.
Fig. 5. Increased S. aureus, S. epidermidis, and P. aeruginosa live (a) and dead (b)
colonies on nanotubular and nanotextured Ti compared to nanorough and conven-
tional Ti after 1 h. Data are mean SEM; n 3; *p <0.05 compared to nanorough Ti;
**p <0.01 compared to nanorough Ti; ***p <0.01 compared to conventional Ti;
#
p <0.05 compared to conventional Ti;
##
p <0.01 compared to nanotextured Ti;
###
p <0.05 compared to nanotextured Ti for respective bacteria lines.
Fig. 6. The highest percentage of live bacteria colonies for S. aureus, S. epidermidis, and
P. aeruginosa attached to the nanorough Ti surfaces after 1 h compared to the
conventional, nanotextured, and nanotubular Ti surfaces. Data are mean SEM; n 3;
*p <0.1 compared to nanotextured Ti; **p <0.01 compared to nanotextured Ti;
***p <0.05 compared to nanotubular Ti;
#
p <0.05 compared to conventional Ti;
##
p <0.01 compared to nanotubular Ti;
###
p <0.1 compared to nanotubular Ti;

p <0.1 compared to conventional Ti;

p <0.1 compared to nanotubular Ti for
respective bacteria lines.
Fig. 7. Increased bronectin adsorption on nanorough, nanotubular, and nanotextured
Ti compared to conventional Ti. Data are mean SEM; n 3; *p <0.01 compared
to conventional Ti; **p <0.1 compared to conventional Ti; ***p <0.05 compared to
nanorough Ti;
#
p <0.01 compared to nanorough Ti;
##
p <0.1 compared to
nanotextured Ti.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 712
[4] Kurtz SM, Lau E, Schmier J, Ong KL, Zhao K, Parvizi J. Infection burden for hip
and knee arthroplasty in the United States. J Arthroplasty 2008;23:98491.
[5] Van Lenthe GH, de Waal Malejt MC, Huiskes R. Stress shielding after total
knee replacement may cause bone resorption in the distal femur. J Bone Joint
Surg Br 1997;79:11722.
[6] Burke DW, OConnor DO, Zalenski EB, Jasty M, Harris WH. Micromotion of
cemented and uncemented femoral components. J Bone Joint Surg Br
1991;73:337.
[7] Lerouge S, Huk O, Yahia LH, Sedel L. Characterization of in vivo wear debris
from ceramicceramic total hip arthroplasties. J Biomed Mater Res
1996;32:62733.
[8] Donlan RM, Costerton JW. Biolms: survival mechanisms of clinically relevant
microorganisms. Clin Microbiol Rev 2002;15:16793.
[9] Gristina AG. Biomaterial-centered infection: microbial adhesion versus tissue
integration. Science 1987;237:158895.
[10] Moroni A, Vannini F, Mosca M, Giannini S. State of the art review: techniques
to avoid pin loosening and infection in external xation. J Orthop Trauma
2002;16:18995.
[11] Mahan J, Seligson D, Henry SL, Hynes P, Dobbins J. Factors in pin tract infec-
tions. Orthopedics 1991;14:3058.
[12] Zlowodzki M, Prakash JS, Aggarwal NK. External xation of complex femoral
shaft fractures. Int Orthop 2007;31:40913.
[13] Green SA, Ripley MJ. Chronic osteomyelitis in pin tracks. J Bone Joint Surg Am
1984;66:10928.
[14] Costerton JW, Stewart PS, Greenberg EP. Bacterial biolms: a common cause of
persistent infections. Science 1999;284:131822.
[15] Rimondini L, Fini M, Giardino R. The microbial infection of biomaterials:
a challenge for clinicians and researchers. A short review. J Appl Biomater
Biomech 2005;3:110.
[16] Fux CA, Costerton JW, Stewart PS, Stoodley P. Survival strategies of infectious
biolms. Trends Microbiol 2005;13:3440.
[17] Baquero F. Gram-positive resistance: challenge for the development of new
antibiotics. J Antimicrob Chemother 1997;39(Suppl. A):16.
[18] Campoccia D, Montanaro L, Arciola CR. The signicance of infection related to
orthopedic devices and issues of antibiotic resistance. Biomaterials 2006;
27:23319.
[19] Webster TJ, Ejiofor JU. Increased osteoblast adhesion on nanophase metals: Ti,
Ti6Al4V, and CoCrMo. Biomaterials 2004;25:47319.
[20] Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion on nanophase ceramics.
Biomaterials 1999;20:12217.
[21] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Enhanced functions of
osteoblasts on nanophase ceramics. Biomaterials 2000;21:180310.
[22] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Enhanced osteoclast-
like cell functions on nanophase ceramics. Biomaterials 2001;22:132733.
[23] Ejiofor JU, Webster TJ. Bone cell adhesion on titanium implants with nanoscale
surface features. Int J Powder Metall 2004;40:4353.
[24] Ward BC, Webster TJ. Increased functions of osteoblasts on nanophase metals.
Mater Sci Eng C 2007;27:5758.
[25] Miller DC, Vance RJ, Thapa A, Webster TJ, Haberstroh KM. Comparison of
broblast and vascular cell adhesion to nano-structured poly(lactic-co-gly-
colic acid) lms. Appl Bionics Biomech 2005;2:17.
[26] Cohen A, Liu-Synder P, Storey D, Webster T. Decreased broblast and increased
osteoblast functions on ionic plasma deposited nanostructured Ti coatings.
Nanoscale Res Lett 2007;2:38590.
[27] Puckett S, Pareta R, Webster TJ. Nano rough micron patterned titanium for
directing osteoblast morphology and adhesion. Int J Nanomed 2008;3:22941.
[28] Yao C, Slamovich EB, Webster TJ. Enhanced osteoblast functions on anodized
titanium with nanotube-like structures. J Biomed Mater Res A 2008;85:
15766.
[29] Woo KM, Chen VJ, Ma PX. Nano-brous scaffolding architecture selectively
enhances protein adsorption contributing to cell attachment. J Biomed Mater
Res A 2003;67:5317.
[30] Khang D, Kim SY, Liu-Snyder P, Palmore GT, Durbin SM, Webster TJ. Enhanced
bronectin adsorption on carbon nanotube/poly(carbonate) urethane: inde-
pendent role of surface nano-roughness and associated surface energy.
Biomaterials 2007;28:475668.
[31] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R. Specic proteins
mediate enhanced osteoblast adhesion on nanophase ceramics. J Biomed
Mater Res 2000;51:47583.
[32] Colon G, Ward BC, Webster TJ. Increased osteoblast and decreased Staphylo-
coccus epidermidis functions on nanophase ZnO and TiO
2
. J Biomed Mater Res
A 2006;78:595604.
[33] Anagnostou F, Debet A, Pavon-Djavid G, Goudaby Z, Helary G, Migonney V.
Osteoblast functions on functionalized PMMA-based polymers exhibiting
Staphylococcus aureus adhesion inhibition. Biomaterials 2006;27:39129.
[34] Kinnari TJ, Peltonen LI, Kuusela P, Kivilahti J, Kononen M, Jero J. Bacterial
adherence to titanium surface coated with human serum albumin. Otol
Neurotol 2005;26:3804.
[35] Eick S, Glockmann E, Brandl B, Pster W. Adherence of Streptococcus mutans to
various restorative materials in a continuous ow system. J Oral Rehabil
2004;31:27885.
[36] Correa CB, Pires JR, Fernandes-Filho RB, Sartori R, Vaz LG. Fatigue and uoride
corrosion on Streptococcus mutans adherence to titanium-based implant/
component surfaces. J Prosthodont 2009;18:3827.
[37] Li B, Logan BE. Bacterial adhesion to glass and metal-oxide surfaces. Colloids
Surf B Biointerfaces 2004;36:8190.
[38] Katsikogianni M, Spiliopoulou I, Dowling DP, Missirlis YF. Adhesion of slime
producing Staphylococcus epidermidis strains to PVC and diamond-like carbon/
silver/uorinated coatings. J Mater Sci Mater Med 2006;17:67989.
[39] Popat KC, Eltgroth M, Latempa TJ, Grimes CA, Desai TA. Decreased Staphylo-
coccus epidermis adhesion and increased osteoblast functionality on antibi-
otic-loaded titania nanotubes. Biomaterials 2007;28:48808.
[40] Yoshinari M, Oda Y, Kato T, Okuda K. Inuence of surface modications to
titanium on antibacterial activity in vitro. Biomaterials 2001;22:20438.
[41] Raulio M, Jarn M, Ahola J, Peltonen J, Rosenholm JB, Tervakangas S, et al.
Microbe repelling coated stainless steel analysed by eld emission scanning
electron microscopy and physicochemical methods. J Ind Microbiol Biotechnol
2008;35:75160.
[42] Hyde FW, Alberg M, Smith K. Comparison of uorinated polymers against
stainless steel, glass and polypropylene in microbial biolm adherence and
removal. J Ind Microbiol Biotechnol 1997;19:1429.
[43] Wentworth JS, Austin FE, Garber N, Gilboa-Garber N, Paterson CA, Doyle RJ.
Cytoplasmic lectins contribute to the adhesion of Pseudomonas aeruginosa.
Biofouling: J Bioadhesion Biolm Res 1991;4:99104.
[44] Palma M, Haggar A, Flock JI. Adherence of Staphylococcus aureus is enhanced
by an endogenous secreted protein with broad binding activity. J Bacteriol
1999;181:28405.
[45] Del Curto B, Brunella MF, Giordano C, Pedeferri MP, Valtulina V, Visai L, et al.
Decreased bacterial adhesion to surface-treated titanium. Int J Artif Organs
2005;28:71830.
[46] Yang B, Uchida M, Kim HM, Zhang X, Kokubo T. Preparation of bioactive tita-
nium metal via anodic oxidation treatment. Biomaterials 2004;25:100310.
S.D. Puckett et al. / Biomaterials 31 (2010) 706713 713