Content uploaded by Edith M Sampson
Author content
All content in this area was uploaded by Edith M Sampson on Jan 08, 2014
Content may be subject to copyright.
Impact of engineered surface microtopography on biofilm formation
of Staphylococcus aureus
Kenneth K. Chung
Department of Materials Science and Engineering, University of Florida, Gainesville, Florida 32611
James F. Schumacher
J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, Florida
32611
Edith M. Sampson
Department of Otolaryngology, University of Florida, Gainesville, Florida 32611
Robert A. Burne
Department of Oral Biology, UF College of Dentistry, University of Florida, Gainesville, Florida 32611
Patrick J. Antonelli
Department of Otolaryngology, University of Florida, Gainesville, Florida 32611
Anthony B. Brennana兲
Department of Materials Science and Engineering, and J. Crayton Pruitt Family Department of Biomedical
Engineering, University of Florida, Gainesville, Florida 32611
共Received 29 March 2007; accepted 30 May 2007; published 29 June 2007兲
The surface of an indwelling medical device can be colonized by human pathogens that can form
biofilms and cause infections. In most cases, these biofilms are resistant to antimicrobial therapy and
eventually necessitate removal or replacement of the device. An engineered surface
microtopography based on the skin of sharks, Sharklet AF™, has been designed on a poly共dimethyl
siloxane兲elastomer 共PDMSe兲to disrupt the formation of bacterial biofilms without the use of
bactericidal agents. The Sharklet AF™ PDMSe was tested against smooth PDMSe for biofilm
formation of Staphylococcus aureus over the course of 21 days. The smooth surface exhibited
early-stage biofilm colonies at 7 days and mature biofilms at 14 days, while the topographical
surface did not show evidence of early biofilm colonization until day 21. At 14 days, the mean value
of percent area coverage of S. aureus on the smooth surface was 54% compared to 7% for the
Sharklet AF™ surface 共p⬍0.01兲. These results suggest that surface modification of indwelling
medical devices and exposed sterile surfaces with the Sharklet AF™ engineered topography may be
an effective solution in disrupting biofilm formation of S. aureus.© 2007 American Vacuum
Society. 关DOI: 10.1116/1.2751405兴
I. INTRODUCTION
Bacterial biofilms are a major concern in the development
of biomaterials, ultrafiltration systems, and underwater ves-
sels. In the biomedical arena, bacterial colonization of sur-
faces compromises the effectiveness of implanted materials
and ultimately can result in persistent infections.1,2 Biomate-
rial surfaces for tissue constructs and implants are subject to
a competition between microbial adhesion and tissue integra-
tion, where an ideal surface would prevent the former while
promoting the latter.3However, the chemical, mechanical,
and physical properties of tissue construct materials are in-
herently meant to enhance all forms of biogrowth, making it
equally likely that such a surface would submit to bacterial
colonization.
Microorganisms colonize biomedical implants by devel-
oping biofilms, structured communities of microbial cells
embedded in an extracellular polymeric matrix that are ad-
herent to the implant and/or the host tissues.2,4 Biofilms ren-
der microorganisms resistant to host defenses and antibiotic
therapy.5,6 The clinical means of managing biofilms has been
prevention.
Preventing biofilm-associated infections has traditionally
been through use of prophylactic antibacterial agents,
whether delivered systemically or released directly from the
biomaterial.7Pharmacokinetics and toxicity of the antibacte-
rial agents incorporated in biomaterials have limited the ef-
fectiveness of such therapies.7–10 Antibiotics, antibodies, and
phagocytes can clear planktonic cells released by the biofilm,
but the sessile communities themselves are resistant to such
agents.11,12 Antibiotic therapy resulting in incomplete eradi-
cation of biofilm has been linked with the emergence of
antibiotic-resistant bacteria, which may compromise the ef-
fectiveness of these agents for even non-biofilm-mediated
infections.13–16
Another strategy for preventing the development of bio-
films has been to alter the biomaterial surface properties.
Surface modification techniques to tailor the surface energy
via surface chemistry and surface topography have been de-
veloped to study the effects of changes in these surface prop-
a兲Electronic mail: abrennan@mse.ufl.edu
89 89Biointerphases 2„2…, June 2007 1934-8630/2007/2„2…/89/6/$23.00 ©2007 American Vacuum Society
erties on biofilm formation.17,18 Bacterial adhesion has been
investigated on surface topographies that range from random
structures to ordered arrays. There appears to be a trend to-
ward increased bacterial coverage as the Ra-roughness values
increased on electropolished steel.19 Conversely, P. aerugi-
nosa was less likely to foul hydrophilic, electrically neutral,
smooth polymeric surfaces.20 Interestingly, bacterial adhe-
sion was reduced on stainless steel surface microtopogra-
phies that were generated by a one-directional polishing fin-
ish relative to smooth surfaces.21 The aforementioned studies
examined surfaces that were randomly roughened and did
not examine specific surface features. The effects of a non-
random topography consisting of etched grooves of varying
widths in silicon coupons with P. aeruginosa and P. fluore-
scens showed that rates of attachment were independent of
groove size and greatest on the downstream edges of
grooves.22 More recently, microbial retention on a defined
microtopography in the form of etched pits was determined
to be dependent on both the size of the surface defect and the
cell.23
The purpose of our study was to investigate the potential
for bacterial attachment and colonization on an engineered
topography with a well-defined structure. Our research is fo-
cused on the design and characterization of surface microto-
pographies that effectively control bioadhesion, with the goal
to produce a biomaterial with topography variants that can
effectively switch from biofilm forming to biofilm inhibition.
To achieve this, a surface energy model was created corre-
lating wettability with bioadhesion to characterize and de-
velop settlement-enhancing as well as antifouling
topographies.24 The application of this model led to the de-
sign concept of a biomimetic structure inspired by the skin of
fast-moving sharks 共Sharklet AF™兲, and an engineered
roughness index 共ERI兲was developed as an extension of this
model to study and develop other unique engineered
topographies.25
In this article, a surface of uniform chemistry with an
engineered microtopography was investigated for the inhibi-
tion of bacterial biofilm formation. The most successful de-
sign to date is Sharklet AF™. This particular microtopogra-
phy is unique from the aforementioned ones in that it has
nonrandom, clearly defined surface features that are typically
tailored to the critical dimensions of the fouling organism.
Recent results on the Sharklet AF™ and other engineered
microtopographies designed at a 2
m feature width and
spacing have shown a strong correlation between the ERI
and the inhibition of settlement by the zoospores 共⬃5
min
diameter兲of the most common ship fouling alga, Ulva.25 In
addition, the Sharklet AF™ microtopography designed at a
20
m feature width and spacing has been demonstrated to
be a strong inhibitor of the settlement of barnacle cyprids of
B. amphitrite.26 This organism’s attachment disk measures
⬃25–30
m.27 For the present study, the Ulva-specific Shar-
klet AF™ surface was selected for the potential to inhibit
biofilm formation of Staphylococcus aureus based on the ap-
proximate match between the size of the bacteria and the
critical dimensions of this surface 共2
m feature width and
spacing, 3
m feature height兲. It was hypothesized that the
dimensions of the topography would be slightly too large to
effectively reduce the attachment of the bacteria in the size
range of ⬃1–2
m but could be effective at physically dis-
rupting the further colonization of additional bacteria and
subsequent formation of biofilm 共Fig. 1兲.S. aureus was se-
lected as the bacterial pathogen due to both its size and its
association with nosocomial infections in implanted devices,
such as cochlear implants, sutures, and heart valves.10,28
II. MATERIALS AND METHODS
A. Materials
Dow Corning® Silastic® T-2, a platinum-catalyzed poly-
共dimethyl siloxane兲elastomer 共PDMSe兲, was used for its low
modulus, low surface energy, and propensity for minimal
bioadhesion.29 Silastic® brand silicone elastomers are bio-
materials used in numerous medical devices including tub-
ing, catheters, and pacemaker leads. The elastomer was pre-
pared by mixing one part by weight of curing agent with ten
parts by weight of resin, then degassing under vacuum
共28–30 in. Hg兲for 30 min. The mixture was cured at
⬃22 °C for 24 h.
FIG. 2. Locations 关共A兲–共C兲兴 for scanning electron microscopy 共SEM兲analy-
sis for 8 mm circular PDMSe samples.
FIG. 1. Sharklet AF™ topography on poly共dimethyl si-
loxane兲elastomer 共PDMSe兲with 2
m feature width
and spacing and 3
m feature height. 共A兲Light micro-
graph with top-down view. 共B兲Scanning electron mi-
crograph with top down view. 共C兲Scanning electron
micrograph taken at 35° tilt to show protruding
features.
90 Chung et al.: Impact of engineered surface microtopography 90
Biointerphases, Vol. 2, No. 2, June 2007
B. Sharklet AF™ design and fabrication of
topographical molds
The Sharklet AF™ design24,25 consists of 2
m wide rect-
angular ribs of varying lengths ranging from 4 to 16
m.
The ribs of varying lengths are combined into a periodic,
diamondlike array at a fixed spacing of 2
m between neigh-
boring features 共Fig. 1兲. The final, resultant Sharklet AF™
topography in PDMSe was created by replication of silicon
wafer molds. Silicon wafer molds were fabricated by first
transferring the Sharklet AF™ design to photoresist-coated
silicon wafers using photolithographic techniques as previ-
ously described.30 Next, the patterned silicon wafers were
deep reactive ion etched to a depth of 3
m before being
cleaned of residual photoresist with an O2plasma etch. The
etched silicon wafer surfaces were then methylated 共via va-
por deposition兲with hexamethyldisilazane to prevent adhe-
sion. These wafers served as negative molds for topographi-
cal replication of the Sharklet AF™ topography at a feature
height of 3
m onto a PDMSe surface.
C. Sample preparation
Silicon molds were replicated into PDMSe to produce
⬃0.4 mm thick films containing protruding topographical
features. Briefly, the PDMSe material components were pre-
pared and mixed as described,24,25 poured over the silicon
wafer molds, and the entire system was pressed between two
larger glass plates with the appropriate spacers to produce a
0.4 mm thick film. After curing for 24 h, the PDMSe film
was removed from the silicon wafer with protruding topo-
graphical features forming the Sharklet AF™ topography on
the PDMSe surface 共Fig. 1兲. PDMSe samples of smooth
共replicated from an unmodified silicon wafer兲and topo-
graphically modified were punched out with a circular die
8 mm in diameter. Five replicates each of smooth PDMSe
and Sharklet AF™ PDMSe disks were placed into 3 in. Petri
dishes 共one dish for each day examined兲for the growth assay
and gas sterilization.
D. S. aureus biofilm formation assay
Staphylococcus aureus 共ATCC 29213兲was subcultured in
tryptic soy broth 共TSB兲growth medium and grown at 37 °C
overnight in static conditions. Optical absorbance was mea-
sured, serial dilutions were performed, and growth curve and
linear optical density-colony-forming unit 共CFU兲regression
were plotted. Bacterial concentration was determined via
spectrophotometry by interpolating CFU per milliliter from
the linear optical density-CFU regression. Samples were
statically immersed in a 107CFU/ml bacterial suspension
for up to 21 days. Every day, dishes were put on a rocker for
1 min at 30 rpm and the medium was replaced to allow for
continued bacterial growth. Dishes were removed on days 0,
2, 7, 14, and 21. For each removed dish, areas surrounding
FIG.3. 关共A兲and 共B兲兴 SEM images 共2000⫻兲of S. au-
reus on smooth and topographically modified PDMSe
surfaces after 7 day exposure. 关共C兲and 共D兲兴 Processed
SEM images of smooth and Sharklet AF™ PDMSe sur-
faces using MACROMEDIA FIREWORKS®. Bacteria covered
areas were outlined and blackened.
FIG. 4. Grouping and numbering of bacteria colonies as
measured by IMAGEJ software. The analysis pictured
was conducted on the processed SEM images in Fig. 3.
共A兲Bacteria coverage on a smooth PDMSe surface was
detected as two colonies. 共B兲Bacteria coverage on a
Sharklet AF™ PMDSe surface was detected as over 40
individual colonies of bacteria.
91 Chung et al.: Impact of engineered surface microtopography 91
Biointerphases, Vol. 2, No. 2, June 2007
the samples on the placement grid were rinsed with de-
ionized water using a Pipet-Aid® and aspirated to eliminate
nonadherent cells. This rinsing procedure was repeated for a
total of three times, and the dish was then put on an orbital
rocker for 1 min at 30 rpm. Each sample in the dish was then
treated with 20 ml of 10 mM cetyl-pyridinium chloride fixa-
tive and allowed to air dry overnight. Another dish exposed
only to sterile medium was incubated with the samples for
21 days and served as a negative control.
E. Characterization
Samples were dehydrated in a graded ethanol series of
25%, 50%, 75%, 95%, and 100% at 10 min intervals.
Samples were washed twice with hexamethyldisilazane with
a 5 min interval between washings, followed by drying using
a vacuum desiccator. Each sample was attached to a beveled
disk and sputter coated with Au/Pd and imaged with a JEOL
6400 scanning electron microscope 共SEM兲. SEM images at
2000⫻from areas A, B, and C for each replicate were used
to quantify bacterial growth on each surface 共Fig. 2兲.
Biofilms were identified by the presence of microorgan-
isms and exopolymeric matrix. Biofilm growth was esti-
mated by measuring the percent area of coverage of bacteria.
To obtain this value, SEM images were first processed using
MACROMEDIA FIREWORKS® software to outline and blacken
the area containing bacteria and/or biofilm 共e.g., Fig. 3兲.
Processed SEM images were analyzed for percent cover-
age using IMAGEJ software.31 Areas of coverage were num-
bered and outlined, and the total summation of area covered
was measured 共e.g., Fig. 4兲. For each replicate, the results for
areas A, B, and C are combined to reflect the percentage
coverage of bacteria for that specific replicate. Results for
each replicate are reported as percent coverage of bacterial
colonies.
F. Statistical analysis
The mean value 共⫾standard error兲of percent area cover-
age of bacteria on both smooth and Sharklet AF™ PDMSe
surfaces at days 0, 2, 7, 14, and 21 was calculated. Statistical
differences were evaluated by a two-way analysis of variance
for the factors of “surface” 共smooth versus Sharklet AF™兲
and “day” 共0, 2, 7, 14, and 21兲followed by Tukey’s test for
multiple comparisons. Statistical differences were considered
at the 95% confidence level.
III. RESULTS
On day 0, individual cells were seen on the surfaces of
smooth PDMSe 关Fig. 5共A兲兴, while individual cells appeared
in the recesses between the protruding features for Sharklet
AF™ PDMSe surfaces 关Fig. 5共B兲兴. On day 2, microcolonies
of bacteria began to form on the smooth surfaces 关Fig. 5共C兲兴,
and the Sharklet AF™ surfaces continued to have isolated
cells accrete between features 关Fig. 5共D兲兴. Growth of the
microcolonies increased on day 7 for the smooth surfaces
into early-stage biofilms 关Fig. 5共E兲兴. The Sharklet AF™ sur-
faces 共day 7兲continued to exhibit small-sized clusters of bac-
teria, with no evidence of early-stage biofilm development;
the clusters were positioned similar to day 2 in the recesses
between protruding topographical features 关Fig. 5共F兲兴.On
day 14, the smooth surfaces had the first evidence of mature
biofilms 关Fig. 5共G兲兴, while the Sharklet AF™ surfaces had a
slight increase in the number of small clusters of cells com-
pared to day 7 but still no evidence of early biofilm devel-
opment or formation 关Fig. 5共H兲兴. On day 21, a significant
portion of the smooth PDMSe surfaces was colonized by
biofilms 关Fig. 5共I兲兴, and biofilms first appeared in isolated
areas on the Sharklet AF™ PDMSe surfaces 关Fig. 5共J兲兴.Ar-
eas surrounding the large bacterial colonies on the topo-
graphical surfaces for day 21 were virtually devoid of adher-
ent bacteria. SEM images of the negative control samples
exposed to only TSB media showed no cells.
FIG. 5. Representative SEM images of S. aureus on PDMSe surfaces over
the course of 21 days 共areas of bacteria highlighted with color to enhance
contrast兲. On the left are smooth PDMSe surfaces and the right column
shows Sharklet AF™ PDMSe surfaces. 共A兲and 共B兲day 0, 共C兲and 共D兲day
2, 共E兲and 共F兲day 7, 共G兲and 共H兲day 14, and 共I兲and 共J兲day 21.
92 Chung et al.: Impact of engineered surface microtopography 92
Biointerphases, Vol. 2, No. 2, June 2007
Image and statistical analysis indicated that smooth
PDMSe samples had significant increases in bacterial cover-
age for pooled time points 共Tukey’s test, p⬍0.05兲, with the
first evidence of biofilm on day 7 samples. The Sharklet
AF™ samples had significantly lower values of percent area
coverage for days 7, 14, and 21 共Tukey’s test, p⬍0.05兲, with
biofilm colonies not appearing until day 21. Even on day 21,
biofilm colonies covered only isolated areas on Sharklet
AF™ samples, with little to no evidence of biofilms or bac-
terial cells in other areas. The mean value 共⫾standard error兲
of percent area coverage of bacteria on both smooth and
Sharklet AF™ PDMSe surfaces at days 0, 2, 7, 14, and 21
was calculated and is graphically displayed 共Fig. 6兲.
IV. DISCUSSION
Most in vitro studies involving S. aureus have examined
the adhesion behavior over the course of a few hours.16,19
However, for transcutaneous devices such as catheters, the
time frame for biofilm formation is typically 14 days.7Thus,
the focus of this study was to test the effects of an engineered
microtopography on bacterial colonization and biofilm for-
mation for a period of time that extended beyond 14 days.
The growth assay parameters included optimized conditions
for S. aureus colonization and spanned 21 days. To date, this
is the first in vitro study involving surface topography and S.
aureus over a time period approximating that of short-term
indwelling devices.
Material selection in this study was predicated upon the
popularity of silicone as the choice material for molded im-
plants, such as cochlear implants and long-term catheters,
despite being shown to have nearly a tenfold greater risk of
infection than other polymer materials.32 The static culture
provided for biofilm growth was chosen to represent the
most challenging environment for the material surface in the
presence of bacteria, as opposed to the low shear dynamics
of indwelling catheters. The results of this study strongly
suggest that the surface modification of existing silicone de-
vices with the Sharklet AF™ topography may prolong the
service life and improve the efficacy of these devices. It is
also encouraging to note that the topographical modification
of a surface used in this study involves no chemical changes
of the biomaterial surface and does not rely on the release of
any antibacterial agents.
Both qualitative 共e.g., development of extracellular ma-
trix兲and quantitative measures of biofilm formation revealed
inhibition of biofilm development on PDMSe with Sharklet
AF™ microtopography. The results confirm the hypothesis
that cells can fit in the recessed regions between the pro-
truded topographical features, but evidence of biofilm forma-
tion did not occur on the Sharklet AF™ features until day 21.
The raised features could potentially reduce the surface area
exposed to bacteria if the channels failed to fill with growth
media. However, finding bacteria strictly in the Sharklet
AF™ channels speaks against this possibility. Observations
of adhered bacteria would suggest that the protruded features
of the topographical surface provided a physical obstacle to
deter the expansion of small clusters of bacteria present in
the recesses into microcolonies. It was at day 21 when bac-
teria were observed to form small, multilayered colonies
within the recesses in order to extend over the protruding
features and connect to other isolated colonies. This phenom-
enon may be the explanation for the delay of early-stage
biofilm development to day 21 that was evident on the
smooth surface at day 7.
In the context of the “race for the surface” in
biomaterials,3our engineered surface approach suggests the
use of a hierarchy of surface topographies26 to control bio-
adhesion. Previous results detailing an engineered topogra-
phy capable of promoting cell growth33 can be integrated
into the Sharklet AF™ in this study to produce a hierarchical
topography for desirable competitive adhesion at the bioma-
terial surface. One can envision a surface that repels and
delays biofilm formation to the extent that host cells, vital to
the integration of the biomaterial with the physiological en-
vironment, can be established and proliferated on the de-
signed surface.
This study was performed using a highly simplified in
vitro model. A great deal of work is needed to determine if
FIG. 6. Mean value of percent area coverage of bacteria
on smooth and Sharklet AF™ PDMSe surfaces at vari-
ous time points. Bars represent± standard error, n=5.
93 Chung et al.: Impact of engineered surface microtopography 93
Biointerphases, Vol. 2, No. 2, June 2007
observations from this in vitro model are borne out in vivo.
Current research is evaluating the adhesion and biofilm for-
mation tendencies of other biofilm-forming bacteria on Shar-
klet AF™. Also, the application of the engineered roughness
index is being used to predict other engineered topographies
that may be effective at inhibiting biofilm formation. Consid-
erations for designing the optimal microtopography for an
implantable device will include the interactions with host
molecules and impact on fibrous capsule formation.
V. CONCLUSIONS
S. aureus was cultured on PDMSe surfaces for up to
21 days. Biofilms were established after 14 days on the
smooth PDMse surfaces whereas an engineered surface mi-
crotopography with nonrandom, clearly defined features elic-
ited a negative response. The Sharklet AF™ microtopogra-
phy disrupted S. aureus colonization and biofilm formation
without the use of bactericidal agents. Engineered surface
microtopographies present a promising means of blocking
biofilm development and reducing the rate of biomedical im-
plant infections.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the financial support
of the Office of Naval Research, Award No. N00014-02-1-
0325 to one of the authors 共A.B.B兲. Special thanks to Sean
Royston for his technical assistance in production and fabri-
cation of the engineered topography.
1J. W. Costerton, Z. Lewandowski, D. E. Caldwell, D. R. Korber, and H.
M. Lappin-Scott, Annu. Rev. Microbiol. 49,711共1995兲.
2J. Costerton, P. Stewart, and E. Greenberg, Science 284,1318共1999兲.
3A. Gristina, Science 237,1588共1987兲.
4J. W. Costerton, G. G. Geesey, and K.-J. Cheng, Sci. Am. 238共1兲,86
共1978兲.
5C. Fux, J. Costerton, P. Stewart, and P. Stoodley, Trends Microbiol. 13,
34 共2005兲.
6W. Costerton, R. Veeh, M. Shirtliff, M. Pasmore, C. Post, and G. Ehrlich,
J. Clin. Invest. 11 2,1466共2003兲.
7J. W. Costerton, G. Cook, M. Shirtliff, P. Stoodley, and M. Pasmore, in
Biomaterials Science, edited by Buddy D. Ratner, Allan S. Hoffman,
Frederick J. Schoen, and Jack E. Lemons 共Elsevier, San Diego, CA,
2004兲, pp. 345–353.
8C. Gordon, N. Hodges, and C. Marriott, J. Antimicrob. Chemother. 22,
667 共1988兲.
9W. Nichols, M. Evans, M. Slack, and H. Walmsley, J. Gen. Microbiol.
135,1291共1989兲.
10L. G. Harris and R. G. Richards, Injury 37,S3共2006兲.
11J. C. Nickel, I. Ruseska, J. B. Wright, and J. W. Costerton, Antimicrob.
Agents Chemother. 27, 619 共1985兲.
12R. M. Donlan and J. W. Costerton, Clin. Microbiol. Rev. 15,167共2002兲.
13R. Schwalbe, J. Stapleton, and P. Gilligan, N. Engl. J. Med. 316,927
共1987兲.
14F. Biavasco, E. Giovanetti, M. P. Montanari, R. Lupidi, and P. E. Varaldo,
J. Antimicrob. Chemother. 27,71共1991兲.
15K. Hiramatsu, H. Hanaki, T. Ino, K. Yabuta, T. Oguri, and F. C. Tenover,
J. Antimicrob. Chemother. 40, 135 共1997兲.
16P. Vaudaux, P. Francois, B. Berger-Bachi, and D. P. Lew, J. Antimicrob.
Chemother. 47, 163 共2001兲.
17Y. H. An and R. J. Friedman, J. Biomed. Mater. Res. 43,338共1998兲.
18W. Teughels, N. Van Assche, I. Sliepen, and M. Quirynen, Clin. Oral
Implants Res. 17,68共2006兲.
19E. Medilanski, K. Kaufmann, L. Y. Wick, O. Wanner, and H. Harms,
Biofouling 18, 193 共2002兲.
20M. Pasmore, P. Todd, S. Smith, D. Baker, J. Silverstein, D. Coons, and C.
N. Bowman, J. Membr. Sci. 194,15共2001兲.
21A. Allion, J.-P. Baron, and L. Boulange-Petermann, Biofouling 22,269
共2006兲.
22T. Scheuerman, A. Camper, and M. Hamilton, J. Colloid Interface Sci.
208,23共1998兲.
23K. Whitehead, J. Colligon, and J. Verran, Colloids Surf., B 41,129
共2005兲.
24M. Carman et al., Biofouling 22,11共2006兲.
25J. F. Schumacher et al., Biofouling 23,55共2007兲.
26J. F. Schumacher, N. Aldred, M. E. Callow, J. A. Finlay, J. A. Callow, A.
S. Clare, and A. B. Brennan, Biofouling 共to be published兲.
27A. S. Clare and J. A. Nott, J. Mar. Biol. Assoc. U.K. 74,967共1994兲.
28P. Antonelli, J. Lee, and R. Burne, Otol. Neurotol. 25, 953 共2004兲.
29L. Hoipkemeier-Wilson et al., Biofouling 20,53共2004兲.
30A. W. Feinberg et al., ACS Symp. Ser. 838, 196 共2003兲.
31W. S. Rasband, IMAGEJ, U.S. National Institutes of Health, Bethesda, MD,
http://rsb.info.nih.gov/ij/, 1997–2006.
32R. J. Sherertz, in Infections Associated with Indwelling Medical Devices,
edited by F. A. Waldogel and A. L. Bisno 共ASM, Washington, DC, 2000兲,
pp. 111–125.
33A. W. Feinberg, W. R. Wilkerson, C. A. Seegert, A. L. Gibson, L.
Hoipkemeier-Wilson, and A. B. Brennan, J. Biomed. Mater. Res. 共to be
published兲.
94 Chung et al.: Impact of engineered surface microtopography 94
Biointerphases, Vol. 2, No. 2, June 2007