Cell-Binding Domain Context Affects Cell Behavior on
Sarah C. Heilshorn,†Julie C. Liu,†and David A. Tirrell*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Mail Code 210-41,
Pasadena, California 91125
Received June 29, 2004; Revised Manuscript Received October 5, 2004
A family of artificial extracellular matrix proteins developed for application in small-diameter vascular
grafts is used to examine the importance of cell-binding domain context on cell adhesion and spreading.
The engineered protein sequences are derived from the naturally occurring extracellular matrix proteins
elastin and fibronectin. While each engineered protein contains identical CS5 cell-binding domain sequences,
the lysine residues that serve as cross-linking sites are either (i) within the elastin cassettes or (ii) confined
to the ends of the protein. Endothelial cells adhere specifically to the CS5 sequence in both of these proteins,
but cell adhesion and spreading are more robust on proteins in which the lysine residues are confined to the
terminal regions of the chain. These results may be due to altered protein conformations that affect either
the accessibility of the CS5 sequence or its affinity for the R4?1integrin receptor on the endothelial cell
surface. Amino acid choice outside the cell-binding domain can thus have a significant impact on the behavior
of cells cultured on artificial extracellular matrix proteins.
A common goal in biomaterials design is the engineering
of cell-adhesive materials. The identification within native
proteins of numerous cell-binding domains that mediate cell-
matrix interactions,1-7most notably the tripeptide sequence
RGD,8-10has greatly aided the pursuit of this goal. Im-
mobilization of RGD and other adhesion peptides onto a
variety of polymeric substrates has been utilized to enhance
cell adhesion.11Although many of these peptides have been
reduced to their “minimal binding sequences”, the identity
of flanking amino acid residues has been shown to alter
activity.10,12A variety of experiments, including site-directed
mutagenesis and phage display, have been used to determine
neighboring sequences that enhance cell adhesion.10,12-15For
example, replacing the serine residue in the wild-type peptide
GRGDSPC with asparagine yields a peptide that is six times
more effective at inhibiting cell attachment to fibronectin.12
Furthermore, RGD sequences flanked by cysteines that form
disulfide bridges and cause the peptide to cyclize exhibit
stronger cell adhesion through the RIIb?3integrin than linear
RGD sequences, perhaps because the cyclic peptide mimics
the conformation of the native ligand.13,16
While engineered peptides grafted onto synthetic polymer
substrates yield cell-adhesive materials, an alternative ap-
proach involves the design and synthesis of more complex
polypeptides that mimic some of the essential properties of
the extracellular matrix (ECM). Genetic engineering allows
highly specialized artificial proteins to be obtained in good
yield with high fidelity,17-19and the modular nature of
recombinant DNA methodology allows facile synthesis of
engineered proteins containing structural sequences and cell-
binding domains derived from naturally occurring proteins
or created by rational design. This approach has yielded cell-
adhesive biomaterials containing cadherin-like domains,
RGD sequences, and CS5 sequences derived from the
alternatively spliced IIICS fragment of fibronectin.20-29
The protein engineering approach to biomaterials design
allows incorporation of full-length cell-binding domains, as
opposed to the minimal binding sequences typically used to
modulate cell adhesion to synthetic biomaterials. It is
reasonable to anticipate that full-length cell-binding domains
may elicit more authentic responses. For example, engineered
elastin-like proteins containing the full CS5 sequence were
shown to promote cell adhesion,21while those incorporating
the minimal binding sequence REDV were not adhesive.30
Furthermore, protein engineering offers the prospect of
additional levels of control in that the complete primary
sequence can (and must) be specified, not just the sequences
integral and proximal to the cell-binding domains. To what
extent do more remote elements of amino acid sequence
affect cell-adhesion properties? We address this question here
by characterizing cell response to engineered proteins that
present the CS5 cell-binding domain within two different
The elastin-like proteins examined in this work were
originally designed for application in small-diameter vascular
grafts.22,24,25Cardiovascular disease afflicts more than 61
million Americans31and causes 4 million deaths in Europe
every year.32Although large-diameter grafts in regions of
high blood flow remain patent for many years, replacement
of small- and medium-diameter vessels has met with limited
success.33-35Unopposed proliferation of myofibroblasts leads
* To whom correspondence should be addressed. Phone: (626) 395-
3140. Fax (626) 793-8472. E-mail: email@example.com.
†These authors contributed equally to this work.
Biomacromolecules 2005, 6, 318-323
10.1021/bm049627q CCC: $30.25© 2005 American Chemical Society
Published on Web 12/04/2004
to stenosis of such grafts and subsequent thrombosis. It is
believed that (i) the inability of the graft material to support
the development of an endothelial cell monolayer and (ii)
the compliance mismatch between the prosthetic graft and
the host tissue both contribute to graft failure. Therefore,
our initial design criteria in engineering materials for small-
diameter vascular grafts were (i) enhancing endothelial cell
adhesion and (ii) tuning the elastic modulus of the material
to match that of the affected artery. To meet these goals,
CS5 cell-binding domains were used to enhance endothelial
cell adhesion and elastin-like repeats were included to confer
elastomeric behavior. Past studies have shown that these
engineered proteins can support adhesion of endothelial cells
in physiologically relevant fluid flows.24Lysine residues were
incorporated into the sequences as specific cross-linking sites,
allowing formation of free-standing films with tensile moduli
similar to those of native elastins.22,25,36,37
In this study, we examine the importance of amino acid
context in affecting the response of human umbilical vein
endothelial cells (HUVEC) to the CS5 cell-binding domain.
The engineered protein sequences are listed in Figure 1. Ki
contains lysine residues (K) at intervals of 25 amino acids
internal (i) to the elastin-like domain. The CS5 cell-binding
domain and elastin-like domain are repeated three times
within the protein. Ki* is a similar protein except for the
fact that the minimal binding sequence of the CS5 domain
has been scrambled to provide a negative control. Compari-
son of HUVEC adhesion to Ki and Ki*, in addition to
peptide inhibition studies, has shown that cell adhesion to
the CS5 cell-binding domain in Ki is sequence-specific.26
Kt also includes three repeats of the CS5 cell-binding domain
and the elastin-like cassette; however, the lysine residues are
located only at the N- and C-termini (t). Peptide inhibition
studies performed on a protein similar to Kt, containing five
repeats of the CS5 and elastin-like domains, have demon-
strated that HUVEC adhesion is primarily a consequence of
sequence-specific interactions with the CS5 cell-binding
domain.24Although endothelial cells adhere specifically to
the cell-binding domain in this family of engineered elastin-
like proteins, the slight modification in primary sequence
from Ki to Kt is shown here to significantly affect HUVEC
behavior. These results show clearly that the context of the
engineered protein, i.e., the identity of amino acids at sites
distant from the cell-binding domain, can affect cell spread-
ing and adhesion.
Materials and Methods
Protein Expression and Purification. Ki, Ki*, and Kt
were expressed in E. coli and purified as previously
described.22,24,25Purity was assessed by SDS-PAGE, mass
spectrometry, and Western blotting with anti-T7 tag-
horseradish peroxidase conjugate antibody (Amersham). The
molecular weights of the three proteins are 37 120, 37 120,
and 42 974, respectively.
Protein Adsorption. To determine protein adsorption
isotherms, 50 µL of each protein solution (0.05-8 mg/mL
in PBS) was adsorbed onto a 96-well tissue culture poly-
styrene plate overnight at 4 °C. The substrates were rinsed
three times with 100 µL of PBS, and adsorbed protein was
quantified via the bicinchoninic acid (BCA) method.38,39
Briefly, 50 µL of PBS was added to the wells containing
adsorbed protein. To create a calibration curve, 50 µL
aliquots of protein solutions of known concentrations were
added to the 96-well plate. An equal volume of the working
reagent containing an alkaline Cu2+solution was added and
incubated at 60 °C for 1 h. After the plates equilibrated to
room temperature, the absorbance was read at 562 nm. The
absorbance values observed for the protein standards were
fit to a linear calibration curve, which was used to determine
the amount of adsorbed protein in each sample.40Each
concentration was tested in at least two independent experi-
ments in triplicate. For cell culture experiments, solutions
of engineered proteins (1 mg/mL in PBS unless otherwise
noted) and fibronectin (10 µg/mL in PBS) were adsorbed
onto tissue culture polystyrene overnight at 4 °C. The treated
substrates were rinsed with PBS, blocked with 0.2% heat-
inactivated bovine serum albumin (BSA fraction V, Sigma)
for 30 min, and rinsed again with PBS.
Cell Culture. Human umbilical vein endothelial cells
(HUVEC, Bio Whittaker) were maintained in a 37 °C, 5%
CO2 humidified environmental chamber. The cells were
grown in Endothelial Growth Medium-2 (2% serum, Bio
Whittaker), which was replaced every 2 days. Near confluent
HUVEC cultures were passaged nonenzymatically by treat-
ment with 0.61 mM EDTA (Gibco). Passages 2-10 were
Cell Spreading. Scanning Electron Microscopy. HUVEC
were seeded on engineered proteins in serum-free Endothelial
Basal Medium-2 (EBM-2, Bio Whittaker) and incubated for
30 min. Substrates were rinsed twice with PBS and fixed
with formaldehyde. Fixed samples were subjected to critical
point drying and gold/platinum sputter coating prior to
Figure 1. Amino acid sequences of the artificial extracellular matrix proteins. Ki has three cassette repeats with lysine residues internal to the
elastin-like domain. Ki* is similar but contains a scrambled CS5 binding domain as a negative control. Kt has three cassette repeats with lysine
residues at the termini.
Cell-Binding Domain ContextBiomacromolecules, Vol. 6, No. 1, 2005
imaging on a JEOL 6400 V scanning electron microscope.
Two independent experiments were performed for each
Phase Contrast Microscopy. HUVEC in serum-free EBM-2
were allowed to spread on engineered proteins and imaged
at 15 min intervals by using a 10× phase contrast objective
on a Nikon Eclipse TE300 inverted microscope. Images were
density-sliced to determine the number of well-spread (i.e.,
dark) versus nonspread (i.e., bright and refractive) cells using
Scion Image for Windows.26Three independent experiments
were performed. A one-tailed two-sample t-test that assumed
equal variances was applied to determine statistical signifi-
Cell Resistance to Detachment. Cell resistance to normal
detachment forces was measured as previously described.26
Briefly, cells were fluorescently labeled with calcein ac-
etoxymethyl ester (Molecular Probes). Fluorescently labeled
HUVEC were incubated on adsorbed protein substrates for
30 min. A solution of Percoll (21% w/w in PBS, Sigma)
was added to each well, and the plates were centrifuged
upright for 10 min at 1, 100, 1000, 2000, and 3000g.
Nonadherent cells were removed, and the remaining cells
were quantified by their fluorescence (excitation at 485 nm
and emission at 538 nm). A cell adhesion index (CAI) was
calculated as the fluorescence reading of a test well divided
by the fluorescence reading of HUVEC attached to fibronec-
tin subjected to 1g (2.6 pN). Because Percoll has a higher
density (1.123 g/mL) than the cells (∼1.07 g/mL), a buoyant
force is exerted on the cells. Using Archimedes’ theorem,41,42
the range of detachment forces applied was estimated as 2.6-
780 pN. At least three independent experiments with six
replicates each were performed.
Results and Discussion
Protein Synthesis and Characterization. A typical wet
cell mass from a 10 L batch fermentation was 200-250 g,
and expression yields for proteins Ki, Ki*, and Kt were 10-
20 mg/g wet cell mass. The expressed proteins were readily
purified to provide multigram quantities of material. The
engineered proteins were adsorbed onto tissue-culture poly-
styrene, and their adsorption isotherms were determined
(Figure 2) to ensure that the adsorbed films contained similar
densities of cell-binding domains. At high (g4 mg/mL) and
low (e0.05 mg/mL) solution concentrations, both lysine
variants (Ki and Kt) adsorbed similarly to the polystyrene
substrate. At intermediate concentrations (0.08-2 mg/mL),
small but reproducible differences in adsorption levels were
detected. Ki and Ki* exhibited indistinguishable isotherms
(data not shown).
HUVEC Resistance to Detachment Forces. Although
previously published results showed that HUVEC adhesion
to this family of engineered elastin-like proteins was
primarily due to sequence-specific interactions with the CS5
cell-binding domain, we show here that amino acid choice
within the elastin-like domains can significantly affect
HUVEC resistance to detachment. To ensure that these
effects were not due to differences in cell-binding domain
density, we examined HUVEC adhesion over a range of
protein concentrations (Figure 3). Assuming that each cell
has 5.8 × 106R4?1 integrin receptors for the REDV
sequence, of which one-half are available for surface
interactions,43and a spread cell area of 150 µm2after 30
min of incubation (see Figure 5b, below), the adsorbed
protein surfaces displayed 10-100 times more cell-binding
domains than available receptors per cell. Over this range
of cell-binding domain densities, the cell adhesion indices
(CAI) on the terminal lysine protein (Kt) were consistently
higher than those observed on the protein with lysine residues
internal to the elastin domain (Ki). At high cell-binding
domain densities, the CAI on Kt after a 10 min exposure to
an estimated 260 pN detachment force was nearly 100%
(relative to HUVEC adhesion on fibronectin after a 10 min
exposure to a 2.6 pN detachment force). In contrast, the CAI
on Ki at these high cell-binding domain densities was ∼20%.
Thus, Kt appears to be a more highly adhesive substrate than
Figure 2. Adsorption isotherms of engineered proteins with terminal
lysine residues, Kt (O), and lysine residues internal to the elastin-
like domain, Ki (B), on tissue-culture polystyrene. Data are from one
representative experiment performed in triplicate. Error bars represent
one standard deviation.
Figure 3. HUVEC resistance to a 260 pN normal detachment force
after 30 min of incubation on tissue-culture polystyrene treated with
varying amounts of Ki (B) or Kt (O) protein. Three independent
experiments with six replicates were performed. Error bars represent
one standard deviation.
Biomacromolecules, Vol. 6, No. 1, 2005 Heilshorn et al.
Ki, even though both engineered proteins include the same
The relationships between CAI and cell-binding domain
density also differed for the two substrates. On Kt, cell
adhesion increased with increasing protein concentration up
to a cell-binding domain density of 40 × 1010/mm2, which
corresponds to 20 cell binding domains per available
receptor, and then showed no further dependence on protein
concentration. Over a limited range of cell-binding domain
densities (35-90 × 1010/mm2), the CAI on Ki was ∼50%;
at higher densities, the CAI was reduced to ∼20%. The origin
of this complex relationship between cell-binding domain
density and CAI is unknown but may reflect a change in
the conformation of the protein (or of the cell-binding
domain) as a function of surface coverage. An alternative
explanation might be cell loss due to substrate fracture at
higher cell-binding domain densities; however, no evidence
of protein delamination was observed by SEM (data not
shown). Furthermore, no differences in cell viability were
detected after 24 h between cells grown on either of the
engineered protein substrates and those grown on fibronectin
(data not shown). Cell viability was assessed by monitoring
cleavage of the tetrazolium salt WST-1. Subsequent HUVEC
studies were conducted on adsorbed protein surfaces created
from 1 mg/mL bulk solutions, which correspond to cell-
binding domain densities of 80 and 100 × 1010/mm2for
proteins Ki and Kt, respectively. This choice of bulk
concentration for the adsorption solution allows us to
compare cell behavior on Ki and Kt at cell-binding domain
densities that elicit the highest cell adhesion indices on each
Over a range of detachment forces (26-780 pN), HUVEC
cultured on Kt had consistently higher CAI than those on
Ki (Figure 4). As expected, the CAI for both Kt and Ki
decreased as higher detachment forces were applied. The CAI
for HUVEC on Ki*, the scrambled negative control protein,
did not exceed 23% even at the lowest detachment forces
examined in this work. These trends were also observed when
the proteins were adsorbed on glass surfaces (see Supporting
Information). Because there is a heptahistidine tag in Ki and
Ki* but not in Kt, we verified that removal of the tag from
Ki does not affect the measured CAI at 26 pN (see
Supporting Information). Thus, we believe that the change
in context of the cell-binding domain, and not the supporting
substrate or the heptahistidine tag, leads to a higher CAI on
Kt as compared to Ki.
HUVEC Spreading on Engineered Proteins. The amino
acid context of the cell-binding domain was found to affect
cell spreading as well as cell attachment. After 30 min of
incubation, HUVEC on Ki remained small and rounded
(Figure 5a) while those on Kt exhibited extended processes
and had larger spread-cell areas (Figure 5b). When compared
to cells spread on fibronectin, cells on Ki and Kt have
smaller spread-cell areas. The SEM images in Figure 5 are
representative of typical HUVEC morphologies on Ki and
Kt. To confirm these morphological observations, the
percentages of well-spread cells were quantified by phase
contrast microscopy. A larger percentage of HUVEC were
well spread on Ki (g20%) than on the sequence-scrambled,
negative control protein Ki* (e10%) after 60 min. At least
twice as many HUVEC were well spread on the terminal
lysine protein (Kt) as on the internal lysine protein (Ki) after
15-60 min of incubation (Figure 6). Furthermore, Kt elicited
more rapid cell spreading than Ki, although the rate of cell
spreading was substantially lower than that observed on
The results presented here demonstrate that context, i.e.,
amino acid choice, can affect the activity of cell-binding
domain sequences in engineered proteins. It has been
demonstrated previously that flanking amino acid residues
can alter cell adhesion to binding sequences, but this work
shows that amino acids located 15 or more residues away
from the primary binding sequence can have a significant
impact on HUVEC spreading and adhesion. Cell adhesion
to this family of engineered proteins is primarily a conse-
quence of sequence-specific recognition of the CS5 cell-
Figure 4. HUVEC resistance to normal detachment forces after 30
min of incubation on tissue-culture polystyrene treated with various
proteins: Kt (O), Ki (B), and Ki* (2). Three independent experiments
with six replicates were performed. Error bars represent one standard
Figure 5. Scanning electron micrographs of HUVEC after 30 min of
incubation on tissue-culture polystyrene treated with (a) Ki and (b)
Kt. Two independent experiments were performed for each substrate.
Both images are at the same magnification. Scale bar represents 7
Cell-Binding Domain ContextBiomacromolecules, Vol. 6, No. 1, 2005
binding domain regardless of context. Although all of the
engineered proteins displayed similar adsorption isotherms,
HUVEC morphology and adhesion to the adsorbed protein
films were found to be context dependent. We suggest that
changes in amino acid context may modify the protein
conformation, which in turn may alter the accessibility or
receptor affinity of the cell-binding domain.44
Acknowledgment. This work was supported by NIH
Grant 5 R01 HL59987-03, NSF Grant BES-9901648, and a
Whitaker graduate fellowship. We thank Kathleen Di Zio
for helpful discussion regarding protein purification, Paul
Nowatzki for providing protein samples, the Electron Mi-
croscopy Laboratory in the Biology Division at Caltech and
Robert Strittmatter for help with the scanning electron
micrographs, and Krystle Wang, Gustavo Olm, and Regina
Wilpiszeski for help with the BCA and cell detachment
Supporting Information Available. Supporting figures
include 1. Cell resistance to detachment forces on engineered
proteins adsorbed to glass substrates, 2. Western analysis
confirming complete cleavage of heptahistidine- and T7-tags,
and 3. Cell resistance to detachment forces on engineered
proteins with heptahistidine- and T7-tags removed. This
material is available free of charge via the Internet at http://
References and Notes
(1) Humphries, M. J.; Akiyama, S. K.; Komoriya, A.; Olden, K.; Yamada,
K. M. J. Cell Biol. 1986, 103, 2637-2647.
(2) Humphries, M. J.; Komoriya, A.; Akiyama, S. K.; Olden, K.; Yamada,
K. M. J. Biol. Chem. 1987, 262, 6886-6892.
(3) Komoriya, A.; Green, L. J.; Mervic, M.; Yamada, S. S.; Yamada,
K. M.; Humphries, M. J. J. Biol. Chem. 1991, 266, 15075-15079.
(4) Graf, J.; Iwamoto, Y.; Sasaki, M.; Martin, G. R.; Kleinman, H. K.;
Robey, F. A.; Yamada, Y. Cell 1987, 48, 989-996.
(5) Graf, J.; Ogle, R. C.; Robey, F. A.; Sasaki, M.; Martin, G. R.;
Yamada, Y.; Kleinman, H. K. Biochemistry 1987, 26, 6896-6900.
(6) Sephel, G. C.; Tashiro, K. I.; Sasaki, M.; Greatorex, D.; Martin, G.
R.; Yamada, Y.; Kleinman, H. K. Biochem. Biophys. Res. Commun.
1989, 162, 821-829.
(7) Tashiro, K.; Sephel, G. C.; Weeks, B.; Sasaki, M.; Martin, G. R.;
Kleinman, H. K.; Yamada, Y. J. Biol. Chem. 1989, 264, 16174-
(8) Pierschbacher, M.; Hayman, E. G.; Ruoslahti, E. Proc. Natl. Acad.
Sci. 1983, 80, 1224-1227.
(9) Pierschbacher, M. D.; Ruoslahti, E. Nature 1984, 309, 30-33.
(10) Pierschbacher, M. D.; Ruoslahti, E. Proc. Natl. Acad. Sci. 1984, 81,
(11) For review of RGD biomaterials, see: Hersel, U.; Dahmen, C.;
Kessler, H. Biomaterials 2003, 24, 4385-4415.
(12) Pierschbacher, M. D.; Ruoslahti, E. J. Biol. Chem. 1987, 262, 17294-
(13) Koivunen, E.; Wang, B. C.; Ruoslahti, E. Bio-Technology 1995, 13,
(14) Healy, J. M.; Haruki, M.; Kikuchi, M. Protein Pept. Lett. 1996, 3,
(15) Richards, J.; Miller, M.; Abend, J.; Koide, A.; Koide, S.; Dewhurst,
S. J. Mol. Biol. 2003, 326, 1475-1488.
(16) Cheng, S.; Craig, W. S.; Mullen, D.; Tschopp, J. F.; Dixon, D.;
Pierschbacher, M. D. J. Med. Chem. 1994, 37, 1-8.
(17) Krejchi, M. T.; Atkins, E. D. T.; Waddon, A. J.; Fournier, M. J.;
Mason, T. L.; Tirrell, D. A. Science 1994, 265, 1427-1432.
(18) Petka, W. A.; Harden, J. L.; McGrath, K. P.; Wirtz, D.; Tirrell, D.
A. Science 1998, 281, 389-392.
(19) Yu, S. J. M.; Conticello, V. P.; Zhang, G. H.; Kayser, C.; Fournier,
M. J.; Mason, T. L.; Tirrell, D. A. Nature 1997, 389, 167-170.
(20) Urry, D. W.; Pattanaik, A.; Xu, J.; Woods, T. C.; McPherson, D. T.;
Parker, T. M. J. Biomater. Sci.-Polym. Ed. 1998, 9, 1015-1048.
(21) Panitch, A.; Yamaoka, T.; Fournier, M. J.; Mason, T. L.; Tirrell, D.
A. Macromolecules 1999, 32, 1701-1703.
(22) Welsh, E. R.; Tirrell, D. A. Biomacromolecules 2000, 1, 23-30.
(23) Halstenberg, S.; Panitch, A.; Rizzi, S.; Hall, H.; Hubbell, J. A.
Biomacromolecules 2002, 3, 710-723.
(24) Heilshorn, S. C.; Di Zio, K.; Welsh, E. R.; Tirrell, D. A. Biomaterials
2003, 24, 4245-4252.
(25) Nowatzki, P. J.; Tirrell, D. A. Biomaterials 2004, 25, 1261-1267.
(26) Liu, J. C.; Heilshorn, S. C.; Tirrell, D. A. Biomacromolecules 2004,
(27) Nagaoka, M.; Ise, H.; Akaike, T. Biotechnol. Lett. 2002, 24, 1857-
(28) Asakura, T.; Tanaka, C.; Yang, M. Y.; Yao, J. M.; Kurokawa, M.
Biomaterials 2004, 25, 617-624.
(29) Kobatake, E.; Onoda, K.; Yanagida, Y.; Aizawa, M. Biomacromol-
ecules 2000, 1, 382-386.
(30) Nicol, A.; Gowda, D. C.; Parker, T. M.; Urry, D. W. In Biotechnology
and BioactiVe Polymers; Gebelein, C. G., Carraher, C. E., Eds.;
Plenum Press: New York, 1994.
(31) American Heart Association. Heart Disease and Stroke Statistics-
2003 Update; American Heart Association: Dallas, TX, 2002.
(32) Rayner, M.; Petersen, S. European CardioVascular Disease Statistics;
British Heart Foundation: London, 2000.
(33) Bos, G. W.; Poot, A. A.; Beugeling, T.; van Aken, W. G.; Feijen, J.
Arch. Physiol. Biochem. 1998, 106, 100-115.
(34) Salacinski, H. J.; Tiwari, A.; Hamilton, G.; Seifalian, A. M. Med.
Biol. Eng. Comput. 2001, 39, 609-618.
(35) Nugent, H. M.; Edelman, E. R. Circ. Res. 2003, 92, 1068-1078.
(36) Fung, Y. C. Biomechanics: Mechanical Properties of LiVing Tissues,
2nd ed.; Springer-Verlag: New York, 1993.
(37) Di Zio, K.; Tirrell, D. A. Macromolecules 2003, 36, 1553-1558.
(38) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner,
F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B.
J.; Klenk, D. C. Anal. Biochem. 1985, 150, 76-85.
(39) Wiechelman, K. J.; Braun, R. D.; Fitzpatrick, J. D. Anal. Biochem.
1988, 175, 231-237.
(40) Calibration curves were identical when using either protein solutions
of known concentrations or known amounts of protein dried down
(41) Channavajjala, L. S.; Eidsath, A.; Saxinger, W. C. J. Cell Sci. 1997,
(42) Koo, L. Y.; Irvine, D. J.; Mayes, A. M.; Lauffenburger, D. A.;
Griffith, L. G. J. Cell Sci. 2002, 115, 1423-1433.
(43) Hubbell, J. A.; Massia, S. P.; Desai, N. P.; Drumheller, P. D. Bio-
Technology 1991, 9, 568-572.
(44) One of the referees suggested that the observed differences in cell
spreading and adhesion might be due to the difference in molar mass
between Ki and Kt (molecular weights 37 and 43 kDa, respectively).
We performed experiments using Kt and a similar protein with a
molar mass of 70 kDa (instead of three repeats of the CS5 cell-
Figure 6. Percent of spread HUVEC on engineered proteins at
various time points. Comparison of spread HUVEC on Kt (O), Ki (B),
and Ki* (2). Three independent experiments were performed.
Asterisks indicate p-values e 0.05 for HUVEC on Kt and Ki. Error
bars represent one standard deviation.
Biomacromolecules, Vol. 6, No. 1, 2005Heilshorn et al.
binding domain and elastin cassette; this protein has five repeats) Download full-text
and observed no differences in cellular response. Because a molecular
weight difference of 27 kDa does not alter cellular response to the
Kt sequence design, we do not expect that a molecular weight
difference of 6 kDa contributes to the observed differences in cellular
response to Kt and Ki.
Cell-Binding Domain ContextBiomacromolecules, Vol. 6, No. 1, 2005