|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, Vol. 94 No. 8 (October 15), 1999:
pp. 2716-2724
By
From the Department of Chemical Engineering, University of Delaware,
Newark, DE; and the Sol Sherry Thrombosis Research Center, Temple
University School of Medicine, Philadelphia, PA.
Interaction of biomaterials with blood components including
neutrophils is responsible for some of the clinical complications that
have occurred in cardiopulmonary bypass, hemodialysis, and ventricular
assist procedures. The possibility of inhibiting the initial adhesion
of neutrophils to biomaterials has been studied extensively, but the
problem remains unsolved. In this study, we investigated the effect of
HK adsorption on polyurethane, a widely used component of
extracorporeal and intracorporeal devices. HK and HKa were allowed to
adsorb on 4 different charged polyurethanes: noncharged (PU), cationic
(NR4), anionic (SO3), and zwitterionic (GPC)
polyurethanes. The effect of kininogen adsorption on neutrophil adhesion, the surface density of the adsorbed kininogen, and the exposure of HK domains 3 and 5 (D3 and D5H),
which are responsible for the binding of HK to the neutrophil integrin
THE INTERACTION OF leukocytes with
biomedical devices is an important event in the host response to
biomaterials, because the inflammatory mediators released from adherent
or activated leukocytes induces complications in the clinical use of
dialyzers, oxygenators, vascular grafts, and hemofiltration
units.1,2 After adhesion to a biomaterial, neutrophils
undergo a number of processes, such as the secretion of neutral
proteases and respiratory burst, which can result in the deterioration
of the implant material and peripheral tissue damage.
Segmented polyurethane elastomers were introduced as biomaterials by
Boretos and Pierce,3 because they possessed superior mechanical properties and exhibited good in vitro stability. Cooper et
al have shown that the thromboresistance of PU can be increased by
sulfonate incorporation.4,5 However, this modification in
the absence of plasma proteins enhances neutrophil adhesion and
neutrophil respiratory burst and could well induce complications in
certain biomaterial applications.6,7 Recently, we reported reduced bacterial and neutrophil adhesion on polyurethanes containing the phosphorylcholine moiety.8,9 Other researchers have
also demonstrated a decrease in plasma protein adsorption and in
platelet adhesion on poly-alkyl-methacrylate-based polymers with
pendant phosphorylcholine groups.10,11
Kininogens are plasma proteins that are natural substrates of plasma
and tissue kallikrein and yield, on hydrolysis, the nonapeptide bradykinin. Bradykinin release has been implicated in hypotensive reactions during hemodialysis and cardiopulmonary bypass for patients receiving angiotensin converting enzyme (kininase 2) inhibitors. Two
forms of purified human plasma kininogen (kinin precursors) were
described by Jacobsen and Kriz12 in 1967: high molecular weight kininogen (HK) and low molecular weight kininogen (LK). HK
consists of 6 domains, of which domains 3 and 5 (D3 and
D5H) are responsible for the binding of HK to
neutrophils.13 D5H contains 2 anionic surface
binding subdomains: 1 rich in histidine and glycine14,15
and the other rich in histidine, glycine, and lysine.15 LK
is coded from the same single-copy gene as HK by differential splicing
of the primary transcript, but LK does not contain regions resembling
D5 and D6 and, instead, has a unique short
light chain D5L. Plasma kallikrein in solution cleaves HK
and eventually yields a stable product (HKa) composed of 2 disulfide-linked chains of molecular weight (Mr) 64 kD and 45 kD. Scott et al16 showed that the cleaved form, HKa,
binds to a greater extent to an activating surface than HK, indicating that HK exists as a procofactor that can be activated via cleavage by kallikrein.
We postulated that the effects of adsorbed HK on neutrophil-surface
interactions would vary with the charge of the surface and that binding
of HK to anionic surfaces would prevent or reduce neutrophil adhesion
and associated elastase release and superoxide production. We examined
neutrophil adhesion to a series of polyurethanes containing different
charges to which various forms of HK had been preadsorbed. The surface
density of the adsorbed protein after adsorption was first quantified.
To elucidate the role of the Mac-1 receptor on cell adhesion, the
availability of Mac-1 binding sites on adsorbed HK and HKa was also
studied, because D3 and D5H of HK are known to
be responsible for mediating the binding of HK to neutrophils.
Polymers
Plasma Proteins
Isolation of Human Neutrophils Neutrophils were isolated from human whole blood anticoagulated with citrate phosphate dextrose using centrifugation with NIM (Cardinal Associates, Santa Fe, NM) as the density gradient isolation medium. The isolated neutrophils were suspended in 2 mL of Hank's balanced salt solution (HBSS) containing 2 mg/mL bovine serum albumin (BSA) at 4°C. The purity, viability, and concentration of the cells were evaluated using a hemacytometer. The viability of the isolated cells in each experiment was always greater than 95%, as determined by trypan blue exclusion, and the concentration was always greater than 107 cells/mL. Untreated neutrophils were then diluted to a standard concentration of 105 cells/mL for the adhesion experiments. Fully activated neutrophils were obtained using the method previously described by Khan et al.18 In brief, isolated neutrophils were incubated with 5 µg/mL cytochalasin B for 3 minutes, followed by the addition of 10 7 M fMLP for 5 minutes
at 37°C to fully activate the neutrophils and to maximally express
Mac-1.18 After the activation, neutrophils were mixed 1:1
with HBSS at 4°C, followed by centrifugation at 250g for 5 minutes. Finally, they were diluted to 105 cells/mL for the
adhesion experiments. This concentration was high enough to provide
good visualization of cell adhesion, but dilute enough to minimize
interactions between cells and to prevent the formation of cell aggregates.
Protein Adsorption To analyze neutrophil interactions with protein adsorbed surfaces, the amount of plasma protein adsorbed was first determined with 125I-labeled proteins at room temperature. HK and HKa were radiolabeled with Na 125I using the IODO-GEN method,19 such that the specific radioactivity of the proteins was greater than 2 mCi/mg. One and one-half inch glass tubings (ID, 12 mm) were coated with 1% solution of PU, NR4, SO3, or GPC in DMAc and dried with nitrogen at room temperature. This process was performed in triplicate and the uniformity of the polymer coating was checked under the optical microscope. Tubings with unevenly coated patterns were not used. Shunts composed of these polymer-coated tubings were used in the adsorption study. Each shunt was injected with HBSS for 2 hours, followed by the adsorption of different concentrations (10, 50, 100, 200, and 500 pmol/mL) of 125I-labeled HK or HKa to the inner surface of the shunt for 6 hours. Nonadsorbed protein was removed with HBSS rinsing at the end of the adsorption period. The radioactivity of the glass tubing was counted using a Beckman Gamma 5500 -counter (Beckman Coulter, Fullerton, CA). We verified
that the radiolabeling did not modify the adsorption process by using
mixtures of labeled and nonlabeled protein solutions.
Neutrophil Adhesion Assay Radial flow chamber.
Neutrophil adhesion was performed using a radial flow chamber with
automated video microscopy and image analyzing systems according to the
design by Dickinson and Cooper.20 In brief, the radial flow
geometry provides a well-defined laminar flow field with a continuous
variation of shear rate for the study of neutrophil adhesion. A
schematic of the radial flow chamber used in the experiments is shown
in Fig 1. The flow cell was constructed according to a similar design by Cozens-Roberts et al.21
The chamber was made of a Plexiglas housing, with 2 optically flat glass discs, one 75 mm and the other 50 mm in diameter. The 50-mm glass
disc was permanently cemented to the housing and bored with a
0.125-inch diameter inlet port through its center. The 75-mm glass disc
was removable and was pretreated under different conditions before the
adhesion experiment. The neutrophil suspension entered at the center of
the chamber and flowed radially outward between the 2 glass discs
through a 300-µm gap.
Disc preparation. The 75-mm glass discs were cleaned by soaking in concentrated chromium trioxide/sulfuric acid solution overnight, in deionized water twice for 1 hour, and were dried overnight in an oven at 60°C. The cleaned discs were spin-coated with 1 wt% polymer solution and dried. The coated discs were incubated in HBSS for approximately 24 hours before the cell adhesion experiments were performed. For experiments involving kininogen adsorption, discs were incubated with HK or HKa for 6 hours, followed by HBSS to remove nonadsorbed protein. Adhesion experiments.
A background scan at every scanned flow field was performed with only
buffer flowing through the chamber. The 105 cells/mL
neutrophil suspension (activated or untreated) was then introduced into
the chamber for the actual cell adhesion experiment with a continuous
field scan over time. The volumetric flow rate was maintained at 2.5 mL/min. This produced a wall shear rate range of approximately 20 to 90 s Purification of Monoclonal Antibodies (MoAbs) Murine MoAb binding to HK were performed as described by Schmaier et al.22 The MoAb 2B5 reacts on immunoblotting with an epitope on the HK heavy chain (D1-D3) and prevents the inhibition of calpain by kininogen. The epitope of 2B5 has been further localized to D2 and D3 of HK.23 The MoAb C11C1 reacts with the light chain of HK on immunoblot and inhibits the coagulant activity.22 Dela Cadena and Colman14 have further shown that C11C1 reacts with D5, the histidine-glycine-lysine rich region of HK.Enzyme-Linked Immunosorbant Assay (ELISA) The conformation of adsorbed HK and HKa, specifically the exposure of the D3 and D5H regions, was determined using an ELISA. Glass discs with 6.4-mm diameter (Richland Glass, Richland, NJ) were coated with polymers using the method described in the previous section. The polymer-coated glass discs were equilibrated with HBSS in a 96-well microtiter plate followed by 6 hours of HK or HKa (100 pmol/mL) adsorption. One percent BSA was then added for 2 hours to block nonspecific binding. Primary antibody at 20 µg/mL (2B5 or C11C1) was added for 2 hours, followed by 1 hour incubation of alkaline phosphatase-conjugated goat antimouse IgG (1:10,000 dilution). p-nitrophenyl phosphate (2 mg/mL) was then added for 0.5 hours for color development. The optical density of each well was detected using a Multiskan Bichromatic plate reader (ICN Biomedicals, Inc, Horsham, PA) at absorbance wavelength of 405 nm.
Kininogen Adsorption Study The surface density of HK and HKa on the polymer-coated surfaces (PU, NR4, SO3, and GPC) at different incubation concentration is shown in Fig 2A and B, respectively. For PU, NR4, and SO3, the plateau region between 200 and 500 pmol/mL concentration indicated the formation of a kininogen monolayer. Assuming the surface density of the adsorbed protein reaches monolayer coverage at 500 pmol/mL condition, the adsorption of both HK and HKa on these 3 surfaces covered more than 80% of surface when the 100 pmol/mL or more concentrated solution was used. The trends of HK and HKa adsorption on all polymer-coated surfaces were similar, with somewhat higher amounts of HKa adsorption seen on all surfaces. The cationic polyurethane-coated surface (NR4) adsorbed the highest amount of protein, followed by the underivatized polyurethane (PU)-, the anionic polyurethane (SO3)-, and the zwitterionic polyurethane (GPC)-coated surface. The amount of adsorbed protein on NR4 was significantly higher (P < .01) when compared with other surfaces, whereas the GPC surface had significantly less protein adsorption (P < .01).
Neutrophil Adhesion Study on HK and HKa Adsorbed Polymer Surfaces Figures 3 and 4 show the adherent cell density on all surfaces studied with untreated neutrophils as the adsorption concentration of kininogen increases from 0 to 100 pmol/mL. At 90 s1, neutrophil adhesion was
observed on the PU, NR4, and SO3 surfaces, but
not on the GPC-coated surface (Figs 3A and 4A). Incubating surfaces
with 100 pmol/mL HK (Fig 3) or HKa (Fig 4) virtually eliminated any
adhesion on the PU- and SO3-coated surfaces. Incubating SO3 with 50 pmol/mL HK or HKa reduced cell adhesion
substantially, but this effect was not observed on PU. On both
SO3 and PU, incubating the surfaces with 10 pmol/mL
kininogen did not result in any decrease in cell adhesion. On the
NR4 surface, no difference in cell adhesion was observed
with or without kininogen adsorption. At 20 s1,
neutrophil adhesion on the PU, NR4, and SO3
surfaces without kininogen preadsorption was the highest, and again no
adhesion was found on the GPC surface (Figs 3B and 4B). By increasing
the kininogen adsorption, PU and SO3 showed a decrease in
cell adhesion, and the decrease was most significant on the
SO3 surface. In fact, almost all adhesion was eliminated on
SO3 preadsorbed with 100 pmol/mL HKa. No decrease in the
cell adhesion was again observed on NR4 after incubating
with HK or HKa. The degree of cell adhesion on surfaces preadsorbed
with 200 pmol/mL was similar to that preadsorbed with 100 pmol/mL and
is not shown here.
Morphology of Adherent Neutrophils The degree of spreading of adherent neutrophils was quantified using the CI shown in Table 1. In the absence of kininogens, the CI on SO3 was the lowest compared to PU and NR4 surfaces. CI increased on all 3 surfaces after the preadsorption of HK and HKa, and the increase was most significant on the SO3 surface. Lim and Cooper7 have shown a direct correlation between neutrophil respiratory burst and spreading on the bare SO3 surface. Other researchers have also shown that the increased adherence and spreading of neutrophil to nonphagocytosable surfaces lead to increased activation of oxygen metabolism and to defective bactericidal activity.25,26 We show here that, in the presence of adsorbed HK or HKa, cell adhesion and also the degree of spreading of the adherent cells is reduced significantly on the SO3 surface.
The Adhesion of Activated Neutrophils on Surfaces Adsorbed With
Kininogens
Exposure of Mac-1 Binding Sites on Adsorbed Kininogen
High molecular weight kininogen, especially the activated form (HKa),
has a high affinity towards negatively charged surfaces and can
displace fibrinogen from glass surfaces, a phenomenon that has been
referred to as the Vroman effect.28,29
The authors acknowledge the Blood Bank of Delaware for donating the
whole blood used in the neutrophil adhesion study.
Submitted July 20, 1998; accepted June 11, 1999.
Supported by the National Institutes of Health through Grants No.
HL-47179 and HL-56914 (Program Project).
The publication costs of this
article were defrayed in part by
page charge payment. This article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. section
1734 solely to indicate this fact.
Address reprint requests to Robert W. Colman, MD, Sol Sherry Thrombosis
Research Center, Temple University School of Medicine, Philadelphia, PA
19140; e-mail: colmanr{at}astro.ocis.temple.edu.
1.
Bohler J, Kramer P, Gotze O, Schartz O, Scheler F:
Leukocyte counts and complement activation during pump-driven and arteriovenous haemofiltration.
Cont Hephrol
36:15, 1983
2.
Ryhanen P, Eleyva E, Hollman A, Naolinen R, Pihlajanieme R, Saarela E:
Changes in peripheral blood leukocyte counts, lymphocyte subpopulation and in vitro transformation after heart valve replacement.
J Thorac Cardiovasc Surg
77:259, 1979[Medline]
[Order article via Infotrieve]
3.
Boretos JW, Pierce WS:
Segmented polyurethane: A new elastomer for biomedical applications.
Science
158:1481, 1967
4.
Grasel TG, Cooper SL:
Properties and biological interactions of polyurethane anionomers: Effect of sulfonate incorporation.
J Biomed Mater Res
23:311, 1989[Medline]
[Order article via Infotrieve]
5.
Okkema AZ, Yu XH, Cooper SL:
Physical and blood contacting characteristics of propyl sulphonate grafted Biomer.
Biomaterials
12:3, 1991[Medline]
[Order article via Infotrieve]
6.
Lim F, Cooper SL:
Effect of sulphonate incorporation on in vitro leucocyte adhesion to polyurethanes.
Biomaterials
16:457, 1995[Medline]
[Order article via Infotrieve]
7.
Lim F, Cooper SL:
Chemiluminescent oxidative products generated by in vitro leukocyte materials interactions.
J Mater Sci Mater Med
7:69, 1996
8.
Baumgartner JN, Yang CZ, Cooper SL:
Physical property analysis and bacterial adhesion on a series of phosphonated polyurethanes.
Biomaterials
18:831, 1997[Medline]
[Order article via Infotrieve]
9.
Yung LL, Cooper SL:
Neutrophil adhesion on phosphorylcholine-containing polyurethanes.
Biomaterials
19:31, 1998[Medline]
[Order article via Infotrieve]
10.
Ishihara K, Inoue H, Kurita K, Nakabayashi N:
Selective adhesion of platelets on a polyion complex composed of phospholipid polymers containing sulfonate groups and quaternary ammonium groups.
J Biomed Mater Res
28:1347, 1994[Medline]
[Order article via Infotrieve]
11.
Iwasaki Y, Kurita K, Ishihara K, Nakabayashi N:
Effect of reduced protein adsorption on platelet adhesion at the phospholipid polymer surfaces.
J Biomater Sci Polymer Ed
8:151, 1996
12.
Jacobsen S, Kriz M:
Some data on two purified kininogens from human plasma.
Br J Pharmacol
29:25, 1967[Medline]
[Order article via Infotrieve]
13.
Wachtfogel YT, DeLa Cadena RA, Kunapuli SP, Rick L, Miller M, Schulze RL, Altieri DC, Edgington TS, Colman RW:
High molecular weight kininogen binds to Mac-1 on neutrophils by its heavy chain (domain 3) and its light chain (domain 5).
J Biol Chem
269:19307, 1994
14.
DeLa Cadena RA, Colman RW:
The sequence HGLGHGHEQQHGLGHGH in the light chain of high molecular weight kininogen serves as a primary structural feature for zinc-dependent binding to an anionic surface.
Prot Sci
1:151, 1992[Medline]
[Order article via Infotrieve]
15.
Kunapuli SP, DeLa Cadena RA, Colman RW:
Deletion mutagenesis of high molecular weight kininogen light chain: Identification of two anionic surface binding subdomains.
J Biol Chem
268:2486, 1993
16.
Scott CF, Silver LD, Schapira M, Colman RW:
Cleavage of high molecular weight kininogen markedly enhances its coagulant activity: Evidence that this molecule exists as a procofactor.
J Clin Invest
73:954, 1984
17.
Goddard RJ, Cooper SL:
Polyurethane cationomers with pendant trialkylammonium groups: Effects of ion content, alkyl group, and neutralizing anion.
J Polym Sci Polym Phys Ed
17:1557, 1994
18.
Khan MMH, Kunapuli SP, Lin Y, Majluf-Cruz A, DeLa Cadena RA, Cooper SL, Colman RW:
Three noncontiguous peptides comprise binding sites on high-molecular-weight kininogen to neutrophils.
Am J Physiol
275:H145, 1998
19.
Fraker PJ, Speck JC:
Protein and cell membrane iodinations with a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a-diphrenylglycoluril.
Biochem Biophys Res Commun
80:849, 1978[Medline]
[Order article via Infotrieve]
20.
Dickinson RB, Cooper SL:
Analysis of shear-dependent bacterial adhesion kinetics to biomaterials surfaces.
Am Inst Chem Eng J
41:2160, 1995
21.
Cozens-Roberts C, Quinn JA, Lauffenburger DA:
Receptor-mediated adhesion phenomena: Model studies with the radial-flow detachment assay.
Biophys J
58:107, 1990[Medline]
[Order article via Infotrieve]
22.
Schmaier AH, Schutsky D, Farber A, Silver LD, Bradford HN, Colman RW:
Determination of the bifunctional properties of high molecular weight kininogen by studies with monoclonal antibodies directed to each of its chains.
J Biol Chem
262:1405, 1987
23.
Jiang Y, Nawarawong W, Meloni FJ, Schmaier AH:
Insights on monoclonal antibodies to kininogens' heavy chain which influence kininogens' binding to platelets.
Thromb Haemost
68:143, 1992[Medline]
[Order article via Infotrieve]
24.
Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC:
Measurement of protein using bicinchoninic acid.
Anal Biochem
150:76, 1985[Medline]
[Order article via Infotrieve]
25.
Klock JC, Bainton DF:
Degranulation and abnormal bactericidal function of granulocytes procured by reversible adhesion to nylon wool.
Blood
48:149, 1976
26.
Wright DG, Gallin JI:
Secretory responses of human neutrophils: Exocytosis of specific (secondary) granules by human neutrophils during adherence in vitro and during exudation in vivo.
J Immunol
123:285, 1979
27. Yung LL, Lim F, Khan MMH, Kunapuli SP, Rick L, Colman RW, Cooper
SL: High molecular weight kininogen preadsorbed to surfaces markedly
reduces leukocyte adhesion. Biomaterials (in press)
28.
Vroman L, Adams AL:
Findings with the recording ellipsometer suggesting rapid exchange of specific plasma proteins at liquid/solid interfaces.
Surface Sci
16:438, 1969
29.
Schmaier AH, Silver L, Adams AL, Fischer GC, Munoz PC, Vroman L, Colman RW:
The effect of high molecular weight kininogen on surface-adsorbed fibrinogen.
Thromb Res
33:51, 1984[Medline]
[Order article via Infotrieve]
30.
Ishihara K, Nomura H, Mihara T, Kurita K, Iwasaki Y, Nakabayashi N:
Why do phospholipid polymers reduce protein adsorption?
J Biomed Mater Res
39:323, 1998[Medline]
[Order article via Infotrieve]
31.
Nagel JA, Dickinson RB, Cooper SL:
Bacterial adhesion to polyurethane surfaces in the presence of pre-adsorbed high molecular weight kininogen.
J Biomater Sci Polymer Ed
7:769, 1996
32.
Gustafson EJ, Lukasiewicz H, Wachtfogel YT, Norton KJ, Schmaier AH, Niewiarowski S, Colman RW:
High molecular weight kininogen inhibits fibrinogen binding to cytoadhesins of neutrophils and platelets.
J Cell Biol
109:377, 1989
33.
Sheng N, Fairbanks MB, Heinrikson RL, Canziani G, Chaiken IM, Mosser DM, Colman RW:
The integrin CD11b/CD18 (Mac-1) is a major cell receptor for high molecular weight kininogen.
FASEB J
11:A920, 1997
34.
DeLa Cadena RA, Wachtfogel YT, Colman RW:
Contact activation pathway: Inflammation and coagulation, in
Colman RW,
Hirsh J,
Marder VJ,
Salzman EW
(eds):
Hemostasis and Thrombosis: Basic Principles and Clinical Practice (ed 3). Philadelphia, PA, Lippincott, 1994, p 219
35.
Chapman D:
Biomembranes and new hemocompatible materials.
Langmuir
9:39, 1993
36.
Ishihara K, Oshida H, Endo Y, Ueda T, Watenabe A, Nakabayashi N:
Hemocompatibility of human whole blood on polymers with a phospholipid polar group and its mechanism.
J Biomed Mater Res
26:1543, 1992[Medline]
[Order article via Infotrieve]
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|||||||
 |
 |
 |
 |
 |
 |
 |
 |
||
| Copyright © 1999 by American Society of Hematology Online ISSN: 1528-0020 | |||||||||
TOP
ABSTRACT
INTRODUCTION
m
2 or Mac-1, were examined. On PU,
NR4, and SO3, kininogen adsorption reached 80%
of monolayer coverage when 100 pmol/mL or higher concentration of
protein solutions were used. The NR4 surface adsorbed the
most kininogen along with a high exposure of D3 and
D5H. The availability of D3 and D5H
allowed neutrophils to bind to the surface via the Mac-1 receptor;
thus, on the NR4 surface, adsorbed kininogens lost their
antiadhesive property, which resulted in a high degree of neutrophil
adhesion. Increasing Mac-1 expression by exposure to fMLP increased the
neutrophil adhesion on this surface. In contrast, exposure of
D3 and D5H on SO3 was significantly
less, because HK binds to anionic surfaces with similar protein
sequences used for cell binding. This low binding site exposure
preserved the antiadhesive property of HK. GPC was resistant to
neutrophil adhesion even in the absence of adsorbed kininogens because
of its phosphorylcholine moiety. Thus, both SO3 coupled
with kininogen (or kininogen peptides) and GPC have the potential to
markedly reduce neutrophil adhesion to biomaterial devices.
-glycerophosphorylcholine.
[observed area])/(observed
perimeter)
)
NR4, (
) SO3, and (
) GPC as a function of
solution concentration. Error bars indicate the standard deviation of
the mean.
) adsorbed with 10 pmol/mL HK, (
) adsorbed with 100 pmol/mL HK (n = 3). (*)
indicates no adherent cells found. Error bars indicate the standard
deviation of the mean. §P < .01 versus bare polymer
surface.
