Regulating MCP-1 diffusion in affinity hydrogels for enhancing immuno-isolation

Article abstract of DOI:10.1016/j.jconrel.2009.11.022

Delivering cells using semi-permeable hydrogels is becoming an increasingly important direction in cell based therapies and regenerative medicine applications. Synthetic hydrogels have been functionalized with bioactive motifs to render otherwise inert po

Full text of DOI:10.1016/j.jconrel.2009.11.022

Journal of Controlled Release 142 (2010) 384–391
Contents lists available at ScienceDirect
Journal of Controlled Release
journal homepage: www.elsevier.com/locate/jconrel
Regulating MCP-1 diffusion in affinity hydrogels for enhancing immuno-isolation
Chien-Chi Lin a,b, Patrick D. Boyer a, Alex A. Aimetti a, Kristi S. Anseth a,b,
⁎
a
Department of Chemical and Biological Engineering, University of Colorado, 424 UCB, Boulder, CO 80309, USA
b
Howard Hughes Medical Institute, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Delivering cells using semi-permeable hydrogels is becoming an increasingly important direction in cell based
therapies and regenerative medicine applications. Synthetic hydrogels have been functionalized with
bioactive motifs to render otherwise inert polymer networks responsive. However, little effort has been
focused on creating immuno-isolating materials capable of retarding the transport of small antigenic
molecules secreted from the cells delivered with the synthetic carriers. Toward the goal of developing a
complete immuno-isolation polymeric barrier, affinity peptide-functionalized PEG hydrogels were developed
with the ability to sequester monocyte chemotactic protein 1 (MCP-1), a chemokine known to induce the
chemotaxis of monocytes, dendritic cells, and memory T-cells. Affinity peptides capable of sequestering MCP-1
were identified from CCR2 (a G protein-coupled receptor for MCP-1) and incorporated within PEG hydrogels
via a thiol-acrylate photopolymerization. The release of encapsulated recombinant MCP-1 from PEG hydrogels
is readily tuned by: (1) incorporating affinity peptides within the network; and/or (2) altering the spacer
distance between the affinity peptide and the crosslinking site. Furthermore, when pancreatic β-cells were
encapsulated within these novel peptide-functionalized hydrogels, the release of cell-secreted MCP-1 was
significantly reduced, demonstrating the potential of this new gel formulation to reduce the host innate
immune response to transplanted cells by decreasing the recruitment and activation of host monocytes and
other immune cells.
Received 23 September 2009
Accepted 22 November 2009
Available online 29 November 2009
Keywords:
Chemokine
Affinity hydrogels
Photopolymerization
Tissue engineering
Diabetes
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
hydrogels, however, requires consideration of not only antibodies and
immune cells, but also small molecular weight antigenic, chemotactic,
The long-term survival of transplanted islets is key to the
successful reversal of type 1 diabetes [1]. To overcome graft rejection
induced by host T-cells, the administration of immunosuppressants
(e.g., sirolimus/rampamycin and tacrolimus) is inevitable, albeit these
drugs are toxic to islet cells and may cause unfavorable systemic
complications [2–6]. Recently, poly(ethylene glycol) (PEG) hydrogels
are in development for many cell delivery and regenerative medicine
applications [7,8], including the encapsulation of pancreatic β-cells for
cell transplantation to treat Type 1 diabetes [9,10]. The needs for
immunosuppressants following islet transplantation may be alleviat-
ed when the islet grafts are encapsulated and protected by a physical
hydrogel barrier. One of the major characteristics of PEG hydrogels is
their high water content that leads to preferential high diffusivity for
small molecular weight nutrients and metabolic products, such as
glucose and insulin [11,12]. The crosslinked PEG polymer network
provides a selective “immuno-isolation barrier” to protect the
encapsulated cells from immune destruction initiated by host
antibodies and immune cells [1]. The design of immuno-isolating
and cytotoxic immune-mediators (e.g., reactive oxygen species and
pro-inflammatory cytokines) [13,14]. While preventing direct cell–
cell contact and antibody diffusion is relatively easy with size
exclusive barriers, such as PEG hydrogels, it remains a challenging
task to preclude the diffusion of small molecular weight soluble
immune-mediators in permissive and bio-inert hydrogels [1,11].
The concentration of a variety of cytokines and chemokines is
known to be elevated at the site of injury, infection, or transplantation
[15–17]. The increased cytokine/chemokine concentration plays an
important role in the dynamic inflammatory response. For example,
donor antigens (e.g., chemokines) secreted by transplanted cells are
responsible for recruiting host inflammatory and immune cells [18–
20]. In this regard, unmodified PEG hydrogels fail to prevent the
leakage of small molecular weight chemokines secreted from cells
encapsulated in the gel capsule. These chemokines, which readily
diffuse out of the gel, can recruit additional immune cells that sub-
sequently secrete excess cytotoxic cytokines that can damage the
encapsulated cells [1,11,13]. Thus, implementing mechanisms to
regulate the transport and escape of these small molecular weight
antigenic/cytotoxic molecules produced by transplanted cells would
be highly advantageous, and fabrication of new cell carriers that
enable this suppression of transport of undesired molecules while
maintaining cell viability would be highly desirable.
⁎
Corresponding author. Dept. of Chemical and Biological Engineering, University of
Colorado, Boulder, CO 80309, USA Tel.: +1 (303) 492 3147.
E-mail address: Kristi.Anseth@Colorado.edu (K.S. Anseth).
0168-3659/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2009.11.022

C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
385
Chemokines are small proteins (∼8 to 13 kDa) known to attract
2.2. In vitro cellular activities of monocytes in the presence of
recombinant MCP-1
inflammatory and immune cells. For example, the major function of
monocyte chemotactic protein 1 (MCP-1 or CCL2), a member of the CC
chemokine superfamily, is to recruit monocytes [21–23], dendritic
cells [24,25], and memory T-cells [26] to the site of injury, infection,
and transplantation. These immune cells, when recruited and
activated, secrete a variety of cytotoxic cytokines that subsequently
cause damage to the allogenic or xenogenic cells, even when these
cells are encapsulated in hydrogels. Chemokines are important
immune-regulators in a variety of autoimmune and inflammatory
diseases [23], including type 1 diabetes [27–29]. Constitutive
expression of chemokines, such as MCP-1, from pancreatic islets has
been demonstrated and is known to play a critical role in the
induction of autoimmune response that leads to the onset of type 1
diabetes [30]. The secretion of MCP-1 from pancreatic islets, as well as
insulin-producing INS-1E and RINm5F β-cells, is enhanced following
the stimulation with angiotensin-II (Ang-II) [31,32]. It has also been
shown that the expression and secretion of MCP-1 from murine
pancreatic β-cell line, MIN6, is up-regulated upon stimulation with
pro-inflammatory cytokines [33]. Clinical studies have also revealed
elevated serum MCP-1 concentration in type 1 diabetic patients [34],
as well as reduced transplanted graft survival associated with
increased circulating donor MCP-1 concentrations [35]. Given the
importance of MCP-1 in type 1 diabetes, we hypothesize that it would
be beneficial to design functional gel barriers that could eliminate or
minimize the release of donor MCP-1 from transplanted allogenic/
xenogenic β-cells.
A monocyte/macrophage cell line, J774a.1 (ATCC: TIB-67TM), was
used to test the dose-dependent response of monocytes to MCP-1.
(1) Chemotaxis: J774a.1 cells were seeded in a 24-well trans-well
device (pore size: 7 μm) and placed in the wells containing different
concentrations of recombinant murine MCP-1 in high glucose DMEM
(with 10% FBS, 1% penicillin–streptomycin, and 0.5 μg/mL fungizone)
for 16 h. Migrating cells that adhered to the bottom of the trans-well
devices were fixed with 4% formalin for 15 min, washed with PBS, and
stained with crystal violet (0.2 wt% in 20% methanol/PBS) for 30min.
Excess crystal violet was removed by extensively washing with PBS.
Stained cells were imaged with a phase contrast microscope and
counted in 5 different fields per sample (n=4). The number of
migrating cells without MCP-1 treatment was set as 100%. (2)
Proliferation: J774a.1 cells were seeded in 24-well plates at 100,000
cells/well overnight. Cells were incubated in serum-free DMEM
overnight prior to MCP-1 treatment. Various concentrations of murine
MCP-1 were added to cells that were further incubated for 24 h. The
relative cell proliferation was determined by AlamarBlue® reagent
according to the manufacturer's protocol. (3) IL-1β secretion: J774a.1
cells were seeded in a 24-well plate at 100,000 cells/well for
overnight, followed by treatment of MCP-1 in serum-free DMEM.
After 24 h of culture, media were collected and analyzed with murine
IL-1β ELISA kit to determine the amount of IL-1β secretion.
Significant efforts have been dedicated to the design of bioactive
materials for regulating local inflammation [13,14,36]. Previously,
strategies were developed based on affinity protein–peptide binding
to control growth factors release [37] from PEG-based hydrogels, as
well as to antagonize a pro-inflammatory cytokine, tumor necrosis
factor α (TNFα), within the local gel environment [13]. In the context
of regulating the inflammatory response, affinity peptides were
conjugated in photopolymerized PEG hydrogels, creating cytokine-
antagonizing hydrogels that sequester and antagonize infiltrated
TNFα, thereby decreasing apoptosis of the encapsulated cells,
including pancreatic islets [13]. Here, we present the identification
and characterization of affinity peptides capable of binding to murine
MCP-1 and describe the fabrication of affinity PEG–peptide hydrogels
to reduce the diffusion of chemokine MCP-1 from highly permissive
PEG hydrogels. Specifically, several synthetic peptides derived from
extracellular loops of murine CCR2, a 7 trans-membrane G protein-
coupled receptor (GPCR) for murine MCP-1, were synthesized and co-
polymerized with PEG-diacrylate (PEGDA), via a thiol-acrylate
photopolymerization [10,37,38], to form peptide-functionalized PEG
hydrogels. These affinity PEG–peptide hydrogels were found to
sequester MCP-1 secreted by the encapsulated MIN6 β-cells and
subsequently reduce the activation and chemotaxis of macrophages.
2.3. Design, synthesis, and purification of peptide analogs
Peptide sequences were derived form murine CCR2, a GPCR for MCP-1.
Prior studies have suggested that the first and second extracellular loops
of CCR2 play an important role in MCP-1/CCR2 binding [39,40]. Thus, two
sequences derived from loop-1 and loop-2 of CCR2, respectively, were
chosen as initial affinity peptides candidates for murine MCP-1 binding.
While the complete sequence of loop-1 (112AHYAANEWVFGNIMCK127 or
P1 in Fig. 2A) was synthesized, only a portion of loop-2 (211WKNFQTI217
or P2 in Fig. 2A) was selected and synthesized. This sequence (P2) is a
comparison result of sequence alignment between murine and human
CCR2, as a peptide sequence (WNNFHTIMR) derived from human CCR2
was previously found to exhibit high affinity for human MCP-1 [40]. Thus,
it was hypothesized that the selected P2 sequence should contribute to
affinity binding between murine MCP-1 and CCR2, due to the high
sequence homology of chemokines and their receptors between species.
A terminal cysteine residue and an additional glycine spacer were added
for conjugating the peptide within crosslinked PEGDA hydrogels.
All peptides were synthesized (Applied Biosystems 433A) using
solid phase Fmoc chemistry with HBTU/HOBt amino acid activation.
Peptides were deprotected and cleaved from the solid support using
5 wt.% phenol in 95% TFA, 2.5% triisopropylsilane (TIPS), and 2.5%
water. Peptides were purified by reverse-phase HPLC and charac-
terized by MALDI-TOF mass spectroscopy. In all peptide analogs, an
additional cysteine residue was added to the N-terminus for covalent
conjugation within PEGDA hydrogel network via a thiol-acrylate
photopolymerization [10,37,38].
2. Materials and methods
2.1. Materials
Standard amino acids (Fmoc-capped), Fmoc-Rink-amide MBHA
resin, and reagents for solid phase peptide synthesis were obtained
from Anaspec (Fremont, CA). Diethylene glycol spacer (Fmoc-PEG2-
Suc-OH) was obtained from Bachem (Torrance, CA). Recombinant
murine MCP-1 protein (M.W. 13.8 kDa) and ELISA development kits
for murine MCP-1 and IL-1β were purchased from Peprotech (Rocky
Hill, NJ). Rabbit anti-mouse MCP-1 antibody was acquired from Santa
Cruz Biotechnology (Santa Cruz, CA). Goat anti-rabbit HRP-conjugated
antibody was purchased from Abcam (Cambridge, MA). AlamarBlue®
reagent was obtained from Invitrogen (Carlsbad, CA). All other chemi-
cals were received from Sigma-Aldrich (St. Louis, MO) unless noted
otherwise.
2.4. Synthesis of PEGDA macromer, LAP photoinitiator, and PEG
hydrogels
The synthesis of poly(ethylene glycol) diacrylate (PEGDA, 10 kDa)
has been described elsewhere [37,41]. Briefly, hydroxyl-terminated
PEG (PEG, 10 kDa) was first dissolved in toluene at above 60 °C and
reacted with 4 molarity excess (to hydroxyl group) of acryloyl
chloride and triethylamine (TEA) at room temperature for overnight.
The product was filtered through neutral alumina to remove TEA-HCl
salt and precipitated in cold ether. The PEGDA macromer was then
filtered and dried in vacuo. The degree of acrylation was over 95% as

386
C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
determined by 1H NMR. The photoinitiator lithium phenyl-2,4,6-
trimethylbenzoylphosphinate (LAP) was synthesized by reacting an
equimolar amount of dimethyl phenylphosphonite with 2,4,6-
trimethylbenzoyl chloride [42,43]. Briefly, 2,4,6-trimethylbenzoyl
chloride was added drop-wise into dimethyl phenylphosphonite
and allowed to react for 18 h, followed by the addition of 4 molarity
excess of lithium bromide in 2-butanone. The reaction mixture was
heated to 50 °C until a solid precipitate formed (∼10 min). The mixture
was cooled to room temperature and allowed to rest for 4 h. The
resulting product was filtered and washed extensively with excess
amount of 2-butanone and diethyl ether, then dried in vacuum. Previous
studies have shown that the photoinitiator LAP is more efficient,
compared to the widely used 2-hydroxy-4′-(2-hydroxyethoxy)-2-
methylpropiophenone (Irgacure-2959) in fabricating hydrogels from
PEGDA with 365 nm light. The photo-curing time required to obtain a
cured hydrogel is reduced to less than 2 min with ∼7 mW/cm2 of
365 nm light, while maintaining cytocompatibility [42].
PEG–peptide hydrogels were synthesized via a thiol-acrylate
mixed-mode photopolymerization reaction [10,38]. Briefly, 10 wt.%
of PEGDA (M.W. 10 kDa) macromer was dissolved in pH 7.4 PBS
containing 0.05 wt.% of photoinitiator (LAP) and selected amounts of
peptides. The pre-polymer solutions were mixed and injected
between two glass slides separated by 1 mm thick Teflon spacers,
followed by photopolymerization under a UV lamp (365 nm, 7 mW/
cm2) at ambient temperature for 2 min. Equal sized hydrogel disks
were obtained using a 7 mm diameter biopsy punch.
MCP-1 was mixed with 10 wt.% PEGDA (Mn ∼10 kDa), 0.05 wt.% LAP,
and peptide analogs at desired concentrations. The mixed macromer
solutions were injected between two glass slides separated by 0.5 mm
Teflon spacers and exposed to UV light (365 nm, 7 mW/cm2) for 2min
to allow photo-curing. The resulting gels were cut into identical
circular disks with a 7 mm diameter biopsy punch and placed in at
least 2 mL of release buffer (0.1% BSA, 1 mM EDTA in PBS, pH7.4). To
maintain the bioactivity of the released MCP-1, gel samples were
placed at 4 °C on an orbital shaker. After sampling at predetermined
time intervals, gels were transferred into fresh release buffer. Samples
were stored at −20 °C until determination of released MCP-1
concentrations using a murine MCP1/JE ELISA kit. It should be noted
that storage of the collected cytokine samples at low temperature
(−20 °C) for extended period of time often leads to the loss of
immunoreactivity. However, for recombinant MCP-1 samples, we did
not observe significant loss of immunoreactivity after 24 h of storage
(93.8 8.4%). Around 50% decrease in immunoreactivity may occur
after storage for 72 h at −20 °C.
2.8. Photo-encapsulation of MIN6 cells and cytokine-stimulated MCP-1
secretion
MIN6 cells were trypsinized from flasks and dispersed into single
cells, mixed with sterile-filtered macromer solutions containing
10 wt.% PEGDA, 0.05 wt.% LAP, and desired concentrations of peptide
analogs. Mixed cell-macromer solutions (30 μL per sample) were
injected into 1 mL sterile syringes with the tips cut-off and placed
under UV light for 2 min to allow photo-gelation and encapsulation.
Encapsulated cells were placed in RPMI-1640 medium for at least 1h
to remove any un-encapsulated cells located on the gel surface. After
which, the cell-laden hydrogels were transferred into fresh media or
media containing a cytokine cocktail, including IFN-γ (750 units/mL),
IL-1β (10 units/mL), and TNFα (500 units/mL) for 6 h or 24 h. Media
were collected and the concentrations of the cell-secreted MCP-1
from gels were determined by ELISA. The amount of cell-secreted
MCP-1 was normalized to cell viability determined by CellTiter-Glo
reagent as described elsewhere [10].
2.5. Rheological analysis of hydrogel equilibrium properties
Hydrogels with or without peptides as described above were
prepared for rheological property measurements. Prior to measure-
ment, gels were incubated in PBS at room temperature for 48 h to reach
equilibrium swelling. Rheological properties of the swollen hydrogels
were measured using a modified ARES rotational rheometer (TA
Instruments). Strain sweep experiments (ω=10 rad s−1) were con-
ducted on equilibrated hydrogels between parallel plates. Gel moduli
were determined within the materials linear viscoelastic regions.
2.6. Characterization of peptide affinity and incorporation into PEG
2.9. Statistical analysis
hydrogels
Results were presented as Mean SEM. All experiments contain-
ing 3 to 4 replicates were repeated independently at least three times.
Statistical analysis was performed using one-way ANOVA, followed by
Turkey test with 95% confidence intervals.
Peptide affinity was qualitatively revealed by a modified ELISA
similar to a prior study [44]. Briefly, purified peptides (100 μM) were
coated in a Maxisorp 96-well ELISA plate at 4 °C overnight, washed 3
times (0.05% Tween-20 in PBS), and blocked with 1% BSA for at least
1 h. The plate was then incubated with recombinant murine MCP-1 at
various concentrations for 2 h at room temperature. Rabbit anti-
mouse MCP-1 antibody and goat anti-rabbit HRP-conjugated antibody
were used as primary (2 h) and secondary (1 h) antibodies, respec-
tively. Finally, TMB substrate was added and the plate was incubated
for 15–20 min before stop solution (0.2 N HCl) was added. Absorbance
of the plate at 450 nm was measured with a microplate reader.
The degree of peptide incorporation in PEG hydrogels was
determined by quantifying the concentrations of unconjugated pep-
tides. Briefly, affinity peptides (1 mM) were included in a 10 wt.%
PEGDA pre-polymer solution and co-polymerized as described above.
The resulting gels of identical size (7 mm dia.×1 mm thickness) were
incubated in pH 7.4 PBS at ambient temperature for 48 h to allow for
release of unconjugated peptides. The supernatants were collected
and the concentrations of the unconjugated peptides were quantified
by Fluoraldehyde reagent (PIERCE).
3. Results and discussion
3.1. Effects of recombinant MCP-1 on monocyte activities in vitro
MCP-1 is known to exert a variety of cellular activities on
monocytes, including chemotaxis, proliferation, and secretion of
cytokines such as interleukin-1β (IL-1β). These cellular activities
were verified in vitro using a model murine monocyte/macrophage
cell line, J774a.1. As shown in Fig. 1A, J774a.1 cells migrate
directionally in response to increasing MCP-1 concentration, demon-
strating the chemotactic effect of MCP-1 to monocytes. In addition to
chemotaxis, MCP-1 also induces enhanced cell proliferation (Fig. 1B)
and IL-1β secretion (Fig. 1C). Collectively, these in vitro results
confirm the importance of MCP-1 on recruiting and activating
monocytes.
3.2. Affinity peptide identification and peptide analog design rationale
2.7. Characterization of recombinant MCP-1 release
Affinity peptides that bind to murine MCP-1 were identified from
its receptor CCR2 (Fig. 2A) [39,40]. The sequence P1 (112AHYAA-
NEQVFQNIMCK127) was derived from the entire extracellular loop-1
Recombinant murine MCP-1 was photo-encapsulated in PEGDA
hydrogels to characterize its in vitro release. Briefly, 100 ng/mL of

C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
387
Fig. 2. (A) Model of extracellular loops of murine CCR2, a murine MCP-1 receptor.
Sequences P1 and P2 were derived from extracellular loop 1 and loop2, respectively.
(B) Affinity ELISA to determine binding affinity of P1 and P2 to recombinant MCP-1. BSA
and a scrambled sequence were used as negative controls (Mean SEM, n=4).
After purification, the identity of P1 and P2 were verified by MALDI-
TOF (See supporting information). A modified ELISA was developed to
determine whether P1 and P2 indeed exhibit affinity to murine MCP-
1. As shown in Fig. 2B, P1 and P2 are both affinity peptides for murine
MCP-1, as increasing the concentrations of soluble MCP-1 in the
incubation buffer enhances its binding to the surface-adsorbed
peptides. Wells coated with BSA, or a control peptide with identical
amino acids but scrambled sequence of P2, were used as controls and
only minimal non-specific interactions were observed at very high
MCP-1 concentrations (Fig. 2B). Since both P1 and P2 exhibit similar
affinity to MCP-1 and P2 (WKNFQTI) is a shorter peptide that allows
facile synthesis and modification, only P2 were used in the subsequent
studies.
3.3. Conjugation of peptide into hydrogels, peptide-functionalized
hydrogel properties
Fig. 1. Effects of recombinant MCP-1 on in vitro (A) chemotaxis, (B) proliferation, and
(C) secretion of IL-1β of a monocyte/macrophage cell line, J774a.1 (Mean SEM,
n=4). Asterisks represent statistical significance compared to non-treated cells
(pb0.05).
Previously, several peptide-functionalized PEG hydrogels were
engineered for a variety of controlled release and tissue engineering
applications [10,13,37]. An efficient and convenient peptide conju-
gation scheme, namely thiol-acrylate photopolymerization (Fig. 3A),
was utilized in this study to conjugate and crosslink affinity peptide
analogs within PEG hydrogels. It has been revealed that the archi-
tecture of affinity peptides crosslinked in PEG hydrogels significantly
affects their affinity binding to target proteins [37]. Since the main
purpose of this study was to retain MCP-1 within hydrogels, peptide
structures that offer high affinity binding are particularly favor-
able. Thus, three P2 analogs with an additional N-terminal cysteine
and different spacers separating the cysteine residue (crosslinking
site) and the affinity binding sequence (Fig. 3B) were synthesized
and conjugated within PEG hydrogels to examine the relationship
between peptide architecture and affinity binding, as well as to
enhance the retention of MCP-1 within the permissive PEG
hydrogels.
of murine CCR2, while the sequence P2 (211WKNFQTI217) was only a
partial sequence of the 2nd loop. The rationale of selecting P2 as a
potential affinity peptide for murine MCP-1 was based on previous
efforts in identifying human MCP-1 binding peptide [40]. It was
shown that the sequence WNNFHTI (derived from loop-2 from
human CCR2) binds to human MCP-1 with an equilibrium dissociation
constant (Kd) of 22 μm [40]. It was therefore hypothesized that a
sequence derived from the same location on murine CCR2 may also
contribute affinity binding to murine MCP-1, in part due to highly
conserved sequences between murine and human MCP-1 and CCR2.
The two sequences (P1 and P2) were synthesized using a solid
phase peptide synthesizer and purified with reverse-phase HPLC.

388
C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
Fig. 3. (A) Reaction scheme of thiol-acrylate photopolymerization. (B) Chemical structure of a control peptide (CGKFIQWNT) with additional terminal cysteine and glycine residues.
(C) Chemical structure of the affinity peptide (CGWKNFQTI) with an additional cysteine residue and a glycine spacer. (D) Chemical structure of the affinity peptide with an additional
cysteine residue and three diethylene glycol spacers (COOOWKNFQTI, O: diethylene glycol spacer).
Prior to studying the affinity binding and retention of MCP-1 in PEG
hydrogels, the efficiency of peptide incorporation within crosslinked
PEG hydrogels was quantified to be at least 80% (data not shown).
Further, it was determined that the incorporation of a small amount of
P2 peptide analogs does not significantly alter the physical properties of
the PEGDA-10 kDa hydrogels (10 wt.% or 10 mM). As shown in Fig. 4,
rheological analysis revealed similar gel mechanical properties, imply-
ing a similar gel crosslinking density. Further, equilibrium gel swelling
was not affected by the incorporation of the affinity peptides (data not
shown), suggesting similar transport properties for small molecular
weight molecules such as glucose, insulin, and most importantly in this
study, cell-secreted MCP-1. It is important to note that maintaining the
high permeability of hydrogels for small biomolecules is of paramount
importance in the encapsulation of pancreatic β-cells due to the critical
role of maintaining β-cells viability and facile insulin diffusion.
Previously, it has been shown that the diffusion of insulin in the highly
swollen PEG hydrogels was not affected by the incorporation of small
quantities of crosslinkable peptides [10].
major disadvantage of highly permeable PEG hydrogels in controlling
the diffusion of small molecular weight biomolecules. Further, an
intermediate release of MCP-1 was obtained when using affinity
peptides with only one glycine spacer separating the crosslinking site
3.4. Recombinant MCP-1 release from affinity hydrogels
Toward the goal of sequestering cell-secreted MCP-1 from PEG
hydrogels, the efficiency of affinity peptide-functionalized PEG
hydrogels in controlling the diffusion of in situ encapsulated
recombinant MCP-1 was examined. Fig. 5 shows a rapid release of
large quantities of MCP-1 from unmodified PEG hydrogels, revealing a
Fig. 4. Rheological analysis of PEG hydrogels incorporating with or without affinity
peptides. (PEGDA: 10 kDa, 10 wt.%∼10 mM; peptide concentration: 1 mM) (Mean
SEM, n=3).

C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
389
various concentrations of cytokine cocktails to stimulate the synthesis
and secretion of MCP-1. Fig. 6A shows that the secretion of MCP-1
from MIN6s encapsulated in unmodified PEG hydrogels increases
significantly and dose dependently under the stimulation of the
cytokine cocktail, demonstrating that unmodified PEG hydrogels were
unable to prevent the diffusion of small antigenic molecules, such as
MCP-1, from permissive hydrogels network. The release of cell-
secreted MCP-1 from affinity peptide-functionalized PEG hydrogels,
however, was decreased significantly (Fig. 6B), with a trend similar to
that shown in the release of recombinant MCP-1 (Fig. 5). Specifically,
more than 60% of MCP-1 was sequestered and retained in PEG
hydrogels after 24 h of cytokine stimulation.
It should be noted that the purpose of using high concentrations of
cytokine cocktail in Fig. 6 was to augment the effects of affinity
binding on retaining MCP-1 within a short period of time (24 h).
Clinically, the average amount of human MCP-1 (M.W. 8 kDa)
secreted by isolated islets is around 30 pg/islet/day (∼0.4 pmol) for
high MCP-1 secreting islets. The affinity hydrogel formulation used in
this study (100 µM peptides in 30 μL gels) theoretically is sufficient to
sequester 3nmol of MCP-1 in its maximal capacity. As noted before,
increasing peptide affinity or concentration can further enhance the
capacity of the affinity hydrogels in retaining cell-secreted MCP-1.
Therefore, it is expected that the formulation of this affinity hydrogel
platform can be readily adjusted to fulfill future clinically-relevant
applications. Future studies will be focused on developing in vitro co-
culture systems containing cell-laden hydrogels and monocytes, as
Fig. 5. Recombinant murine MCP-1 release from PEG gels (10%, 10 kDa) with or without
affinity peptide. Peptide incorporation concentration: 100 µM. (Mean SEM, n=4).
(cysteine) and the affinity binding motif (WKNFQTI). Finally, very
limited release of MCP-1 was observed from affinity peptide with
three diethylene glycol spacers. Fig. 5 demonstrates that the release of
MCP-1 from PEG hydrogels can be reduced significantly due to the
binding of MCP-1 to network-immobilized affinity peptides. Further-
more, the binding can be further enhanced, at the same peptide
incorporating concentration, by using affinity peptides with longer
spacers. This demonstrates an important aspect of affinity binding in
hydrogels using crosslinked peptide. Previously, work has shown that
increasing spacer distance enhances the affinity binding of crosslinked
peptides and soluble growth factor [37]. As shown in Fig. 5, enhanced
binding leads to reduced MCP-1 release. The benefits of using peptides
with specific affinity for cytokines are multifaceted. For example,
similar tunable protein retention/release effects can be achieved by
changing peptide affinity, concentration, architecture, and presenta-
tion in hydrogels. For example, increasing peptide concentrations in
hydrogels or further increasing the length of the di(ethylene glycol)
spacer may also result in similar enhanced protein retention, as
suggested by prior research results [13,37,41,45]. This former
approach, however, will likely decrease the gel crosslinking density
significantly due to the excessive chain transfer and termination
during polymerization, while the latter may decrease the purity of the
peptide due to the increased sequence length. Note that the absolute
amounts of MCP-1 release shown in Fig. 5 may be skewed due to the
practical limitation in the use of frozen MCP-1 aliquots and the short-
term storage of the collected samples that result in decreased
immunoreactivity in ELISA. However, the shape of the curves and
the relative release amounts using different peptide architectures
should remain the same, as all the samples were processed with the
same conditions.
3.5. MCP-1 secreted by MIN6 cells encapsulated in affinity hydrogels
MIN6 cells, a murine β-cell line, have been used in a variety of
in vitro β-cell models focusing on understanding molecular and cell
biology of type 1 diabetes. MIN6 cells have also been shown to secret
MCP-1 under the stimulation of a cytokine cocktail containing tumor
necrosis factor α (TNFα), interleukin-1β (IL-1β), and interferon
γ (IFNγ) [33]. Here, MIN6 cells were encapsulated in either
unmodified PEG hydrogels or gels functionalized with MCP-1 binding
peptides. A high cell encapsulation density (200,000 cells/gel) was
employed to increase cell–cell contact that maintains MIN6 cell
viability during the course of study (data not shown). Following
photo-encapsulation, the cell-laden hydrogels were incubated in
RPMI-1640 medium for one hour to allow for the removal of any sol
fraction. Cell-laden hydrogels were transferred to media containing
Fig. 6. (A) MCP-1 release from MIN6 cells encapsulated in 10% PEG hydrogels incubated
with or without a cytokine cocktail (1x cytokine: 750 units/mL IFN-γ, 10 units/mL IL-1β,
and 500 units/mL TNFα). (B) MCP-1 release from MIN6 cells encapsulated in PEG
hydrogels or PEG–peptide hydrogels incubated in 1x cytokine cocktail (peptide=
100 µM, mean SEM, n=4). Asterisks represent statistical significance between
indicated groups (pb0.05).

390
C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
[10] C.C. Lin, K.S. Anseth, Glucagon-like peptide-1 functionalized PEG hydrogels
promote survival and function of encapsulated pancreatic beta-cells., Biomacro-
molecules 10 (2009) 2460–2467.
[11] C.C. Lin, K.S. Anseth, PEG hydrogels for the controlled release of biomolecules in
regenerative medicine, Pharmaceutical Research 26 (2009) 631–643.
[12] C.C. Lin, A.T. Metters, Hydrogels in controlled release formulations: network
design and mathematical modeling, Advanced Drug Delivery Reviews 58 (2006)
1379–1408.
well as in vivo studies to determine the therapeutic efficacy of this
strategy in reducing local inflammation. Further, affinity peptides
capable of binding to other chemokines (e.g., (IFNγ)-inducible
protein-10 (IP-10), macrophage inflammatory protein (MIP), etc.)
might be implemented together with MCP-1 binding peptides
developed here to augment the effects of reducing chemokine
diffusion.
[13] C.C. Lin, A.T. Metters, K.S. Anseth, Functional PEG–peptide hydrogels to modulate
local inflammation induced by the pro-inflammatory cytokine TNFalpha.
Biomaterials 30 (2009) 4907–4914.
[14] C.Y. Cheung, S.J. McCartney, K.S. Anseth, Synthesis of polymerizable superoxide
dismutase mimetics to reduce reactive oxygen species damage in transplanted
biomedical devices, Advanced Functional Materials 18 (2008) 3119–3126.
[15] R.A.F. Clark, Basics of cutaneous wound repair, Journal of Dermatologic Surgery
and Oncology 19 (1993) 693–706.
[16] Z.D. Yang, M. Chen, J.D. Ellett, J.D. Carter, K.L. Brayman, J.L. Nadler, Inflammatory
blockade improves human pancreatic islet function and viability, American
Journal of Transplantation 5 (2005) 475–483.
[17] S.A. Eming, T. Krieg, J.M. Davidson, Inflammation in wound repair: molecular and
cellular mechanisms, Journal of Investigative Dermatology 127 (2007) 514–525.
[18] M.D. Miller, M.S. Krangel, Biology and biochemistry of the chemokines — a family
of chemotactic and inflammatory cytokines, Critical Reviews in Immunology 12
(1992) 17–46.
[19] P.M. Murphy, The molecular-biology of leukocyte chemoattractant receptors,
Annual Review of Immunology 12 (1994) 593–633.
[20] A. Rot, U.H. von Andrian, Chemokines in innate and adaptive host defense: basic
chemokinese grammar for immune cells, Annual Review of Immunology 22 (2004)
891–928.
[21] Y.L. Jiang, D.I. Beller, G. Frendl, D.T. Graves, Monocyte chemoattractant protein-1
regulates adhesion molecule expression and cytokine production in human
monocytes, Journal of Immunology 148 (1992) 2423–2428.
[22] B.J. Rollins, A. Walz, M. Baggiolini, Recombinant human MCP-1/JE induces
chemotaxis, Calcium Flux, and the Respiratory Burst in Human Monocytes, Blood
78 (1991) 1112–1116.
[23] S. Sozzani, M. Locati, P. Allavena, J. VanDamme, A. Mantovani, Chemokines: a
superfamily of chemotactic cytokines, International Journal of Clinical & Labora-
tory Research 26 (1996) 69–82.
[24] L.L. Xu, M.K. Warren, W.L. Rose, W.H. Gong, J.M. Wang, Human recombinant
monocyte chemotactic protein and other c–c chemokines bind and induce
directional migration of dendritic cells in vitro, Journal of Leukocyte Biology 60
(1996) 365–371.
[25] A. Vecchi, L. Massimiliano, S. Ramponi, W. Luini, S. Bernasconi, R. Bonecchi, P.
Allavena, M. Parmentier, A. Mantovani, S. Sozzani, Differential responsiveness to
constitutive vs. inducible chemokines of immature and mature mouse dendritic
cells, Journal of Leukocyte Biology 66 (1999) 489–494.
[26] P. Loetscher, M. Seitz, M. Baggiolini, B. Moser, Interleukin-2 regulates CC
chemokine receptor expression and chemotactic responsiveness in T lympho-
cytes, Journal of Experimental Medicine 184 (1996) 569–577.
[27] S. Arimilli, W. Ferlin, N. Solvason, S. Deshpande, M. Howard, S. Mocci, Chemokines
in autoimmune diseases, Immunological Reviews 177 (2000) 43–51.
[28] M.J. Cameron, G.A. Arreaza, M. Grattan, C. Meagher, S. Sharif, M.D. Burdick, R.M.
Strieter, D.N. Cook, T.L. Delovitch, Differential expression of CC chemokines and
the CCR5 receptor in the pancreas is associated with progression to type I diabetes,
Journal of Immunology 165 (2000) 1102–1110.
4. Conclusions
An affinity hydrogel platform with the ability to sequester
selectively a chemokine, MCP-1, was developed. In this study, two
peptide sequences with binding affinity to murine MCP-1 were
identified from extracellular loops of murine CCR2. A peptide analog,
namely WKNFQTI, was incorporated within highly swollen PEG
hydrogels to selectively and specifically bind to recombinant murine
MCP-1 without affecting the bulk properties of hydrogels. Further, the
release of in situ encapsulated MCP-1 was reduced by changing the
peptide architecture (i.e., adding a tri-diethylene glycol spacer).
Similarly, the release of MCP-1 secreted by the encapsulated MIN6
cells was reduced significantly. In addition to the encapsulation of
pancreatic β-cells for type 1 diabetes, this strategy also has the
potential to enhance the immuno-isolating function of PEG hydrogels,
through selectively reducing diffusion of antigens, in the encapsula-
tion and delivery of allogenic cell types.
Acknowledgement
This work was supported in part by the National Institutes of Health
(R01DK076084) and the Howard Hughes Medical Institute. The
authors would like to thank the support to P.D.B. from the Research
Experience for Undergraduates (REU) in Functional Materials Science
and Engineering supported by the National Science Foundation. The
authors also thank the support to A.A.A. from Graduate Assistance in
Areas of National Need (GAANN).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jconrel.2009.11.022.
References
[29] M. Rotondi, L. Chiovato, S. Romagnani, M. Serio, P. Romagnani, Role of chemokines
in endocrine autoimmune diseases, Endocrine Reviews 28 (2007) 492–520.
[30] L. Piemonti, B.E. Leone, R. Nano, A. Saccani, P. Monti, P. Maffi, G. Bianchi, A. Sica, G.
Peri, R. Melzi, L. Aldrighetti, A. Secchi, V. Di Carlo, P. Allavena, F. Bertuzzi, Human
pancreatic islets produce and secrete MCP-1/CCL2: Relevance in human islet
transplantation, Diabetes 51 (2002) 55–65.
[31] G. Chipitsyna, Q.K. Gong, C.F. Gray, Y. Haroon, E. Kamer, H.A. Arafat, Induction of
monocyte chemoattractant protein-1 expression by angiotensin II in the
pancreatic islets and beta-cells, Endocrinology 148 (2007) 2198–2208.
[32] B. Kutlu, M.I. Darville, A.K. Cardozo, D.L. Eizirik, Molecular regulation of monocyte
chemoattractant protein-1 expression in pancreatic beta-cells, Diabetes 52 (2003)
348–355.
[33] M.S. Baker, X.J. Chen, A. Rotramel, J. Nelson, D.B. Kaufman, Proinflammatory
cytokines induce NF-kappa B-dependent/NO-independent chemokine gene
expression in MIN6 beta cells, Journal of Surgical Research 110 (2003) 295–303.
[34] I. Zineh, A.L. Beitelshees, J.H. Silverstein, M.J. Haller, Serum monocyte chemoat-
tractant protein-1 concentrations associate with diabetes status but not arterial
stiffness in children with type 1 diabetes, Diabetes Care 32 (2009) 465–467.
[35] A.C. Ogliari, R. Caldara, C. Socci, V. Sordi, N. Cagni, M.P. Moretti, A. Dell'Acqua, A.
Mercalli, M. Scavini, A. Secchi, E. Bonifacio, E. Bosi, L. Piemonti, High levels of donor
CCL2/MCP-1 predict graft-related complications and poor graft survival after
kidney-pancreas transplantation, American Journal of Transplantation 8 (2008)
1303–1311.
[1] J.T. Wilson, E.L. Chaikof, Challenges and emerging technologies in the immunoi-
solation of cells and tissues, Advanced Drug Delivery Reviews 60 (2008) 124–145.
[2] M. Tanemura, A. Saga, K. Kawamoto, T. Machida, T. Deguchi, T. Nishida, Y. Sawa, Y.
Doki, M. Mori, T. Ito, Rapamycin induces autophagy in islets: relevance in islet
transplantation, Transplantation Proceedings 41 (2009) 334–338.
[3] A. Aronovitz, J. Josefson, A. Fisher, M. Newman, E. Hughes, F. Chen, D.S. Moons, H.
Kiyokawa, W.L. Lowe, Rapamycin inhibits growth factor-induced cell cycle
regulation in pancreatic beta cells, Journal of Investigative Medicine 56 (2008)
985–996.
[4] M. Sakata, H. Yasuda, H. Moriyama, K. Yamada, R. Kotani, M. Kurohara, Y.
Okumachi, M. Kishi, T. Arai, K. Hara, H. Hamada, K. Yokono, M. Nagata, Prevention
of recurrent but not spontaneous autoimmune diabetes by transplanted NOD
islets adenovirally Transduced with immunomodulating molecules, Diabetes
Research and Clinical Practice 80 (2008) 352–359.
[5] C.P. Larsen, S.J. Knechtle, A. Adams, T. Pearson, A.D. Kirk, A new look at blockade of
T-cell costimulation: a therapeutic strategy for long-term maintenance immuno-
suppression, American Journal of Transplantation 6 (2006) 876–883.
[6] E. Bell, X.P. Cao, J.A. Moibi, S.R. Greene, R. Young, M. Trucco, Z.Y. Gao, F.M.
Matschinsky, S.P. Deng, J.F. Markman, A. Naji, B.A. Wolf, Rapamycin has
deleterious effect on MIN-6 cells and rat and human islets, Diabetes 52 (2003)
2731–2739.
a
[7] M.C. Cushing, K.S. Anseth, Hydrogel cell cultures, Science 316 (2007) 1133–1134.
[8] C.R. Nuttelman, M.A. Rice, A.E. Rydholm, C.N. Salinas, D.N. Shah, K.S. Anseth,
Macromolecular monomers for the synthesis of hydrogel niches and their
application in cell encapsulation and tissue engineering, Progress in Polymer
Science 33 (2008) 167–179.
[9] L.M. Weber, K.N. Hayda, K.S. Anseth, Cell–matrix interactions improve beta-cell
survival and insulin secretion in three-dimensional culture, Tissue Engineering
Part A 14 (2008) 1959–1968.
[36] L.J. Suggs, M.S. Shive, C.A. Garcia, J.M. Anderson, A.G. Mikos, In vitro cytotoxicity
and in vivo biocompatibility of poly(propylene fumarate-co-ethylene glycol)
hydrogels, Journal of Biomedical Materials Research 46 (1999) 22–32.
[37] C.C. Lin, K.S. Anseth, Controlling affinity binding with peptide-functionalized
poly(ethylene glycol) hydrogels, Advanced Functional Materials 19 (2009)
2325–2331.
[38] C.N. Salinas, K.S. Anseth, Mixed mode thiol-acrylate photopolymerizations for the
synthesis of PEG–peptide hydrogels, Macromolecules 41 (2008) 6019–6026.

C.-C. Lin et al. / Journal of Controlled Release 142 (2010) 384–391
391
[39] K.H. Han, S.R. Green, R.K. Tangirala, S. Tanaka, O. Quehenberger, Role of the first
extracellular loop in the functional activation of CCR2 — the first extracellular loop
contains distinct domains necessary for both agonist binding and transmembrane
signaling, Journal of Biological Chemistry 274 (1999) 32055–32062.
phosphinate: polymerization rate and cytocompatibility, Biomaterials 30 (2009)
6702–6707.
[43] T. Majima, W. Schnabel, W. Weber, Phenyl-2, 4, 6-trimethylbenzoylphosphinates
as water-soluble photoinitiators — generation and reactivity of O P(C6h5)(O−
)
[40] M.Y. Kim, C.W. Byeon, K.H. Hong, K.H. Han, S. Jeong, Inhibition of the angiogenesis
by the MCP-1 (monocyte chemoattractant protein-1) binding peptide, FEBS Letters
579 (2005) 1597–1601.
[41] C.C. Lin, A.T. Metters, Metal-chelating affinity hydrogels for sustained protein
release, Journal of Biomedical Materials Research Part A 83A (2007) 954–964.
[42] B.D. Fairbanks, M.P. Schwartz, C.N. Bowman, K.S. Anseth, Photoinitiated
polymerization of PEG-diacrylate with lithium phenyl-2, 4, 6-trimethylbenzoyl-
radical-anions, Makromolekulare Chemie-Macromolecular Chemistry and Physics
192 (1991) 2307–2315.
[44] J.Y. Lee, J.E. Choo, Y.S. Choi, J.S. Suh, S.J. Lee, C.P. Chung, Y.J. Park, Osteoblastic
differentiation of human bone marrow stromal cells in self-assembled BMP-2
receptor-binding peptide-amphiphiles, Biomaterials 30 (2009) 3532–3541.
[45] C.C. Lin, A.T. Metters, Bifunctional monolithic affinity hydrogels for dual-protein
delivery, Biomacromolecules 9 (2008) 789–795.