Synthetic chemoselective rewiring of cell surfaces: Generation of three-dimensional tissue structures

Article abstract of DOI:10.1021/ja2022569

Proper cell-cell communication through physical contact is crucial for a range of fundamental biological processes including, cell proliferation, migration, differentiation, and apoptosis and for the correct function of organs and other multicellular tissues. The spatial and temporal arrangements of these cellular interactions in vivo are dynamic and lead to higher-order function that is extremely difficult to recapitulate in vitro. The development of three-dimensional (3D), in vitro model systems to investigate these complex, in vivo interconnectivities would generate novel methods to study the biochemical signaling of these processes, as well as provide platforms for tissue engineering technologies. Herein, we develop and employ a strategy to induce specific and stable cell-cell contacts in 3D through chemoselective cell-surface engineering based on liposome delivery and fusion to display bio-orthogonal functional groups from cell membranes. This strategy uses liposome fusion for the delivery of ketone or oxyamine groups to different populations of cells for subsequent cell assembly via oxime ligation. We demonstrate how this method can be used for several applications including, the delivery of reagents to cells for fluorescent labeling and cell-surface engineering, the formation of small, 3D spheroid cell assemblies, and the generation of large and dense, 3D multilayered tissue-like structures for tissue engineering applications.

Full text of DOI:10.1021/ja2022569

ARTICLE
pubs.acs.org/JACS
Synthetic Chemoselective Rewiring of Cell Surfaces: Generation
of Three-Dimensional Tissue Structures
Debjit Dutta, Abigail Pulsipher, Wei Luo, and Muhammad N. Yousaf*
Department of Chemistry and Carolina Center for Genome Science, University of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599-3290, United States
S Supporting Information
b
ABSTRACT: Proper cellÀcell communication through physical
contact is crucial for a range of fundamental biological processes
including, cell proliferation, migration, differentiation, and apoptosis
and for the correct function of organs and other multicellular tissues.
The spatial and temporal arrangements of these cellular interactions
in vivo are dynamic and lead to higher-order function that is
extremely difficult to recapitulate in vitro. The development of
three-dimensional (3D), in vitro model systems to investigate these
complex, in vivo interconnectivities would generate novel methods to
study the biochemical signaling of these processes, as well as provide
platforms for tissue engineering technologies. Herein, we develop
and employ a strategy to induce specific and stable cellÀcell contacts
in 3D through chemoselective cell-surface engineering based on liposome delivery and fusion to display bio-orthogonal functional
groups from cell membranes. This strategy uses liposome fusion for the delivery of ketone or oxyamine groups to different
populations of cells for subsequent cell assembly via oxime ligation. We demonstrate how this method can be used for several
applications including, the delivery of reagents to cells for fluorescent labeling and cell-surface engineering, the formation of small,
3D spheroid cell assemblies, and the generation of large and dense, 3D multilayered tissue-like structures for tissue engineering
applications.
’ INTRODUCTION
a number of lithographic printing techniques14À17 have been
integrated with 3D scaffold designs to produce multitype cellular
arrays9,11,17,18 or 3D cell clusters or spheroids.7,8,13 In a recent
study, 3D aggregates consisting of multiple cell types were
formed within a hydrogel matrix through DNA hybridization
after cell surfaces were engineered with complementary short
oligonucleotides via a metabolic labeling approach.7 However,
for some applications, the presentation of cell-surface DNA may
not be stable for extended time periods in cell culture or in vivo.
Cell-surface engineering methodologies have primarily been
of interest in molecular biology. As such, biosynthetic approaches
have been employed to introduce different functional groups on
cell surfaces. In a pioneering study, an unnatural derivative of N-
acetyl-mannosamine, which bears a ketone group, was converted
to the corresponding sialic acid and metabolically incorporated
onto cell-surface oligosaccharides, resulting in the cell surface
display of ketone groups.19 However, metabolic or genetic
methods may alter many of the biochemical pathways required
for normal cell function and not all cell lines possess this
metabolic machinery. Thus, there is a growing demand for
general tools that can provide simple alternatives to the complex
genetic and biosynthetic methods. Other approaches to cell-surface
Cells that make up tissues and organs exist and communicate
within a complex, three-dimensional (3D) environment. The
spatial orientation and distribution of extracellular matrix (ECM)
components directly influences the manner in which cells
receive, integrate, and respond to a range of input signals.1 As
such, cellular interactions with ECM molecules and/or other
cells have been extensively investigated for fundamental studies
in development, cell motility, differentiation, apoptosis, paracrine
signaling, and applications in tissue engineering.2,3 There has
been tremendous effort toward the design and fabrication of 3D
scaffolds that mimic ECM properties and induce tissue formation
in vitro, utilizing various biomaterials, biodegradable polymers,4
collagen,5 and hydrogels.6,7 Among the major challenges facing
the use of these technologies for tissue engineering are the
abilities to force contact between multiple cell types in 3D to
control the spatial and temporal arrangement of cellular interac-
tions and tailor and mold the biomaterial to recapitulate the 3D,
in vivo environment under laboratory constraints. Without the
use of engineered scaffolds in culture, most cells are unable to
form the necessary higher-order 3D structure required for the
anatomical mimicry of tissue and are limited to random migra-
tion, generating two-dimensional (2D) monolayers. As a result,
several approaches, including the use of dielectrophoretic
forces,8,9 laser-guided writing,10À12 surface manipulation,13 and
Received: March 11, 2011
Published: May 11, 2011
r
2011 American Chemical Society
8704
dx.doi.org/10.1021/ja2022569 J. Am. Chem. Soc. 2011, 133, 8704–8713
|

Journal of the American Chemical Society
ARTICLE
Figure 1. General schematic of liposome preparation for cell-surface tailoring. (A) Dodecanone molecules were incorporated into neutral, egg
palmitoyl-oleoyl phosphatidylcholine (POPC) and cationic, 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP) at a ratio of 5:93:2 to form ketone-
presenting liposomes (1). (B) O-Dodecyloxyamine (A) molecules were incorporated into POPC and DOTAP at a ratio of 5:93:2 to form oxyamine-
presenting liposomes (2). (C) List of liposomes, molecules, and cells used in this study: (1) ketone-functionalized liposomes; (2) oxyamine-
functionalized liposomes; (3) ketone-functionalized Swiss 3T3 albino fbs; (4) rhodamine-functionalized oxyamine (rhod-oxyamine); (5) m-cherry
labeled, oxyamine-functionalized rat2 fbs; (6) oxyamine-functionalized Swiss 3T3 albino fbs; and (7) ketone-functionalized hMSCs.
engineering have also been undertaken to incorporate a func-
tional group into a target biomolecule, such as an endogenous
protein, utilizing a cell’s biosynthetic machinery.20,21 These
strategies aim to produce a site that can then be covalently
modified with its delivered counterpart or probe. However, most
of these protein-based tags are large and bulky and become
problematic when interacting with the other glycans and biomo-
lecules on the cell suface.22,23 Additionally, the perturbation of
cellular physiology with biomolecules at the cell surface may
result in the interference of significant biochemical pathways or
cellular functions.24,25 Thus, a methodology that combines cell-
surface modification, without the use of molecular biology
techniques or biomolecules, and a simple, stable bio-orthogonal
conjugation bottom-up approach that is capable of directing
tissue formation would greatly benefit a range of medical
applications such as wound healing and burn treatment.
Herein, we develop and employ a novel strategy to induce
specific and stable cellÀcell contacts through chemoselective cell-
surface engineering based on liposome delivery and fusion of bio-
orthogonal functional groups to cell membranes.26 Our liposome-
cell membrane methodology was inspired by the work of Wilson
et al.27 and Csiszar et al.28 who reported a noncovalent cell-surface
engineering strategy via cationic graft copolymer adsorption and a
fluorescent cell labeling technique via cationic and aromatic lipid
fusion, respectively. Thus, we incorporated a cationic lipid to
initiate membrane fusion and adsorption to the cell surface. This
strategy enables the presentation of bio-orthogonal ketone and
oxyamine molecules from cell surfaces for subsequent chemose-
lective oxime ligation. No proteins or large biomolecules are used
in this strategy, and therefore, cellular physiology is not perturbed.
We demonstrate how this method may be used in several
applications, including the delivery of reagents to cell surfaces,
formation of 3D spheroid assemblies of cells with controlled
interconnectivity, and patterned multilayered cell tissues. Further-
more, 3D multilayered stem cell and fibroblast (fb) co-cultures
were generated, and differentiation was induced to form tissue-like
structures of adipocytes and fbs. To our knowledge, this is the first
report utilizing tailored liposomes to modify a living cells surface
through membrane fusion for subsequent bio-orthogonal tailoring
to generate 3D tissue-like structures.
’ RESULTS AND DISCUSSION
Vesicle fusion to cell membranes was directed through the
use of a cationic lipid and a molecular recognition pair for chemo-
selective ligation toward the goal of rewiring cell adhesion to
generate 3D multilayers of cells. Vesicles were tailored with
ketone (dodecanone) or oxyamine (O-dodecyloxyamine) mol-
ecules, a neutral lipid, egg palmitoyl-oleoyl phosphatidylcholine
(POPC), and a cationic lipid, 1,2-dioleoyl-3-trimethylammonium-
propane (DOTAP). The resulting two vesicle populations
were then integrated with mammalian cells in culture for applica-
tions in small molecule delivery, cell-surface modification,
and tissue engineering. By employing this membrane tailoring
strategy, the assembly of 3D spheroid clusters and tissue-like
structures were directed after culturing two cell populations func-
tionalized with oxyamine- and ketone-containing groups. Be-
cause this method is general, bio-orthogonal, chemically stable
(oxime bond), and noncytotoxic, patterned multilayered tissue-
like structures of different geometric shapes could also be
fabricated without the use of 3D scaffolds to confine the cell
populations. We also show that this method has promising use in
stem cell transplantation by co-culturing human mesenchymal
stem cells (hMSCs) with fibroblasts (fbs) and inducing adipocyte
differentiation while in a 3D multilayered tissue-like structure.
For membrane fusion studies, dodecanone and O-dodecylox-
yamine (A) were incorporated separately into neutral POPC and
cationic DOTAP lipids at a ratio of 5:93:2 (Figure 1A,B). The
alkane-tethered ketone and oxyamine molecules (A) sponta-
neously insert themselves into the vesicles during syntheses.29
The cationic lipid, DOTAP, was added to enhance membrane
fusion via electrostatic destabilization due to the negatively
charged cell membrane and subsequent lipid component display
at the cell surface. A minimum 2% DOTAP amount was
determined and optimized for liposome-cell fusion initiation by
investigating concentrations that ranged from 0.5% to 5%. The
POPC concentrations were varied accordingly, while maintain-
ing the ketone at 5%. An oxyamine-tethered rhodamine (rhod-
oxyamine, 4) (7 mM in H2O, 100 μL added to 4 mL, 2 h) was
then conjugated to cell surfaces after liposome fusion, and
the fluorescence intensities of each lipid ratio were compared.
8705
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
Figure 2. Liposome-cell fusion characterization by cell patterning, fluorescence microscopy, and fluorescence-activated cell sorting (FACS) analysis. A
schematic and images of cells, rewired to adhere to patterned, synthetic ligands are shown in panels AÀE. Cells were cultured separately with (A) ketone-
(1) or (B) oxyamine- (2) containing liposomes and then seeded on patterned surfaces displaying the complementary oxime conjugation group. (C)
Ketone-functionalized cells (3) adhered to patterned surfaces presenting oxyamine groups, while (D) oxyamine-functionalized cells (6) adhered to
patterned surfaces presenting aldehyde groups. (E) A higher magnification image of the cells in panel D is shown. The cells in panels CÀE were stained
with DAPI (blue, nucleus), phalloidin (red, actin), and Cy-2 (green, anti-vinculin). A general schematic and images for cell-surface tailoring using
liposome fusion and chemoselective oxime chemistry are demonstrated in panels FÀH. (F) Ketone-functionalized lipid vesicles (1) were added and
fused with fbs, displaying ketones from the cell surface (3). Addition of rhod-oxyamine (4) resulted in chemoselective oxime formation and the
fluorescent labeling of fbs. The images displayed represent (G) control fbs, where liposomes not displaying ketones were fused to the membrane,
followed by rhod-oxyamine addition (4, 7 mM in H2O, 2 h), and no fluorescence was observed and (F) fluorescently labeled cells after ketone-
functionalized liposomes (1) were fused to fbs and incubated with rhod-oxyamine (4, 7 mM in H2O, 2 h. The scale bars in panels G and H represent
40 μm. The determination of ketone molecules per cell by flow cytomtery is shown. Fbs were cultured with (1) or without (control) ketone-containing
liposomes for 1, 3, 5, and 7 days. The fbs (3) were then reacted with hydrazide-conjugated biotin, followed by fluorescein-labeled streptavidin. Fbs were
then tested against a standard bead (∼107 beads/mL) with known fluorescein molecule density. Approximately 105 cells were counted for all samples.
Samples were run in triplicate, and the mean fluorescence intensity values are displayed ( RSD. (I) FACS data relating the number of cells counted
(∼105) as a function of fluorescence intensity are shown and labeled as control (without ketone) and days 1, 3, 5, and 7 (with ketone). Fluorescence was
observed for all ketone-displaying populations with a decrease from 1 to 7 days, indicating that the ketone group is carried through cell growth and
division. (J) The number of fluorescein molecules was calculated by comparing the relative mean intensity to a standard bead with known fluorescein
molecule density. The numbers of ketone molecules per cell, as well as the control, are listed, showing a decrease in density on the cell surface over time.
(K) The relationship between the mean fluorescence intensity and the number of days incubated with ketone-containing liposomes is shown. As cells
grow and divide over the course of 7 days, the ketone is carried through.
From 2% to 5% DOTAP, the fluorescence intensities were
almost identical, indicating that 2% DOTAP is sufficient to
initiate fusion. Our general fusion strategy to generate 3D
tissue-like structures is represented in Figure 3A. When cells
are incubated with ketone- (1) or oyamine- (2) containing
liposomes, membrane fusion occurs, resulting in the presentation
of ketone and oyamine groups from cell surfaces (3, 7, and 5, 6,
respectively). When these cell populations are cultured together,
interconnected, 3D tissue-like structures form, mediated through
chemoselective oxime conjugation. These stable tissue structures
can be generated in solution or on a solid support.
ways to manipulate cells and modify biomaterials for a range of
biotechnological and tissue engineering applications. Therefore,
we extended our liposome fusion strategy to rewire cells to
adhere to materials presenting bio-orthogonal, complementary
synthetic ligands (Figure 2AÀE). Two cell populations were
cultured separately with ketone- (1) or oxyamine- (2) containing
liposomes and were then seeded on patterned surfaces displaying
either oxyamine or aldehyde groups, respectively. Employing
previously developed self-assembled monolayer (SAM) technologies,
surfaces were patterned via a microfluidics to present either 10%
oxyamine31 or aldehyde32 and 90% tetra(ethylene)glycol (EG4)
functional groups (Figure 2, panels A and B, respectively). The
remaining, unpatterned region was backfilled to present EG4
terminal groups, which are known to resist nonspecific protein
Rewiring Cell Adhesion. The ability to adhere cells to a
variety of materials through a simple bio-orthogonal approach
that does not rely on proteins or complex ligands enables novel
8706
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
Figure 3. Fluorescent, phase contrast, and scanning electron micrographs (SEM) describing 3D spheroid formation via liposome fusion and
chemoselective cell-surface tailoring. Two fb populations were cultured separately with ketone- (1) or oxyamine- (2) containing liposomes, resulting in
membrane fusion and subsequent tethering of ketones and oxyamines from the cell surface. The oxyamine-tethered rat2 fibroblasts (5) contained a
fluorescent m-cherry nuclear label. The ketone-presenting Swiss albino 3T3 fibroblasts (3) were not fluorescently labeled. (A) Two fibroblast
populations were cultured separately with ketone- (1) or oxyamine- (2) containing liposomes. Because of the presence of a positively charged liposome,
fusion occurred, producing ketone- (3) and oxyamine- (5) tethered cells. Upon mixing these cell populations, clustering and tissue-like formation, based
on chemoselective oxime conjugation, occurred. (B) Control experiments (overlay image) demonstrate no spheroid formation for cells that did not
contain either ketone or oxyamine groups. (C and D) However, when two cell populations displaying ketone (3) and oxyamine (5) recognition groups
are mixed, interconnected spheroid assemblies form (overlay images). (EÀG) Representative SEM images of (E) control cells and (E and F) spheroid
assemblies, as described above, are displayed. For all spheroid assemblies depicted, cell populations were mixed and cultured together for 3 h before
imaging at ∼104 cells/mL.
adsorption and cell attachment.33 Upon seeding cells presenting
ketone (3) or oxyamine (6) groups to the complementary
surfaces, an interfacial oxime reaction occurred and cells adhered,
spread, and proliferated in the patterned regions (Figure 2CÀE).
However, untreated cells and cells cultured with liposomes, not
containing the key functional groups, did not attach to the
surface. This strategy allows for a bottom-up, bio-orthogonal
synthetic approach to rewire how cells adhere to materials and
does not require genetic manipulations of cells.
Cell-Surface Labeling. Using bio-orthogonal approaches to
label specific subcellular organelles and cell-surface membranes is
of great interest to research fields ranging from chemical biology
and tissue engineering to drug delivery. A number of pioneering
approaches involving Staudinger ligation24,34 and click
chemistry20 have been employed for specific covalent and
noncovalent cell-surface engineering applications. In this report,
we use oxime chemistry to tailor and fluorescently label the cell
surface via a novel liposome-cell fusion method. Ketone-functio-
nalized vesicles (1) were delivered to fibroblasts in culture
(Figure 2F). After 4 h of incubation, the liposomes fused to
the cell membranes, causing the cell surface to be decorated with
ketone groups (3). These ketones remain dormant until reaction
with rhod-oxyamine (4, 7 mM in H2O, 2 h) after which the cells
became fluorescent, as shown in the fluorescent image in
Figure 2H. A control was performed where nonfunctionalized
liposomes were added to cells in culture (4 h), followed by
subsequent addition of rhod-oxyamine (4, 7 mM in H2O, 2 h).
The fluorescent image indicated that the reaction and labeling
did not occur without the proper oxime recognition pair
(Figure 2G). Furthermore, under these conditions, we observed
no changes in cell behavior upon liposome fusion to cells. This
observation is a very important feature for future in vivo applica-
tions for cell labeling and targeting via pulse and chase type
methods. Thus, by combining liposome fusion and oxime
chemistry, we were able to tailor the cell surface with either
ketone groups or oxyamine groups, which may act as chemose-
lective cell-surface receptors for a range of small molecules,
ligands, biomolecules, and nanoparticles.
Flow Cytometry. After conducting the above liposome-cell
fusion characterization experiments, we hypothesized that
cell proliferation of the liposome-fused cell tissue carries
unreacted ketone or oxyamine groups on the cell membranes.
This hypothesis was further supported by flow cytometry in
conducting FACS analysis (Figure 2IÀK). Ketone-functio-
nalized cells (1) were reacted with hydrazide-conjugated
biotin (3 mM in CBS, 1 h), followed by fluorescein-tethered
streptavidin (1 mM in CBS, 1 h), and the fluorescence
intensity was compared to a bead of known fluorescein
density and control cells (without ketone groups).35 The
approximate number of ketone and oxyamine molecules
displayed from the cell surface after membrane fusion (12 h),
as well as throughout cell growth and division (3, 5, and 7 days
in culture), was then tabulated. FACS showed that approxi-
mately 9800 ketone or oxyamine molecules were imbedded
into the cell surface after 12 h of liposome incubation
(Figure 2J). Further culturing these cells indicated a decrease
8707
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
Figure 4. General schematic and images of oxime-mediated, 3D tissue-like structure formation with controlled interconnectivity. (A) Ketone- (1) and
oxyamine- (2) containing liposomes were added to two separate fb populations, resulting in membrane fusion and subsequent presentation of the ketone
(3) and oxyamine (6) groups from cell surfaces. Culturing these cells on substrates, alternating cell population seeding layer-by-layer, gave rise to
multilayered, tissue-like cell sheets through stable oxime chemistry. (B) A 3D reconstruction and (C) confocal micrograph showing only a monolayer of
cells after oxyamine-presenting cells (6) were cultured with adhered nonfunctionalized cells. (E) A 3D reconstruction and (F) confocal micrograph of
multiple cell layers after oxyamine-presenting cells (6) were added to substrates presenting ketone-containing cells (3). (D and G) Intact, 3D
multilayered cell sheets can be removed from the surface by gentle agitation as displayed by brightfield and fluorescent images. The insets in panels B and
E show a z-plane cross section that indicates the thickness of the cell layers. Cells were stained with DAPI (blue, nucleus) and phalloidin (red, actin).
in available molecules from 3, 5, and 7 days with 7300, 2100, and
800 molecules per cell, respectively (Figure 2J). Fluores-
cence was still detected, although not as intense, after 7 days
of cell culture, indicating that ketone and oxyamine groups
were carried through cell division. Control cells (without
ketone groups) that were incubated with hydrazide-biotin
and fluorescein-tethered streptavidin demonstrated little to
no fluorescence (Figure 2IÀK).
8708
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
Figure 5. Confocal images representing 2D monolayer and 3D multilayered tissue-like structures of fbs with spatial control. (A) A circular, 2D
monolayer of fbs (control) result after ketone-functionalized fbs (3) and fbs (not functionalized with oxyamines) are patterned on a circular,
microcontact printed region, presenting fibronectin and allowed to grow for 5 days. (BÀD) Fbs, functionalized with ketone groups (3) were seeded onto
microcontact printed regions containing fibronectin and allowed to grow for 2 days. Fbs, functionalized with oxyamine groups (6) were then seeded and
allowed to grow for 2À3 more days. Confocal images demonstrating 3D tissue formation in (B) circle, (C) bar, and (D) square geometries are depicted.
The corresponding z-plane cross sections that indicate the thickness of the cell layers are shown as an inset; scale bars represent 30 μm. Cells were stained
with DAPI (blue, nucleus) and phalloidin (red, actin).
3D Spheroid Assembly. The ability to generate multicellular
connected tissues of multiple cell types in vitro is crucial for
studying the complex interplay of cells in a range of organs in vivo
and for developing strategies for synthetic tissue transplantation.
With varying successes, a number of current strategies to gen-
erate 3D cell connections rely on forcing mixed cell populations
into complex microfabricated wells or vessels. Therefore, we
extended this liposome fusion, oxime-based strategy to generate
3D spheroid assemblies of interconnected cells using two
different cell-type populations (Figure 3). The oxyamine-
presenting rat2 fbs (5) contained a nuclear m-cherry fluorescent
label so that the cell clustering to nonfluorescent ketone-tethered
cells (3) could be easily observed. During a 3-h period of mixed-
culturing (∼104 cells/mL) in solution, cells formed spheroid
structures due to the presence of complementary recognition
groups (Figure 3C,D). Furthermore, when oxyamine-presenting
fbs (5) were cultured with control fbs (cells not functionalized
with ketone groups), spheroid assembly did not occur
(Figure 3B). Studies were also performed to test whether
spheroid size and cell composition could be controlled
(Supporting Information, Figure S1AÀD). Ketone-presenting
hMSCs (7) were co-cultured with oxyamine-functionalized fbs
(6) for 1, 2, 3, and 5 h. After 1 h, clusters comprised only of a few
cells were observed. As the co-culturing duration was increased,
larger spheroid structures were observed. Notably, control
experiments were performed simultaneously to ensure that
spheroid generation was being directed through chemoselective
oxime conjugation. Shown as insets in Figure S1AÀD, tissue
structure formation did not occur without the proper comple-
mentary pair displayed from cell surfaces, regardless of the mixing
duration (1À5 h). Thus, size and composition of 3D cell
assemblies in solution could be controlled, showing great pro-
mise for applications in stem cell transplantation and regenera-
tive medicine.
fbs were reacted with nonfunctionalized cells; fbs spread out on
the surface, migrated, but remained alone (Figure 3E). Notably,
cells were able to form stable, interconnected 3D structures in
solution simply upon mixing two tailored cell populations.
Currently, methods to generate these structures require the
support of a 3D hydrogel matrix and/or assisted assembly
through an external stimulus.5,7À9,13
3D Multilayered Tissues. In addition to forming small, 3D
cell clusters or spheroid structures in solution, this strategy
may be employed to direct larger, dense 3D tissue-like networks
on a surface with geometric control. We used full substrates
(Figure 4), as well as surfaces that were patterned with cell
adhesive and nonadhesive regions to generate multilayered
sheets and patterned tissue structures (Figure 5), respectively.30
Ketone- (1) and oxyamine- (2) tailored liposomes were cultured
with separate fb populations, resulting in membrane fusion and
subsequent presentation of chemoselective sites for oxime con-
jugation from the surface (3 and 6, respectively) (Figure 4A).
Culturing these groups on a solid support (∼105 cells/mL) and
in a layer-by-layer deposition manner gave rise to multilayered,
tissue-like cell sheets, which were characterized by confocal
microscopy, as shown in Figure 4E,F. Fbs naturally only form a
single monolayer once they become contact-inhibited. However,
we have successfully induced fbÀfb clustering though oxime-
mediated, cell-surface engineering based on liposome fusion.
To ensure that oxime chemistry was aiding in the formation
of 3D tissue-like structures, several control experiments were
performed. Cells that did not present ketone or oxyamine
functionality were seeded onto separate surfaces. A second cell
population presenting oxyamine (6) or ketone (3) groups from
the cell surface was added, resulting in the formation of only a 2D
monolayer of cells (Figure 4B,C). Similarly, two different cell
populations that were tethered with oxyamine (6) groups were
mixed together, and only a 2D monolayer was generated after 4
days of culture. The same results were observed after culturing
two different ketone-functionalized cell populations (3) for
4 days. These results further support our hypothesis that
multilayered cell interconnectivity is driven by complementary,
oxime chemistry. We also extended this strategy toward the
generation of 3D multilayered co-cultures with hMSCs and fbs
Spheroid formation was also characterized by scanning elec-
tron microscopy (SEM) (Figure 3EÀG and Figure S2). Cells
functionalized with oxyamine (6) and ketone (3) groups were
able to generate clusters when mixed in solution, as displayed in
Figure 3F,G (Supporting Information, Figure S2). However,
spheroid assemblies were not observed when ketone-presenting
8709
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
Figure 6. General schematic and brightfield images representing oxime-mediated, 3D tissue-like structure formation with hMSC/fb co-cultures and
subsequent induced adipocyte differentiation to generate 3D adipocyte/fb co-culture structures. (A) Ketone-tethered hMSCs (7) were seeded onto a
surface, followed by the addition of oxyamine-functionalized fbs (6). The co-culture was allowed to grow and divide for 3 days at which point, adipogenic
differentiation was induced with the addition of the appropriate media. This resulted in a 3D multilayer of adipocytes and fb. (B) A confluent 2D
monolayer of ketone-presenting hMSCs is represented. (C) A brightfield image displaying a 3D multilayer co-culture of hMSCs (7) and oxyamine-
functionalized fbs (6) is shown. (D) Adipogenic differentiation was induced with media resulting in 3D multilayered adipocyte and fb co-culture
structures, represented by low and (E) high-resolution brightfield images (after 10 days in culture). Adipocytes were stained with Oil Red O (lipid
vacuoles) and Harris Hemotoxylin (nucleus).
(Supporting Information, Figure S3). Ketone-functionalized
hMSCs (7) were first cultured on a substrate (∼105 cells/mL),
and stem cells were allowed to spread out and grow for 2 days.
Oxyamine-presenting fbs (6) were then added (∼105 cells/mL)
and co-cultured for an additional 2 days. As shown by the
confocal images in Figure S3B,C, 3D multilayered cell sheets
(4 layers) were formed. The proper controls were conducted;
without the oxime pair, only a 2D monolayer of stem cells and fbs
was formed (Supporting Information, Figure S3A).
multicellular micropatterns. Microcontact printing30 was used to
produce a variety of patterns and geometries on a gold substrate.
Employing SAM and microfabrication technologies, hexadeca-
nethiol (1 mM in EtOH) was printed on a gold surface. The
surface was then backfilled with EG4 (1 mM in EtOH, 16 h) to
render the remaining regions inert to nonspecific protein absorp-
tion. Fibronetin, a cell-adhesive protein was then added (10 mg/mL
in CBS, 2 h), adhering only to the hydrophobic, patterned
areas. As shown by the confocal image in Figure 5A, only a 2D,
circular cell pattern arises after ketone-presenting fbs (3) were
cultured with fbs, not functionalized with oxyamine molecules.
However, when liposome fusion occurs to display complemen-
tary ketone and oxyamine groups from cell surfaces (3 and 6,
respectively), multilayered 3D cell patterns were formed
(Figure 5BÀD). Circular, bar, and square circular tissue-like
structures are depicted in Figure 5BÀD. The ability to generate
3D tissues with controlled geometry would find great use in
tissue transplantation, in which specifically tailored patches are
required.
3D Stem Cell Co-cultures with Induced Adipocyte Differ-
entiation. We explored the general use of this liposome fusion
method, delivered ketone and oxyamine groups to different cell
lines, and demonstrated that 3D spheroid and multilayer can be
generated using co-cultures of hMSCs and fbs (Supporting
Information, Figures S1 and S3, respectively). We next extended
our methodology toward stem cell differentiation to determine
whether 3D multilayered co-cultures could be induced to gen-
erate tissues of differentiated hMSCs and fbs. As shown in
Figure 6A, ketone-functionalized hMSCs (7) were first cultured
on a substrate for 3 days, producing a 2D monolayer of cells
(Figure 6B). Oxyamine-tethered fbs (6) were then co-cultured
with the hMSCs, and the cells were allowed to grow and
proliferate for 2 days (Figure 6C). Adipogenic induction media
was then added, the 3D multilayered coculture was stained for
nuclei (purple) and lipid vacuoles (red), which are characteristic
3D Tissue Release and Cell Viability. During multilayer
culture, it was possible to control the release of the tissues from
the surface with gentle agitation (Figure 4D,G). The ability to
release tissue after surface-supported growth in vitro shows great
potential for applications in tissue engineering and cellular
transplantation. Cell viability was also tested for 3D spheroid
and multilayered structures of fbs and hMSC/fb co-cultures
using the trypan blue assay (Supporting Information, Figure S4).35
After spheroid (1, 2, 3, and 5 h of mixing in solution) and
multilayer (3, 5, and 7 days on a surface) formation, cells were
incubated with trypan blue (0.4%, 2 min). Viability was 100% for
all cells in the spheroid assemblies (1À5 h) and multilayer
structure at day 3. After 5 and 7 days of multilayer genera-
tion, cells showed an approximate viability of 91% and 84%,
respectively. The blue intensity (fluorescence false colored for
enhanced visualization) was compared to a control cell popula-
tion by linescan analysis (Supporting Information, Figure S4).
The control cells were cultured for 7 days to generate 3D
multilayers and were then fixed. Trypan blue was allowed to
react for 2 min, followed by imaging and quantification. Overall,
the viability of cells in conducting membrane fusion to generate
3D tissue-like structures in solution and on a solid support is
high. Therefore, this method may be very useful for applications
in tissue engineering and stem cell transplantation.
3D Tissue Patches with Geometrical Control. We further
demonstrated spatial control by generating a number of 3D
8710
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
of adipocytes (fat cells). The phase contrast images in Figure 6D,
E demonstrate the successful generation of tissue-like structures,
comprising induced adipocytes and fbs. The ability to co-culture
stem cells with many other cell types and induce differentiation
shows great promise in the field of regenerative medicine and
stem cell transplantation.
University of North Carolina (UNC). Rat2 fb transfected with m-cherry
were obtained from the Bear Lab (UNC Chapel Hill, NC). Human
Mesenchymal stem cells (hMSCs) were purchased from Lonza (Basel,
Switzerland).
Syntheses. Tetra(ethylene glycol)-terminated alkanethiol (EG4)33
and rhodamine-oxyamine (rhod-oxyamine, 4)43 were synthesized as
previously reported. The synthesis44 and proton nuclear magnetic
resonance (1H NMR) characterization of O-dodecyloxyamine (A) is
presented in Supporting Information (Scheme S1).
’ CONCLUSION
In this study, we developed a simple liposome delivery and
fusion method to display ketone or oxyamine functional groups
from cell surfaces for applications in bio-orthogonal ligand
conjugation, rewiring cell adhesion, and the generation of stable,
3D spheroid assemblies and multilayered tissue-like structures.
This strategy may have diverse applications in the field of tissue
engineering and regenerative medicine, from growing biocom-
patible tissues and organs in vitro to their cellular transplantation
in vivo.36,37 For example, assembled tissue patches with geome-
trically defined shape can be grown in culture and transplanted or
grafted to specific locations.38 Furthermore, this strategy may
allow for time-lapse observation of cell movement in vivo by using
a pulse (delivery of labeled cells via liposome fusion), followed by
a chase (bio-orthogonal reagent to target only the labeled cells).
When applied in vivo, this method may allow for the monitoring
of many spatiotemporal developmental events and tumor me-
tastasis. Since the liposome fusion method is general, many other
types of chemistries in a single liposome can be delivered to the
membrane surface simultaneously. For example, liposomes con-
taining ketones, alkynes, dienes, azides, hydrazides, or dieno-
philes in varying combinations may be delivered to a cell surface
for iterative or simultaneous postfunctionalization via bio-orthogonal
ligation reactions.
We have also been able to perform liposome fusion to the
same cells several times, which may be important to tailor the
membrane with multiple groups or to increase the concentration
of a particular surface functional group. By rewiring cell adhesion,
a number of materials, surfaces, nanoparticles, and biomedical
devices for various biotechnological applications may be deco-
rated with cells. Since no biomolecules are used with this strategy,
no long-term stability and degradation issues in complex cell
culture media or in vivo will affect cell targeting or cell assembly.
Combining this strategy with polymer scaffolds, 3D tissues and
organs may be generated for paracrine signaling studies, tissue
replacement therapies, stem cell plasticity studies, or as a model
platform for various high-throughput screening studies.39À42
Finally, integrating this strategy with traditional liposome delivery,
where the interior of the liposome contains small molecules or
nanoparticle cargoes, a multiplex system with the delivery of
reagents to the interior of cells and simultaneous labeling of the
exterior of the cells may be possible for entirely new diagnostic
and biomedical applications.
Liposome Preparation. To generate keto- (1) and oxy-LUVs (2),
dodecanone (55 μL, 10 mM solution in CHCl3 at 5 mol %) and alkane-
tethered oxyamine (60 μL, 10 mM solution in CHCl3 at 5 mol %) were
dissolved with egg-POPC (424 μL, 10 mg/mL in CHCl3 at 93 mol %)
and 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP, 10 μL, 10 mg/mL
in CHCl3 at 2 mol %) in chloroform followed by concentration under
high vacuum for 4 h. The dried lipid samples were then reconstituted and
brought to a final volume of 3 mL in PBS buffer, pH 7.4. The contents of
the vial were warmed to 50 °C and sonicated for 20 min, in a tip
sonicator, until the solution became clear, and LUVs containing ketone
or oxyamine groups were formed.
Rewiring Cell Adhesion. Briefly, mixed self-assembled mono-
layers (SAMs) presenting oxyamine (OA) or aldehyde (Ald) and EG4
groups were patterned using microfluidic lithography31 and microfluidic
oxidation,32 respectively. The OA and Ald group percentages were
minimal (1/9 OA/EG4, 1 mM in EtOH total) to ensure resistance to
nonspecific protein and cell adhesion. Swiss 3T3 albino mouse fbs were
separately cultured with keto- (1) or oxy- (2) LUVs, as previously
described, and were then seeded (∼104 cells/mL) to surfaces patterned
with OA or Ald, respectively. Over the course of 4 days, the cells adhered,
spread, and proliferated, filling out the patterned regions of the surface
due to the interfacial oxime reaction. The cells that were cultured with
liposomes not containing the key functional groups did not attach to the
patterned surfaces. The substrates were stained (see staining procedure
below) and imaged by fluorescence microscopy. An exposure time of
400 and 1200 ms was used to image nuclei and actin, respectively.
Cell-Surface Reaction to Ketone-Presenting Cells. Fbs were
incubated with the keto-LUVs (1) for 4 h and washed with PBS (4 Â
25 mL), followed by addition of rhod-oxyamine (4, 7 mM in H2O,
100 μL added to 4 mL) dye and incubation for 2 h. The cells were then
washed with PBS (4 Â 4 mL) and removed with a solution of 0.05%
trypsin and 0.53 mM EDTA and resuspended in serum-free medium
(∼104 cells/mL). The cells were then seeded to a fibronectin-coated
surface for 2 h. After 2 h, serum-containing media was added for cell
growth and imaged after 3 days. Phase contrast and fluorescent imaging
was performed and processed using a Nikon TE2000-E inverted
microscope and Metamorph software, respectively.
Flow Cytometry. Fluorescence-activated cell sorting (FACS)
analysis was performed to quantify the approximate number of ketone
and oxyamine groups at the cell surface after membrane fusion.
Liposomes (1) were prepared as described above and were delivered
to fbs in culture (3 mM in tris buffer, 400 μL added to 4 mL, 12 h). A
time course assay was also conducted using FACS to determine whether
the chemistry was being carried on after cell growth and division. Fbs (3)
were reacted with hydrazine-conjugated biotin (3 mM in CBS, 1 mL
added to 4 mL CBS in cell culture, 1 h) after culture with ketone-
containing liposomes (1) for 1, 3, 5, and 7 days, followed by fluorescein-
conjugated streptavidin (1 mM in CBS, 0.5 mL added to 4 mL CBS in
cell culture, 1 h). A control cell population (not displaying ketone
groups) was only incubated with biotin-hydrazide and streptavidin-
fluorescein for 1 h each, under the same conditions. The cells were
then centrifuged (5 min, 1000 rpm), resuspended in RPMI (without
phenol red), centrifuged (5 min, 1000 rpm), and resuspended in RPMI
(∼107 cells/2 mL). Fluorescence measurements were calibrated using
RCP-5-30 beads (∼107 beads/mL, Spherotech, Inc., Lake Forest, IL) of
’ EXPERIMENTAL SECTION
All chemical reagents were of analytical grade and used without
further purification. Lipids egg palmitoyl-oleoyl phosphatidylcholine
(POPC) and 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP)
were purchased from Avanti Polar Lipids (Alabaster, AL). Antibodies
and fluorescent dyes were obtained from Invitrogen (Carlsbad, CA).
Trypan blue viability dye was obtained from Hyclone (Fisher Scientific,
Pittsburgh, PA), and all other chemicals, including dodecanone, were
obtained from Sigma-Aldrich or Fisher. Swiss 3T3 albino mouse
fibroblasts (fb) were obtained from the Tissue Culture Facility at the
8711
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
known fluorescein equivalent molecule density. Fluorescent intensities
based on number of cells counted were compared to the standard bead
and control cells lacking fluorescent molecule conjugation and approximate
numbers of fluorescent compound bound to the surface were calculated. Flow
cytometry was performed using a Dako CyAn ADP (Beckman-Coulter, Brea,
CA), and data was analyzed with Summit 4.3 software.
3D Spheroid Generation. Keto- (1) and oxy-LUVs (2) were
added to two separate fb populations in culture (3 mM in tris buffer,
400 μL added to 4 mL, 12 h), resulting in fusion and display of ketones
and oxyamines from the cell surface. Oxyamine-presenting Rat2 fb (5)
contained an m-cherry label (nucleus) for enhanced visualization, while
the ketone-presenting Swiss 3T3 albino mouse fb (3) contained no
fluorescent label. These two cell populations were then trypsinized and
mixed together (∼204 cells/mL, 4 mL total) in serum containing (10%
CBS, pH of 7.4) media in a 10 mL-flask and incubated at 37 °C and 5%
CO2 for 3 h. After mixing, the cells were seeded on a glass surface (∼204
cells/mL, 1 mL) and visualized under a Nikon TE2000-E inverted
microscope or by scanning electron microscopy. Image acquisition and
processing was performed using Metamorph software. An exposure time
of 75 ms was used to image all spheroids.
Scanning Electron Microscopy (SEM) of 3D Spheroids.
Spheroids were assembled in solution (reaction for 3 h as described
above), delivered to a glass slide (∼204 cells/mL, 1 mL, 0.8 Â 0.8 cm2),
and then fixed with 10% formalin in PBS for 15 min. The substrate was
then washed with water (15 min), and cells were then dehydrated
stepwise in 30, 50, 70, 90, and 100% ethanolic solutions for 15 min each.
After critical point drying and sputtering 2 nm of gold, the sample
was ready for imaging using a Hitachi S-4700 field emission scanning
electron microscope (Hitachi High Technologies America, Inc.,
Schaumburg, IL).
Fibroblast Culture. Swiss 3T3 albino mouse fb and Rat2 fb were
cultured in Dulbecco’s Modified Eagle Medium (Gibco) containing 10% calf
bovine serum (CBS) and 1% penicillin/streptomycin at 37 °C in 5% CO2.
Human Mesenchymal Stem (hMSC) Cell Culture. hMSCs and
basic, growth, and differentiation media were obtained from Lonza (Basel,
Switzerland). hMSCs were cultured in Dulbecco’s Modified Eagle Medium
(Gibco) containing 10% fetal bovine serum (FBS) and 1% penicillin/
streptomycin at 37 °C in 5% CO2. Culturing with induction medium as
described in the Lonza protocol induced Adipogenic differentiation.
Immunohistochemistry. After the growth of 3D tissue-like structures
and co-culture with Swiss 3T3 albino mouse fb, surfaces were fixed with
formaldehyde (4% in PBS, 30 min). Substrates were then immersed in a
solution containing water and 60% isopropyl alcohol (3À5 min),
followed by staining with Oil Red O (5 min) and Harris Hemotoxylin
(1 min) (6,7). Substrates were visualized by phase contrast microscopy
using a Nikon TE2000-E inverted microscope. Image acquisition and
processing was performed using Metamorph software. An exposure time
of 75 ms was used to image all HMSCs.
mounting medium (Dako, Carpinteria, CA), which enhances the
visualization of cells when viewed under a fluorescent microscope on
a glass coverslip. An exposure time of 400 and 1200 ms was used to
image nuclei and actin, respectively.
Confocal Microscopy. Cell clusters and tissue formation were
visualized with a Nikon Eclipse TE2000-E inverted microscope (Nikon
USA, Inc., Melville, NY). The data were recorded using Leica software
and a spectral confocal microscope (LeicaMicrosystems, Bannockburn, IL).
An average of 84 image scans was used to generate the 3D reconstructions
with Volocity software.
3D Co-culture Spheroid and Multilayer Generation.
Spheroids. Keto- (1) and oxy-LUVs (2) were generated as previously
reported, added to hMSCs and fbs (3 mM in tris buffer, 400 μL added to
4 mL, 12 h), respectively, and were cultured, resulting in fusion and
display of ketones and oxyamines from the cell surface. These two cell
populations were then trypsinized and mixed together in serum contain-
ing (10% FBS, pH of 7.4) media in a 10 mL flask and incubated at 37 °C
and 5% CO2 for 1, 2, 3, and 5 h. After mixing for the allotted time, cells
were seeded onto a glass surface and visualized under a Nikon TE2000-E
inverted microscope under the brightfield setting (75 ms exposure
time). Controls were also performed where hMSCs displaying ketone
groups were co-cultured with fbs (not displaying oxyamine groups) for
each of the corresponding time points, 1, 2, 3, and 5 h, seeded onto glass,
and imaged under the brightfield setting (75 ms). Image acquisition and
processing was performed using Metamorph software.
Multilayers. Keto- (1) and oxy-LUVs (2) were added to hMSC and
fbs (3 mM in tris buffer, 400 μL added to 4 mL, 12 h), respectively, and
were cultured, resulting in fusion and display of ketones and oxyamines
from the cell surface. hMSCs (7) displaying ketone groups were
trypsinized and cultured on glass slides (105 cells/mL) and allowed to
grow for 2 days. Fbs presenting oxyamines (6) were then trypsinized and
added (105 cells/mL) to the hMSCs. These cells were co-cultured in
media (10% FCS) for 3, 5, and 7 days, resulting in the formation of 3D
multilayered, tissue-like structures of hMSCs and fbs.
Cell Viability Assay. Cell viability of 3D spheroid and multilayered
tissue-like structures was assessed using a trypan blue viability assay
(Hyclone, Fisher Scientific, Pittsburgh, PA). Fb spheroid and multilayer
structures were prepared as previously described. A solution of 0.4%
trypan blue in PBS was made and diluted in CBS (1:1) containing the
spheroids (1, 3, and 5 h after mixing, 204 cells/mL) in solution and
multilayer cell sheets (3, 5, and 7 days after a second fb population was
added, 105 cells/mL) on a glass slide. Trypan blue was allowed to react
with the cells for 2 min, at which time spheroids and surfaces were
imaged and false colored with blue for enhanced visualization using a
Nikon TE2000-E inverted microscope. As a control, cells were cultured
for 7 days to generate a multilayer and were then fixed as mentioned
above. Trypan blue was allowed to react for 2 min, and cells were imaged.
For phase contrast and fluorescent imaging, exposure times of 75 and
400 ms were used, respectively.
Directed 3D Tissue-Like Multilayers. Ketone-functionalized fbs
(3) were seeded (∼104 cells/mL) to microcontact printed patterned
(1 mM hexadecanethiol in EtOH, printed on gold 5 s, backfilled with
1 mM EG4 in EtOH, 16 h) surfaces presenting fibronectin (10 mg/mL,
2 h) for 2 h. The cells were allowed to grow for 3 days (37 °C in 5%
CO2).29 Oxyamine-functionalized fbs (6) (∼104 cells/mL) were then
seeded to surfaces for 2 h, followed by addition of serum-containing
(10% CBS) media to promote cell growth. The cells were cultured for
3 more days before imaging. After generation, substrates were fixed,
stained, and imaged by confocal microscopy as described below.
Cell Staining for Imaging. Cells were fixed with formaldehyde
(4% in PBS) and permeated (PBS containing 0.1% Triton X-100). A
fluorescent dye mixture, containing phalloidin-TRITC (actin) and
DAPI (nucleus) was then made in PBS containing 5% normal goat
serum and 0.1% Triton X -100. Cells were incubated with the dye
solution for 2 h. The substrates were then secured in fluorescence
’ ASSOCIATED CONTENT
S
Supporting Information. Additional synthetic details
b
1
and H NMR characterization of O-dodecyloxyamine, phase
contrast and SEM images demonstrating control over composi-
tion and size of 3D spheroid formation, confocal images display-
ing 3D co-culture tissues with stem cells and fibroblast, and cell
viability data. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
mnyousaf@email.unc.edu
8712
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713

Journal of the American Chemical Society
ARTICLE
’ ACKNOWLEDGMENT
(30) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.;
Whitesides, G. M. Chem. Rev. 2000, 105, 1103–1170.
(31) Lamb, B. M.; Barrett, D. G.; Westcott, N. P.; Yousaf, M. N.
Langmuir 2008, 24, 8885–8889.
(32) Westcott, N. P.; Pulsipher, A.; Lamb, B. M.; Yousaf, M. N.
Langmuir 2008, 24, 9237–9240.
(33) Pale-Grosdemange, C.; Simons, E. E.; Prime, K. L.; Whitesides,
G. M. J. Am. Chem. Soc. 1991, 113, 12–20.
(34) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007–2010.
(35) Hsiao, S. C.; Shum, B. J.; Onoe, H.; Douglas, E. S.; Gartner, Z.;
Mathies, R. A.; Bertozzi, C. R.; Francis, M. B. Langmuir 2009,
25, 6985–6991.
The authors thank Dr. Jim Bear for his generous donation of
m-cherry labeled cells, Dr. Nancy Thompson and her lab group
for use of equipment and expertise in liposome chemistry, Lisa
Bixby for conducting FACS, and Dr. Michael Chua for aid with
confocal imaging. This work was supported by the Carolina
Center for Cancer Nanotechnology Excellence, the Burroughs
Wellcome Foundation (Interface Career Award), and the National
Science Foundation.
(36) Mueller-Klieser, W. Am. J. Physiol. 1997, 273, C1109–C1123.
(37) Khademhosseini, A.; Vacanti, J.; Langer, R. Sci. Am. 2009,
300, 64–71.
’ REFERENCES
(1) Nelson, C. M.; Bissel, M. J. Annu. Rev. Cell Dev. Biol. 2006, 22,
287–309.
(38) Badylak, S. F.; Nareem, R. M. Proc. Natl. Acad. Sci. U.S.A. 2010,
(2) Meshel, A. S.; Wei, Q.; Adelstein, R. S.; Sheetz, M. P. Nat. Cell
107, 3285–3286.
Biol. 2005, 7, 157–164.
(3) Isenberg, B. C.; Williams, C.; Tranquillo, R. T. Annu. Biomed.
Eng. 2006, 34, 971–985.
(4) Hollister, S. J. Nat. Mater. 2005, 4, 518–524.
(5) Gillette, B. M.; Jensen, J. A.; Tang, B.; Yang, G. J.; Bazargan-Lari,
A.; Zhong, M.; Sia, S. K. Nat. Mater. 2008, 7, 636–640.
(6) Tanaka, H.; Murphy, C. L.; Murphy, C.; Kimura, M.; Kawai, S.;
Polak, J. M. J. Cell Biochem. 2004, 93, 454–462.
(7) Gartner, Z. J.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2009,
106, 4606–4610.
(39) Khetani, S. R.; Bhatia, S. N. Nat. Biotechnol. 2008, 26, 120–126.
(40) Zeisberg; Yang, C.; Martino, M.; Duncun, M. B.; Rieder, F.;
Tanjore, H.; Kalluri, R. J. Biol. Chem. 2007, 282, 23337–23347.
(41) Bhatia, S. N.; Balis, U.; Yarmush, M. L.; Toner, M. Biotechnol.
Prog. 1998, 14, 378–387.
(42) Kunz-Schughart, L. A.; Freyer, J. P.; Hofstaedter, F.; Ebner, R.
J. Biomol. Screening 2004, 9, 273–285.
(43) Chan, E. W. L.; Yousaf, M. N. J. Am. Chem. Soc. 2006,
128, 15542–15546.
(44) Park, S.; Yousaf, M. N. Langmuir 2008, 24, 6201–6207.
(8) Albrecht, D. R.; Underhill, G. H.; Wassermann, T. B.; Sah, R. L.;
Bhatia, S. N. Nat. Methods 2006, 3, 369–375.
(9) Gray, D. S.; Tan, J. L.; Voldman, J.; Chen, C. S. Biosens.
Bioelectron. 2004, 19, 1765–1774.
(10) Odde, D. J; Renn, M. J. Biotechnol. Bioeng. 2000, 67, 312–318.
(11) Nahmias, Y.; Odde, D. J. Nat. Protocol 2006, 1, 2288–229626.
(12) Barron, J. A.; Krizman, D. B.; Ringeisen, B. R. Annu. Biomed.
Eng. 2005, 33, 121–130.
(13) Inaba, R.; Khademhosseini, A.; Suzuki, H.; Fukuda, J. Bioma-
terials 2009, 30, 3573–3577.
(14) Ringeisen, B. R.; Othon, C. M.; Barron, J. A.; Young, D.; Spargo,
B. J. Biotechnol. 2006, 1, 930–948.
(15) Chiou, P. Y.; Ohta, A. T.; Wu, M. C. Nature 2005, 436,
370–372.
(16) Falconnet, D.; Csucs, G.; Grandin, H. M.; Textor, M. Bioma-
terials 2006, 27, 3044–3063.
(17) Khademhosseini, A.; Langer, R.; Borenstein, J.; Vacanti, J. P.
Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2480–2487.
(18) Luo, W.; Chan, E. W. L.; Yousaf, M. N. J. Am. Chem. Soc. 2010,
132, 2614–2621.
(19) Mahal, L. K.; Yarema, K. J.; Bertozi, C. R. Science 1997, 276,
1125–1128.
(20) Prescher, J. A.; Bertozzi, C. R. Nat. Chem. Biol. 2005, 1, 13–21.
(21) Chen, I.; Howarth, M.; Lin, W.; Ting, A. Y. Nat. Methods 2005,
2, 99–104.
(22) Keppler, A.; Pick, H.; Arrivoli, C.; Vogel, H.; Jonhsson, K. Proc.
Natl. Acad. Sci. U.S.A. 2004, 101, 9955–9959.
(23) Miller, L. W.; Sable, J.; Goelet, P.; Sheetz, M. P.; Cornish, V. W.
Angew. Chem. Int. Ed. 2004, 43, 1672–1675.
(24) Kellam, B.; De Bank, P. A.; Shakesheff, K. M. Chem. Soc. Rev.
2003, 32, 327–337.
(25) Rabuka, D.; Forstner, M. B.; Grovers, J. T.; Bertozzi, C. R. J. Am.
Chem. Soc. 2008, 130, 5947–5953.
(26) Sarkar, D.; Vemula, P. K.; Zhao, W.; Gupta, A.; Karnik, R.; Karp,
J. M. Biomaterials 2010, 31, 5266–5274.
(27) Wilson, J. T.; Krishnamurthy, V. R.; Cui, W.; Qu, Z.; Chaikof,
E. L. J. Am. Chem. Soc. 2009, 131, 18228–18229.
(28) Csiszar, A.; Hersch, N.; Dieluweit, S.; Biehl, R.; Merkel, R.;
Hoffmann, B. Bioconjugate Chem. 2010, 21, 537–543.
(29) Dutta, D.; Pulsipher, A.; Yousaf, M. N. Langmuir 2010, 26,
9835–9841.
8713
dx.doi.org/10.1021/ja2022569 |J. Am. Chem. Soc. 2011, 133, 8704–8713