Protease-Sensitive Hydrogel Biomaterials with Tunable Modulus and Adhesion Ligand Gradients for 3D Vascular Sprouting

Article abstract of DOI:10.1021/acs.biomac.8b00519

Biomaterial strategies focused on designing scaffolds with physiologically relevant gradients provide a promising means for elucidating 3D vascular cell responses to spatial and temporal variations in matrix properties. In this study, we present a photopolymerization approach, ascending photofrontal free-radical polymerization, to generate proteolytically degradable hydrogel scaffolds of poly(ethylene) glycol with tunable continuous gradients of (1) elastic modulus (slope of 80 Pa/mm) and uniform immobilized RGD concentration (2.06 ± 0.12 mM) and (2) immobilized concentration of the RGD cell-adhesion peptide ligand (slope of 58.8 μM/mm) and uniform elastic modulus (597 ± 22 Pa). Using a coculture model of vascular sprouting, scaffolds embedded with gradients of elastic modulus induced increases in the number of vascular sprouts in the opposing gradient direction, whereas RGD gradient scaffolds promoted increases in the length of vascular sprouts toward the gradient. Furthermore, increases in vascular sprout length were found to be prominent in regions containing higher immobilized RGD concentration.

Full text of DOI:10.1021/acs.biomac.8b00519

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
Protease-Sensitive Hydrogel Biomaterials with Decoupled Modulus
and Adhesion Ligand Gradients for 3D Vascular Sprouting
Yusheng J. He, Daniel A. Young, Merjem Mededovic, Kevin Li, Chengyue
Li, Kenneth Tichauer, David Venerus, and Georgia Papavasiliou
Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.8b00519 • Publication Date (Web): 25 Sep 2018
Downloaded from http://pubs.acs.org on September 27, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
Protease-Sensitive Hydrogel Biomaterials with
Tunable Modulus and Adhesion Ligand
Gradients for 3D Vascular Sprouting
10
11
12
13
14
15
16
17
18
19
20
Yusheng J. He,† Daniel A. Young,† Merjem Mededovic,† Kevin Li,† Chengyue Li,†
Kenneth Tichauer,† David Venerus,‡ and Georgia Papavasiliou∗,†
1
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†Illinois Institute of Technology, Biomedical Engineering Department, Chicago, IL, USA
‡Illinois Institute of Technology, Chemical and Biological Engineering Department,
Chicago, IL, USA
E-mail: papavasiliou@iit.edu
Phone: 1-312-567-5959
2
Abstract
3
4
Biomaterial strategies focused on designing scaffolds with physiologically relevant
gradients provide a promising means for elucidating 3D vascular cell responses to spatial
and temporal variations in matrix properties. In this study we present a photopoly-
merization approach, ascending photo-frontal free-radical polymerization, to generate
proteolytically degradable hydrogel scaffolds of poly(ethylene) glycol (PEG) with tun-
able continuous gradients of (1) elastic modulus (slope of 80 Pa/mm) and uniform
immobilized RGD concentration (2.06±0.12 mM) and (2) immobilized concentration of
the RGD cell-adhesion peptide ligand (slope of 58.8 µM/mm) and uniform elastic mod-
ulus (597± 22Pa). Using a co-culture model of vascular sprouting, scaffolds embedded
with gradients of elastic modulus induced increases in the number of vascular sprouts
in the opposing gradient direction, while RGD gradient scaffolds promoted increases in
5
6
7
8
9
10
11
12
13
1
ACS Paragon Plus Environment

Biomacromolecules
Page 2 of 42
1
2
3
4
5
6
7
8
1
2
3
the length of vascular sprouts towards the gradient. Furthermore, increases in vascular
sprout length were found to be prominent in regions containing higher immobilized
RGD concentration.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
Introduction
5
6
7
8
9
The inability to promote a stable microvascular network within implantable scaffolds is cur-
rently a major hurdle to the clinical translation of biomaterial-based strategies in tissue
engineering.1 The engineering of large volume, metabolically demanding tissue requires the
formation of rapid, stable, and functional neovascularization (new blood vessel formation)
for oxygen and nutrient transport and removal of waste products to support and maintain
10 viability, function, and restoration of newly formed tissue. Neovascularization is dependent
11 on vascular cell response to multiple spatiotemporal signals including, soluble and immobi-
12 lized biochemical factors, as well as gradients of mechanical properties and physical structure
13 provided by the 3D extracellular matrix (ECM).2 Direct measurement of these gradients in
14 native ECM is challenging; thus, a precise understanding of the role of different types of gra-
15 dients on vascularized tissue formation is difficult to achieve using animal models. To this
16 end, a variety of biomaterial approaches have been developed to create gradient scaffolds to
17 elucidate the role of spatiotemporally presented signals on neovascularization of engineered
18 tissues in vitro and in vivo.
19
Extensive research has focused on creating tissue engineered scaffolds with persistent gra-
20 dients of diffusible angiogenic growth factors to promote rapid and stable neovascularization
21 within implantable scaffolds.3–6 Other types of gradients have proven critical to this process,
22 including gradients of ECM structural and adhesive proteins,7–9 peptide composition10–12 as
23 well as gradients in mechanical and structural properties including matrix modulus13–15 and
24 proteolytic degradation rate.14 While these studies have shown promising results for stimu-
25 lating guided vascular cell response within 3D scaffolds, most have been limited to exploring
26 cellular responses to gradients of a single factor (i.e. ECM components or growth factor
2
ACS Paragon Plus Environment

Page 3 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
1
composition). For example, 3D culture models utilizing collagen scaffolds containing concen-
tration gradients of hyaluronic acid (HA) have demonstrated directional vascular sprouting
from cell spheroids downstream of the HA gradient.8 Depth-wise gradients of the cell ad-
hesive RGD peptide ligand have been shown to enhanced cellular infiltration within fibrous
HA scaffolds towards increasing RGD concentration as compared to the reverse gradient
orientation.12 Finally, immobilized gradients of VEGF in porous collagen scaffolds promote
chemotactic endothelial cell migration in 3D culture.4,16 Gradients in mechanical properties
and physical structure are present in a variety of native tissues and play a critical role in
biological repair processes, such as nerve regeneration and osteochondral tissue repair.17,18
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 Limited studies, however, have investigated the influence of multiple types of scaffold em-
11 bedded gradients (i.e. mechanical, degradative and biochemical) on 3D cell responses, and
12 more specifically on 3D vessel assembly.
13
Various polymerization methods have been developed to create biochemical and mechan-
14 ical continuous gradients within crosslinked hydrogel scaffolds for tissue engineering applica-
15 tions. Among these, two common fabrication approaches have been extensively exploited. In
16 one approach, gradient scaffolds are formed using a two-step process. The first step involves
17 establishment of a monomer concentration gradient using a fluidic device, such as a source-
18 sink diffusion device,19 a tree-like microfluidic device,20,21 or a device involving mixing of
19 prepolymer input streams using multiple syringe pumps.10,22 This is followed by a second
20 step where the gradient is stabilized upon crosslinking. Since diffusion acts to equilibrate
21 the established concentration gradient during the stabilization step, the final gradient profile
22 may be different from that imposed prior to crosslinking. The second approach involves
23 variation in the degree of crosslinking which results in a gradient of mechanical or structural
24 properties. This has been achieved by exposing prepolymer solutions to variable amount
25 of light during photocrosslinking using sliding15 or gradient greyscale photomasks.22 In this
26 method, the established crosslinking gradient may simultaneously induce concentration gra-
27 dients of immobilized bifunctional cues (cell adhesion ligand or growth factor composition)
3
ACS Paragon Plus Environment

Biomacromolecules
Page 4 of 42
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
in addition to spatial variations in mechanical properties.
As an alternative to the above-mentioned techniques, frontal polymerization has been
utilized to created polymer scaffolds with precise and complex gradient profiles. Frontal
polymerization occurs directionally from a localized reaction zone within which an increase
in the rate of polymerization occus due an input of energy (i.e. heat, light, or increase in
viscosity through the addition of a polymer seed) leading to the establishment of a polymer
reaction front.23–25 The ability to create polymeric fronts using an external flux of long-
wavelength light irradiation, photofrontal polymerization, provides a rapid, one-step process
for generating scaffolds with precise control in continuous gradient properties for tissue en-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 gineering applications.
11
In previously published work we developed a frontal photopolymerization approach to
12 create proteolytically degradable poly(ethylene) glycol (PEG) hydrogel scaffolds with tun-
13 able and continuous gradients of mechanical properties and embedded bifunctionality to
14 explore their role on guiding vascular cell response in 3D culture.14 Localized and controlled
15 perfusion of a photoinitiator within a monomeric mixture resulted in the estabilishment of a
16 localized reaction zone and in the formation of a photopolymerizable front that propoagated
17 through the polymer mixture to form crosslinked gradient hydrogel scaffolds upon exposure
18 to visible light. Using this approach, scaffolds with simultaneous gradients in elastic modu-
19 lus, immobilized RGD concentration, and matrix metalloproteinase (MMP)-sensitivity were
20 created.14,26,27 Scaffolds presenting simultaneous gradients of these factors induced 3D bi-
21 directional vascular sprout alignment and invasion; however, the coupled gradients made it
22 difficult to clearly elucidate the effects of these factors on 3D neovascularization responses.
23 This motivated the development of an alternative, distinct and robust method to create
24 proteolytically degradable scaffolds with decoupled gradients of initial (i.e. prior material
25 degradation) mechanical properties and immobilized RGD concentration to elucidate their
26 role on vascular sprouting in 3D culture.
27
Here we present a distinct free-radical frontal photopolymerization technique that en-
4
ACS Paragon Plus Environment

Page 5 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
1
ables modulation of continuous gradients in crosslink density and/or elastic modulus and
immobilized peptide concentration in biomimetic hydrogel scaffolds. This is achieved using
a dual programmable syringe pump system to control the composition and flow rates of
two distinct prepolymer solutions in the feed stream entering a propagating reaction front
in a polymerization reaction chamber exposed to visible light. This allows for rapid sta-
bilization of concentration gradients in the precursor solutions enabling precise control of
the established gradients within the scaffolds. Using this approach, we created two types
of gradient scaffolds: (1) scaffolds with an elastic modulus gradient (slope = 80 Pa/mm)
and uniform immobilized concentration of RGD ( 2.06±0.12 mM) and (2) scaffolds with
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 gradients of immobilized RGD concentration (slope = 58.8 µM/mm) and uniform elastic
11 modulus ( 597± 22Pa). In each case we fully characterized the gradient scaffolds in terms of
12 spatial variations in mechanical properties, immobilized RGD concentration, and proteolytic
13 degradation. Finally, we evaluated the role of gradients of elastic modulus and immobilized
14 RGD concentration on 3D vascular sprouting using an in vitro co-culture assay of sprouting
15 angiogenesis. We hypothesized that scaffolds with individual gradients in elastic modulus or
16 Immobilized RGD would stimulate uni-directional vascular sprouting in response to spatial
17 variations of these factors.
18 Experimental
19 Materials
20 Dimethylformamide (DMF), O-benzotriazole-N,N,N’,N’-tetra- methyl-uronium-hexafluoro-
21 phosphate (HBTU), trifluoroacetic acid (TFA), Fmoc-amino acids and the Wang resin were
22 obtained from AAPPTec (Louisville, KY). N,N-Diisopropylethylamine (DIEA), thioanisole,
23 triisopropylsilane (TIS) and diethyl ether were obtained from Fisher Scientific (Hanover
24 Park, IL). Piperidine, phenol, N-vinylpyrrolidone (NVP), triethanolamine (TEA) and eosin
25 Y were obtained from Sigma Aldrich (St. Louis, MO).
5
ACS Paragon Plus Environment

Biomacromolecules
Page 6 of 42
1
2
3
4
1
Peptide Synthesis, Design, and Purification
5
6
7
8
9
2
3
4
5
6
7
8
9
To render the scaffolds susceptible to cell mediated proteolysis, an MMP-sensitive peptide
sequence (GGVPMS↓ MRGGK, ↓ denotes cleavage point) was synthesized and incorpo-
rated within network crosslinks as described below. This peptide sequence is susceptible
to cleavage by numerous MMPs including collagenase-1 (MMP-1), gelatinase A (MMP-2)
and gelatinase B (MMP-9), which each play an important role in ECM remodeling.28 The
peptide sequence YRGDS was also synthesized and immobilized into the hydrogel matrix
to enable cell adhesion. Both peptides were synthesized using solid-phase peptide synthesis
based on standard Fmoc chemistry on a Focus Xi automated peptide synthesizer (AAPPTec,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 Louisville, KY). Amino acid coupling was carried out on a Wang resin in the presence of
11 DIEA and HBTU. The Fmoc group was deprotected with 20% piperidine in DMF. Peptides
12 were cleaved from the resin and deprotected in a TFA cleavage cocktail (90% TFA, 2.5% TIS,
13 2.5% thioanisole, 2.5% (w/v) phenol and 2.5% deionized water) for 2.5 hr, precipitated in cold
14 diethyl ether and purified by reverse-phase high-perfomance liquid chromatography (HPLC).
15 Peptide purity was confirmed by ion trap time-of-flight (IT-TOF) mass spectroscopy. The
16 HPLC chromatograms and IT-TOF mass spectra for both peptides are included in the Sup-
17 porting Information (Figures S3 and S4.) Purified peptides were lyophilized and stored at
◦
18 -20 C until use.
19 Synthesis of Difunctional and Monofunctional Photopolymerizable
20 PEG Macromers
21 Three types of photopolymerizable macromers were synthesized and used to create cell adhe-
22 sive and proteolytically degradable hydrogel scaffolds with tunable homogeneous and gradi-
23 ent properties. The synthesized macromers included: (1) an MMP-sensitive PEG diacrylate
24 crosslinker (MMP-sensitive PEGDA) to render network crosslinks susceptible to proteolytic
25 degradation, (2) an RGD containing PEG monoacrylate macromer (RGD-PEGMA) to in-
6
ACS Paragon Plus Environment

Page 7 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
1
corporate tethered cell-adhesive peptide ligands in the hydrogel and (3) a non-biofunctional
PEG monoacrylate macromer (PEGMA) produced with molecular weight similar to RGD-
PEGMA used as a surrogate during polymerization. The PEGMA macromer was used to
control hydrogel properties (elastic modulus, degradation and immobilized RGD concentra-
tion) without inducing variations in polymerization kinetics during the gradient fabrication
process as described below.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The MMP-sensitive peptide (GGVPMS↓ MRGGK) was conjugated to Acryl-PEG5000-
SVA (Laysan Bio, Arab, AL, MW = 5kDa) in a 2:1 PEG to peptide molar ratio. This resulted
in the formation of an MMP-sensitive PEGDA difunctional crosslinker Acryl-PEG5000-
10 GGVPMS↓MRGGK- PEG5000-Acryl (MW ∼11kDa). The cell adhesive peptide YRGDS
11 was conjugated to Acrylate-PEG3400-SVA in a 1:1 PEG to peptide molar ratio yielding
12 a photopolymerizable RGD-PEGMA monofunctional macromer, Acryl-PEG3400-YRGDS
13 (MW ∼ 4kDa). A tyrosine residue (Y) was added to the peptide sequence to track its
14 incorporated concentration in the scaffolds via radiolabeling with 125I as described below.
15 Peptides conjugation reactions were carried out in 50mM NaHCO3 solution (pH 8.0) for 4
16 hours while protected from light. The conjugated solution was then dialyzed for 24 hours to
◦
17 removed unreacted reagents, lyophilized, and stored at -20 C until use.
18
The unmodified PEG monoacrylate macromer, PEGMA (MW =5kDa), was synthe-
19 sized through acrylation of monomethoxy-PEG (MW 5kDa) with a 2-fold molar excess of
20 acryloyl chloride relative to free hydroxyls in the presence of trimethylamine in anhydrous
21 dichloromethane under argon overnight in the dark. The resulting product was separated
22 from aqueous byproducts by the addition of 2M K2CO3 and collection of the organic phase
23 in a separatory funnel. The final PEGMA solution was precipitated in ice cold ether, fil-
24 tered, dried under vacuum overnight, and stored at -20 ◦C until use. Acrylation efficiency of
1
25 PEGMA was confirmed to be above 95% using H-NMR.
7
ACS Paragon Plus Environment

Biomacromolecules
Page 8 of 42
1
2
3
4
1
Fabrication of Uniform Hydrogel Scaffolds
5
6
7
8
9
2
3
4
5
6
7
8
9
To create scaffolds with tunable gradients of embedded biochemical composition and me-
chanical properties, polymerization experiments involving the synthesis of hydrogel scaffolds
with uniform properties were first executed to screen formulations that decoupled variation
in immobilized RGD concentration and elastic modulus. Uniform hydrogel scaffolds were
formed using visible light free-radical photopolymerization as described in our previously
published studies.29 Precursor solutions consisted of 37mM N-vinyl pyrrolidone (an acceler-
ator and comonomer), 225mM TEA (co-initiator), 0.05mM eosin Y (photo-sensitizer) and
varying concentrations of the photopolymerizable macromers (RGD-PEGMA, PEGMA and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 MMP-sensitive PEGDA) dissolved in 10mM HEPES buffered saline (HBS; 10 mM HEPES
11 sodium salt and 137 mM NaCl) with pH=7.4 adjusted by addition of HCl. Modulation in
12 modulus were achieved through adjustments in the concentration of MMP-sensitive PEGDA
13 (1.5, 2 and 2.5 mM) with fixed concentration of RGD-PEGMA (2mM). To induce variations
14 in immobilized RGD concentration without imposing variations in scaffold modulus, the
15 molar ratio of RGD-PEGMA to PEGMA (25:75, 50:50 and 75:25) was adjusted while the
16 total concertation of PEG monoacrylate species (RGD-PEGMA and PEGMA) and MMP-
17 sensitive PEGDA macromers was fixed at 2mM and 1.5mM, respectively. Hydrogels were
18 formed by placing 100 µl of the precursor solution into a well of a 96-well plate followed
19 by photopolymerization with visible light ( λ=514nM) for 5 min using an Argon Ion Laser
20 (Coherent, Inc., Santa Clara, CA) at a laser flux of 100 mW/cm2. After polymerization,
21 hydrogels were rinsed with HBS and swollen to equilibrium prior to quantifying their me-
22 chanical and physical properties and embedded biochemical composition.
23 Fabrication of Gradient Hydrogel Scaffolds
24 A schematic of the gradient hydrogel scaffold frontal photopolymerization fabrication process
25 is shown in Figure 1. Hydrogel scaffolds containing gradinets in modulus or immobilized RGD
26 were created by mixing two distinct precursor solutions (precursor A and precursor B defined
8
ACS Paragon Plus Environment

Page 9 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
1
in Table 1), consisting of 37mM of NVP (accelerator and comonomer), 225mM TEA (co-
initiator), 0.05mM eosin Y (photo-initiator) with varying concentrations of RGD-PEGMA,
PEGMA and MMP-sensitive PEGDA dissolved in HBS(pH=7.4). The total macromer com-
position was varied in a manner that allowed us to decouple gradients of immobilized RGD
concentration from elastic modulus as well as modulate gradient characteristics (i.e. magni-
tude and slope of the immobilized RGD gradient). Four gradient scaffolds were created using
this approach: (1)scaffolds with an elastic modulus gradient and a uniform immobilized RGD
concentration, and scaffolds with uniform elastic modulus containing (2) steep, (3) interme-
diate, and (4) shallow RGD gradient slopes. To create scaffolds with an elastic modulus
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 gradient, two distinct precursor concentrations of the MMP-sensitive PEGDA crosslinker
11 were used (2mM for precursor A and 1.5mM for precursor B) containing an equivalent con-
12 centration of RGD-PEGMA in precursor A and precursor B. Sscaffolds containing an RGD
13 gradient and uniform elastic modulus were created by polymerizing precursor A and B with
14 a the concentration of the MMP sensitive PEGDA crosslinker while the concentrations of
15 RGD-PEGMA and PEGMA were varied in each precursor (A and B) to induce variations
16 in gradient slope. Specifically, the RGD-PEGMA precursor concentration was adjusted to
17 2mM, 1mM and 0.5mM in Precursor A and kept at 0mM in Precursor B, while the con-
18 centration of PEGMA was varied from 0mM, 1mM and 1.5mM in Precursor A and kept at
19 2mM in Precursor B targeting steep, intermediate, and shallow RGD gradients, respectively.
20 Table1 summarizes the precursor composition used to create the different types of gradient
21 scaffolds.
22
The targeted gradients were formed as a result of continuous feeding of an ascending pho-
23 topolymerizable reaction front with imposed changes in macromer composition over time.
24 This was achieved by perfusion of distinct precursor solutions whose delivery was controlled
25 using two programmable syringe pumps (NE-1000, New Era Pump System Inc., Farming-
26 dale, NY) shown in Figure1A. The flow rates of each precursor solution were programmed
27 to be linearly increasing form 0 µl/min to 120 µl/min (Precursor A) or decreasing from 120
9
ACS Paragon Plus Environment

Biomacromolecules
Page 10 of 42
1
2
3
4
Table 1: Precursor compositions for gradient hydrogel fabrication
5
6
7
8
Gradient
Group
modulus gradient (uni-
Scaffold
MMP-sensitive
PEGDA (µM)
PEGMA-RGD
(µM)
PEGMA
(µM)
Pump
A
B
A
B
2
2
2
2
0
0
0
0
2
9
form RGD)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.5
1.5
1.5
steep RGD Gradient
(uniform modulus)
intermediate RGD gra-
dient (uniform modu-
lus)
A
1.5
1
1
B
A
B
1.5
1.5
1.5
0
2
shallow RGD gradient
(uniform modulus)
0.5
2
1.5
0
1
2
3
4
5
6
7
8
9
µl/min to 0 µl/min (Precursor B) over a period of 5 min (Figure1B). Subsequently precur-
sor solutions were mixed and dripped into a transparent reaction chamber (5mm x 10mm x
20mm) and immediately photopolymerized by exposure to visible light ( λ=514 nm) using an
Argon Ion Laser (Coherent, Inc., Santa Clara, CA) at a laser flux of 100mW/cm2. Hydrogel
crosslinking occured continuously with the ascending reaction front as precursor solutions
were fed to the reaction vessel. While the syringe pumps were programmed to deliver precur-
sor for 5 min, the photopolymerization was carried out for a total of 10 min to ensure that
every region of the precursor was photopolymerized for at least 5 min. Following photopoly-
merization, gradient hydrogels were removed from the reaction vessel, soaked in HBS for 48
10 hours to allow for equilibrium swelling, and then cut into seven 2mm thick sections using an
11 array of razor blades. The first and last sections were discarded (to avoid chracterization of
12 resulting edge effects,) and the remaining five sections were were used to quantify variation
13 in elastic modulus, degradation kinetics and immobilized RGD concentration as described
14 below (Figure1C).
10
ACS Paragon Plus Environment

Page 11 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1: Gradient Hydrogel Fabrication, Charaterization, and Quantification
of 3D Vascular Sprouting in Response to Imposed Gradients (A) Schematic of
gradient hydrogel fabrication process of photo-frontal polymerization. (B) Flow rate profiles
of precursor solutions using two programmable syringe pumps to control their delivery during
crosslinking. (C) Schematic representation of hydrogel sectioning and spheroid placement in
gradient scaffolds for quantifying spatial variations of physical and mechanical properties,
biochemical composition and vascular sprouting. (D) Representation of methodology used to
quantify directional vascular sprouting parameters in responses to the embedded gradients.
Vascular sprout response towards high RGD or modulus (90◦), low RGD or modulus (270◦)
or in directions perpendicular to the imposed RGD or modulus gradients (0◦, 180◦).
11
ACS Paragon Plus Environment

Biomacromolecules
Page 12 of 42
1
2
3
4
1
Fabrication of Fluorecent Gradient Hydrogel Scaffolds
5
6
7
8
9
2
3
4
5
6
7
8
9
A near-infrared fluorophore, indocyanine green (ICG, MW=774.96 g/mol), was physically
entrapped in the scaffolds during crosslinking to visualize and confirm the presence of a
gradient through spatial variations in fluorescence. ICG was selected because its fluorescence
spectrum (Ex 780nm/Em830nm) does not overlap with that of the photo-initiator eosin
Y (Ex 524nm/Em544nm) used during the photopolymerization. To create scaffolds with
gradients of fluorescence, 0.1 mM ICG was included in precursor A. Concentration gradients
of soluble ICG in the scaffolds were imaged immediately after hydrogel formation using a
Pearl near-infrared fluorescent and bioluminescent small animal imaging system (LI-COR,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 Biosciences). Fluorescence intensity gradients were quantified by analyzing the images with
11 MATLAB (R2017a).
12 Quantification of Elastic Modulus Gradients
13 Spatial variations in elastic modulus were quantified using compression experiments. Prior
14 to mechanical testing, hydrogel sections were swollen in HBS for 48 hours with three buffer
15 changes. Scaffolds sections were loaded onto a TA RSA3 mechanical tester (TA Instruments)
16 controlled by TA Orchestrator software. Samples were compressed at a rate of 0.5 mm/min
17 and the elastic modulus of each section was quantified from the slope of the linear region
18 (<10% strain, r2 >0.95) of the stress-strain curve (provided in the supporting information)
19 as previously reported.26). The elastic modulus of scaffolds with uniform properties and
20 biochemical composition was similarly quantified.
21 Quantification of Immobilized RGD concentration Gradients
22 Immobilized RGD concentration in scaffolds was quantified by radiolabeling tyrosine residues
23 (Y) with 125I prior to gel formation according to previously reported techniques.26 Briefly,
24 five mg of YRGDS was dissolved in reaction buffer (1x PBS, pH=6.5), combined with 1mCi
12
ACS Paragon Plus Environment

Page 13 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1
sodium iodine (Perkin-Elmer Health Sciences) in the presence of iodination beads (Pierce)
and allowed to react for 15 min. Radiolabeled RGD (125I-YRGDS) was conjugated to Acryl-
PEG3400-SVA as described above to generate a radiolabeled version of the RGD-PEGMA
macromer. To track the spatial incorporation of the YRGDS adhesion ligand in gradient
scaffolds and in uniform hydrogel controls, radiolabeled Acryl-PEG3400-YRGDS was added
to the precursor in an amount not exceeding 20 wt % of the total peptide. In the case of
gradient scaffolds, hydrogels were sectioned as described above, and sections were swollen in
HBS for 48 hours. Following washing, the radioactivity of each section was measured using
a Packard Cobra II gamma counter (Perkin-Elmer Health Sciences). The immobilized RGD
2
3
4
5
6
7
8
19
20
21
9
10 concentration in the scaffold [RGD]gel was calculated according to following the equation:
R × SA
(1)
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[RGD]gel
=
mswollen − mdried mdried
+
ρwater
ρPEG
11
where R is the measured radioactivity, SA is the specific activity (mole/Ci) that was
12 measured and used to convert the measured radioactivity to moles of peptide, mswollen is
13 the equilibrium the swollen weight of the hydrogel, mdried is the weight of the hydrogel in
14 the dried state, ρwater is density of water at room temperature (0.997 g/ml) and ρPEG is the
15 density of PEG (MW=10 kDa) at room temperature (1.3 g/ml).
16 Quantification of Gradients in Proteolytic Scaffold Degradation
17 Spatial variations in scaffold degradation were quantified from gravimetric measurements in
18 the wet weight of hydrogel sections during incubations in interstitial collagnease (Collagenase-
19 1A, MMP-1,Sigma Aldrich) in 10mM HBS with 1mM CaCl2 (pH =7.4) at a fixed concen-
20 tration of 1 µg/mL. Gradient hydrogel sections were incubated in collagenase degradation
21 buffer solutions at 37◦C over a desired time period, the enzyme solution was removed, and the
22 swollen weight of each section was recorded every hour until complete hydrogel degradation
23 was achieved.
13
ACS Paragon Plus Environment

Biomacromolecules
Page 14 of 42
1
2
3
4
5
6
1
2
Evaluation of 3D Vascular Sprout Formation in Response to Scaffold
Embedded Gradients
7
8
9
3
4
5
6
7
8
9
The effects of the scaffold embedded gradients on neovascularization was evaluated using a
3D co-culture spheroid sprouting assay of sprouting angiogenesis.14 This assay was chosen
because it recapitulates the multicellular and coordinated process of neovascularization in
vivo. Cell spheroids composed of 50% human umbilical vascular endothelial cells (HUVECs,
Lonza) and 50% human umbilical arterial smooth muscle cells (HUASMCs, Lonza) were
formed by suspending both cells types (5000 total cells/well) in culture media containing
0.24% (w/v) methyl cellulose in round-bottom, low binding, 96 well-plates and incubated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 overnight at 37◦C and 5% CO2. Following scaffold fabrication, and swelling for 48 hours in
11 EGM-2, cell spheroids were gently placed inside the scaffolds using a micropipette tip. In
12 the case of gradient scaffolds, aggregates were positioned in three distinct regions along the
13 gradient with a maximum of two spheroids per regions spaced at least 8mm apart. Each
14 region presented distinct scaffold properties based on the type of gradient. In the case of
15 scaffolds containing gradients in elastic modulus, spheroids were seeded in high, intermediate
16 ,or low modulus regions. In the case of RGD gradient scaffolds, spheroids were seeded in
17 regions containing high, intermediate ,and low immobilized concentrations of the cell adhe-
18 sion ligand (Figure 1). Spheroids were imaged using an Axiovert 200 inverted microscope
19 (Carl Zeiss Inc.) at days 1, 3, 5, and 7. Culture media (EGM-2) was changed every other
20 day. Phase contrast images were acquired from days 1 to 7 and used to quantify spheroid
21 sprout invasion over time in terms of projected spheroid area as well as the directionality of
22 vascular sprouting by measuring the length of individual sprouts at day 7 using Axiovision
23 4.2 image analysis software (Carl Zeiss Inc., GÃűttingen, Germany). The projected area and
24 individual sprout was traced manually for each image with acquired measurements blinded
25 to the samples. Examples of analyzed images are provided in the Supporting Information
26 (Figure S1). Specifically, the image field centered on the spheroid was divided into four
27 directional sectors as shown in Figure 1, where the 90◦ and 270◦ angles represent the direc-
14
ACS Paragon Plus Environment

Page 15 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
tions of increasing and decreasing elastic modulus and RGD concentration along the gradient
direction, respectively, while the 0◦ and 180◦ angles correspond to directions perpendicular
to the imposed gradients. The analysis of directionality was evaluated by measuring the
average length and number of sprouts at each of these angles (Figure 1D).
2
3
4
5
Results
17
18
19
20
21
6
Characterization of Uniform Hydrogel Scaffolds with Tunable Elastic
Modulus and Immobilized Concentration of RGD Peptide Ligands
7
22
23
24
25
8
Prior to creating scaffolds with tunable gradients of mechanical properties and immobilized
RGD, studies involving the formation of scaffolds that presented these factors uniformly
9
26
10 (isotropic scaffolds) were performed to determine the polymerization conditions for decou-
11 pling these matrix properties. The elastic modulus and immobilized RGD concentration in
12 isotropic hydrogel scaffolds were individually tuned by adjusting the precursor concentration
13 of MMP-sensitive PEGDA, RGD-PEGMA, and PEGMA. Adjustments in the precursor con-
14 centration of MMP-sensitive PEGDA from 1.5, 2, 2.5 mM resulted in variations in elastic
15 modulus of 1.2±0.05 kPa, 3.2±0.02 kPa, and 6.0 ±0.07kPa (Figure 2A) respectively, while
16 the immobilized concentration of RGD remained constant (1.39±0.05 mM, Figure 2B). Ad-
17 justments in the molar ratio of RGD-PEGMA to PEGMA (25:75, 50:50,75:25) while keeping
18 the total concentration of PEG monoacrylates (RGD-PEGMA and PEGMA) fixed (2mM)
19 and the concentration of the MMP-sensitive PEGDA (1.5 mM) constant led to variations in
20 immobilized RGD concentration of 0.33±.02 mM, 0.61±0.04 mM, and 0.87±0.08mM, respec-
21 tively, (Figure 2D) while the scaffold elastic modulus was fixed at 1.2±0.08 Pa (Figure 2C).
22 It is worth noting that the immobilized RGD concentration is lower than the concentration
23 of RGD-PEGMA in the precursor solution because the RGD concentration in the scaffold is
24 measured in the fully swollen state and the hydrogels typically result in a 10% increase in
25 gel volume from the relaxed state. The immobilized RGD concentration in the fully swollen
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
ACS Paragon Plus Environment

Biomacromolecules
Page 16 of 42
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
gel is calculated using equation 1, which considers the ratio of moles of RGD incorporated
and the volume of the swollen gel which contributes to an observed decrease of incorporated
RGD. Furthermore, the presence of multiple monomers (PEGDA, RGD-PEGMA and NVP)
in this copolymerization system influence the cross-propagation kinetics which impact the
amount of monomers incorporated into the crosslinked network. Regardless, these results
indicate that variation in hydrogel modulus and immobilized RGD concentration may be de-
coupled by controlling crosslinker concentration and the RGD-PEGMA and PEGMA molar
ratio in the precursor.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
B
8000
6000
4000
2000
0
2.0
1.5
1.0
0.5
0.0
*
*
*
1.5 mM
2 mM
2.5 mM
1.5 mM
2 mM
2.5 mM
PEGDA Concentration
PEGDA Concentration
C
D
*
*
2000
1500
1000
500
0
1.0
0.8
0.6
0.4
0.2
0.0
*
25:75
50:50
75:25
25:75
50:50
75:25
PEGMA-RGD to PEGMA Ratio
PEGMA-RGD to PEGMA Ratio
Figure 2: Characterization of isotropic hydrogel scaffolds with decoupled vari-
ations in immobilized RGD concentration and elastic modulus. Variations in
crosslinker (PEGDA) concentration results in changes in (A) elastic modulus without induc-
ing variations in (B) immobilized RGD concentration. Variations in the ratio of PEGMA-
RGD to PEGMA at fixed PEG monoacrylate concentration does not induce alterations in
(C) elastic modulus but results in variation in (D) immobilized RGD concentration in the
scaffolds. Asterisk (*) indicates statistical significance with p < 0.05 (n = 4).
16
ACS Paragon Plus Environment

Page 17 of 42
Biomacromolecules
1
2
3
4
5
6
7
1
Characterization of Hydrogel Formation via Photo-frontal Polymer-
ization
2
8
9
10
11
12
13
14
15
16
17
18
19
20
21
3
To fabricate gradient scaffolds using photo-frontal polymerization, the flow rate of the feed
stream was carefully selected to match the propagation speed of the polymerization front
while maintaining a constant thin liquid layer of precursor solution on the upper surface
of the propagating polymerization front (Supplementary Video 1). When higher flow rates
were selected, convection in the liquid precursor resulted in lack of control of desired gra-
dients while lower flow rates led to discrete spatial variations in scaffold properties rather
than formation of continuous gradients (data not shown). The process of hydrogel formation
4
5
6
7
8
9
22
10 was characterized by terminating the polymerization at various light exposure times ranging
11 from 1 to 10 min after which the thickness of the formed gel as well as the liquid precursor
12 layer were quantified. Figure 3A displays a series of hydrogels produced by photo-frontal
13 polymerization at various exposure times. The thickness of the hydrogels is observed to
14 increase with time and to eventually reached a plateau after 6 min (Figure 3B). The rate of
15 increase in gel thickness, related to the propagation speed of the polymerization front, follow-
16 ing an exposure time of 6 min was determined to be 2.36±0.08 mm/min, which corresponds
17 to the programmed rate of the rising precursor liquid level (2.4mm/min). The thickness of
18 the liquid precursor layer was calculated by subtracting the gel thickness from the overall
19 thickness. Figure 3C shows that the precursor layer thickness is maintained at 0.65± 0.23
20 mm for first five minutes and is completely depleted after 6 min of light exposure. Overall
21 these results confirm that the selected flow rate in the feed stream (120 µl/min) sustains
22 the propagation of the polymerization front and maintains a thin precursor layer to ensure
23 proper gradient generation.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment

Biomacromolecules
Page 18 of 42
1
2
3
4
5
A
B
C
6
7
8
9
15
10
5
1.5
1.0
0.5
0.0
&"*)+'"*)+ ("*)+ ".*)+
%"*)+
$"*)+
#"*)+
""*)+
!"*)+
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
"*)+
0
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
Time (min)
Time (min)
Figure 3: Characterization of photo-frontal polymerization used to create gradient
hydrogel scaffolds. (A) Hydrogels formed at various exposure time. (B) Changes in
scaffold height as a function of visible light exposure times. (C) Variations in precursor
thickness over light exposure time.
1
2
Visualization of Continuous Gradient Formation during Ascending
Photo-frontal Polymerization
3
4
5
6
7
8
9
To verify that continuous gradients could be created using our proposed methodology, the
near-infrared fluorophore, ICG, was included in the precursor solution fed by pump A at
a concentration of 0.1 mM to generate an ICG gradient ranging from 0 to 0.1 mM along
the gradient direction (y-axis). A continuous gradient of fluorescence intensity was visual-
ized along the y-axis with minimal variation along the x-axis within the scaffold following
photopolymerization as demonstrated in Figure 4A. Acquired images were analyzed using
MATLAB to quantify spatial variations in fluorescence intensity within scaffolds. The data
10 in Figure 4B indicate gradual decreases in fluorescence intensity with increasing distance
11 from the region at which precursor was continuously fed to the polymerization front. These
12 results indicate that controlled perfusion of polymeric precursors during free-radical frontal
13 photopolymerization can be used to create continuous and specific gradients within hydrogel
14 scaffolds along the vertical (y-axis) direction.
18
ACS Paragon Plus Environment

Page 19 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
A
y
B
0.6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
0.4
0.2
0.0
0.0
0.4
0.8
1.2
x
Length (mm)
Figure 4: Visualization of gradient formation. (A) Fluorescent image of a hydrogel
scaffold embedded with concentration gradient of ICG. (B) Fluorescence intensity change
along the gradient direction.
1
Characterization of Hydrogel Scaffolds with Elastic Modulus Gradi-
ents
27
28
29
30
31
2
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
4
5
6
7
8
9
Scaffolds with gradients of elastic modulus were created by gradually changing the concen-
tration of the MMP-sensitive PEGDA crosslinker in the precursor feed stream while keeping
the feed concentration of RGD-PEGMA constant during photopolymerization. As shown
in Figure 5A, this resulted in the formation of hydrogel scaffolds with linearly increasing
gradients of elastic modulus from 660 Pa to 1460 Pa over a span of 10mm, with a gradient
slope estimated to be 80 ± 5 Pa/mm. Quantification of spatial variaiton in the immobilized
RGD revealed that the concentration of cell adhesive peptide remained uniform (at 2.06 ±
10 0.12mM) in modulus gradient scaffolds (p=0.996).
11
Gradient hydrogel scaffolds were also rendered proteolytically degradable by including
12 MMP-sensitive peptide sequences within network crosslinks. Since gradients in elastic mod-
13 ulus were induced as a result of gradients in crosslink density due to adjustments in the com-
14 position of the MMP-sensitive PEGDA crosslinker in the precursor, the resultant scaffolds
15 also presented spatial variations in MMP-sensitive crosslinked sites available for enzymatic
19
ACS Paragon Plus Environment

Biomacromolecules
Page 20 of 42
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
cleavage. This was confirmed by quantifying spatial variations in proteolytic degradation ki-
netics obtained by gravimetric measurements in the wet weight of hydrogel sections (2mm)
over time in collagenase incubation. As shown in Figure 5B, scaffolds containing gradients
of elastic modulus exhibited spatial variations in degradation kinetic profiles. The time
required for complete degradation progressively increased with increasing elastic modulus
and/or crosslink density. Hydrogel sections with the lowest modulus (0-2 mm) exhibited the
shortest time for complete degradation at 4 hours, while sections with intermediate elastic
modulus (4-6 mm) required 6 hours for complete degradation, and sections possessing the
highest elastic modulus (8-10 mm) degraded the slowest with complete degradation occurring
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 at 8 hours.
A
B
140
120
100
80
0-2 mm
2-4 mm
4-6 mm
6-8 mm
8-10 mm
6
2000
1500
1000
500
0
RGD Concentration
Elastic Modulus
4
2
60
40
20
0
0
0
2
4
6
8
10
12
0
2
4
6
8
Length (mm)
Time (hrs)
Figure 5: Characterization of hydrogel scaffolds with gradients of elastic modulus
and proteolytic degradation. (A) Scaffolds with spatial variations of elastic modulus and
uniform immobilized RGD concentration. (B) Spatial variations in hydrogel degradation
kinetics in collagenase enzyme solution (1 µg/mL).
11 Characterization of of Hydrogel Scaffolds with Immobilized RGD
12 Concentration Gradients
13 RGD gradient hydrogel scaffolds were created by gradually changing the molar ratio of RGD-
14 PEGMA and PEGMA while keeping the total concentration of the monoacrylate macromers
15 and the crosslinker concentration constant in the precursor solution fed into the reaction
20
ACS Paragon Plus Environment

Page 21 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1
vessel over time. The immobilized RGD concentration gradient linearly increased 205%
from 0.48mM to 0.98 mM over a span of 10 mm, resulting in a gradient slope of 58.8±4.8
µM/mm while the elastic modulus remained relatively uniform at 597±22Pa (Figure 6A).
Quantification in proteolytic degradation kinetics along the RGD gradient scaffolds revealed
similar degradation profiles and time for complete degradation (4 hours, Figure 6B). This
was expected since scaffolds of uniform crosslink density possess crosslinks susceptible to
proteolytic degradation thereby resulting in uniform degradation kinetics. Based on our
observed results with uniform hydrogel scaffolds, changes in the molar ratio of RGD-PEGMA
and PEGMA induced RGD concentration gradients without inducing spatial variations in
2
3
4
5
6
7
8
19
20
21
9
10 crosslink density and/or elastic modulus and consequently proteolytic degradation kinetics.
Characteristics in gradients of biochemical composition, including magnitude and slope,
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
12 have also been found to be important in controlling cell responses.11 Therefore, we created
13 hydrogel scaffolds where the slope of the immobilized RGD concentration gradient was var-
14 ied while the modulus remained uniform. This was achieved by adjusting the concentration
15 range of RGD-PEGMA in the precursor feed stream during gradient hydrogel formation, as
16 indicated in Table 1. The RGD-PEGMA concentration ranges were varied to create shallow,
17 intermediate, and steep gradients while the total concentration of PEGMA species was kept
18 constant at 2mM by supplementing with PEGMA surrogate in the precursor to maintain
19 the same crosslink density (and elastic modulus) in each case. As shown in Figure 7, all
20 three RGD gradient hydrogels exhibited a linearly increasing immobilized RGD concentra-
21 tion with distinct slopes at 9.0±2.1 µM/mm (shallow), 32.7±5.6 µM/mm (intermediate) and
22 58.8±4.8 µM/mm (steep). These results demonstrate that the slope of the RGD gradient
23 can be modulated without inducing variations in scaffold elastic modulus and degradation
24 kinetics by varying the RGD-PEGMA precursor concentration while keeping total concentra-
25 tion of total PEG monoacrylate macromers (RGD-PEGMA and PEGMA) constant during
26 polymerization.
21
ACS Paragon Plus Environment

Biomacromolecules
Page 22 of 42
1
2
3
4
5
6
7
8
A
B
140
0-2 mm
2-4 mm
4-6 mm
6-8 mm
8-10 mm
9
1.5
1.0
0.5
0.0
2000
RGD Concentration
Elastic Modulus
120
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
100
1500
80
1000
60
40
500
20
0
12
0
0
2
4
6
8
10
0
2
4
6
8
Length (mm)
Time (hrs)
Figure 6: Characterization of hydrogel scaffolds with gradients in immobilized
RGD concentration (A) Scaffolds with spatial variations in immobilized RGD concen-
tration and uniform elastic modulus and (B) Uniform degradation kinetics in collagenase
incubation (1µg/mL).
1400
Shallow (9 µM/mm)
1200
Intermediate (34 µM/mm)
1000
Steep (59 µM/mm)
800
600
400
200
0
0
2
4
6
8
10
12
Length (mm)
Figure 7: Characterization of scaffolds with varying slope and magnitude of im-
mobilized RGD concentration.
22
ACS Paragon Plus Environment

Page 23 of 42
Biomacromolecules
1
2
3
4
5
6
7
1
Quantification of 3D Vascular Sprout Formation in Gradient Hydro-
gel Scaffolds
2
8
9
10
11
12
13
14
15
16
17
18
19
20
21
3
A co-culture model of sprouting angiogenesis was used to investigate the effect of the imposed
modulus or immobilized RGD gradients on 3D vascular sprout formation. Cell spheroids
were placed into gradient scaffolds at three distinct regions along the imposed gradients. In
these specific regions, the cells were exposed to different gradient magnitudes but the same
gradient slope. In scaffolds containing a modulus gradient of 80 ±5 Pa/mm and a uniform
immobilized RGD concentration ( ∼2mM) spheroids were placed in three regions along the
gradient, with each region composed of a distinct modulus; (1) a high elastic modulus region
4
5
6
7
8
9
22
10 (∼1300Pa), an intermediate modulus region( ∼1000 Pa), and a low modulus region(∼700
11 Pa). Among the three types of RGD gradient scaffolds, vascular cell response was evaluated
12 only in the case of a steep RGD gradient (58.8 ±4.8µM/mm with spheroids placed in regions
13 containing high (∼900 µM), intermediate (∼700 µM), and low (∼500 µM), immobilized RGD
14 concentrations.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
To evaluate the effects of specific gradients and gradient characteristics on 3D neovascu-
16 larization, we quantified in vitro vascular sprout invasion over time from acquired brightfield
17 images of cell spheroids within the different scaffold regions at days 1, 3, 5, 7 (Figure 8) .
18 As shown in Figure 9, the normalized invasion area in both elastic modulus and RGD gra-
19 dient increased progressively over time with up to a 5-fold increase at day 7. Cell spheroids
20 seeded in the low elastic modulus region exhibited significantly higher sprout invasion area
21 as compared to the high modulus region by days 5 and 7 (Figure 9A). However, no statistical
22 significant differences in invasion areas among the three modulus regions were observed at
23 earlier time points. Differences in invasion area were not found to be statistically signifi-
24 cant when spheroids were placed in different regions (different distances from the imposed
25 gradient) within the RGD gradient scaffolds (Figure 9B).
26
Immobilized or soluble gradients of biochemical factors have been previously reported to
27 result in sprout polarization in the direction towards the imposed gradients.8,30 To determine
23
ACS Paragon Plus Environment

Biomacromolecules
Page 24 of 42
1
2
3
4
5
6
7
8
9
A
D
E
F
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
500µm
B
C
Figure 8: Vascular sprout invasion in gradient hydrogel scaffolds. Day 7 repre-
sentative brightfield images of vascular sprout formation of spheroids embedded in different
regions of scaffolds containing an elastic modulus gradient (A) high, (B) intermediate and
(C) low modulus regions, and in scaffolds containing an immobilized RGD gradient (D) high
RGD, (E) intermediate RGD, and (F) low RGD concentration regions.
24
ACS Paragon Plus Environment

Page 25 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
A
B
8
6
4
2
0
8
High Modulus
High RGD
*
Intermediate Modulus
Low Modulus
Intermediate RGD
Low RGD
6
4
2
0
*
Day1
Day3
Day5
Day7
Day1
Day3
Day5
Day7
Figure 9: Kinetics in 3D Vascular Sprout Invasion Area. Projected area of vascular
sprouts in (A) modulus gradient hydrogels and (B) RGD gradient hydrogels. * indicates
statistical significance with p < 0.05 (n = 8).
1
2
3
4
5
6
7
8
9
whether the imposed gradients result in guided vascular sprout responses, the average sprout
length and sprout number at day 7 were quantified along 4 directions: parallel and towards
gradient ( θ= 90◦), parallel and opposing gradient ( θ= 270◦) and in direction perpendicular
to the gradients ( θ= 0◦ and 180◦). As shown in Figure 10A, in scaffolds containing gradi-
ents of elastic modulus, the length of vascular sprouts was found to be significantly higher
in regions of low modulus (691±167 µm) towards the opposing gradient direction (towards
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
◦
◦
low crosslink density, θ= 270 ), as compared to the sprout length at 90 (494±76 µm).
However, in gradient scaffold regions of high and intermediate modulus, no statistical sig-
nificant differences in directional vascular sprout length was found to occur. In terms of the
10 number of vascular sprouts, however, significant increases in the opposing gradient direction
◦
11 ( θ= 270 ) were noted as compared to all other directions. This result was found to be
12 consistent in all regions of the scaffold (high, intermediate, and low modulus, Figure 10B).
◦
13 Furthermore, the number of sprouts towards the gradient ( θ = 90 ) was also found to be
14 significantly higher as compared to the perpendicular directions ( θ=0 ◦, 180 ◦). In summary,
15 these results indicate that gradients in mechanical properties of the matrix influence vascular
16 sprout formation with increased vascular sprout polarization and sprout length towards the
17 opposing gradient direction. These results are consistent with previous studies that indicate
18 that sprout formation and migration decrease with increase in matrix density as additional
25
ACS Paragon Plus Environment

Biomacromolecules
Page 26 of 42
1
2
3
4
1
proteolytic activity is required for 3D matrix remodeling.29–31
5
6
7
8
A
B
*
*
*
*
*
*
1000
20
15
10
5
*
*
*
*
*
*
0°
0°
*
9
800
600
400
200
0
90°
180°
270°
90°
180°
270°
*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
High
Modulus
Region
Intermediate
Modulus
Region
Low
Modulus
Region
High
Modulus
Region
Intermediate
Modulus
Region
Low
Modulus
Region
Figure 10: Directional Analysis of Vascular Sprout Parameters in Response to
Gradients of Elastic Modulus. Quantification of directional (A) mean sprout length and
(B) number of sprouts in gradient scaffold regions containing high, intermediate and low
modulus at day 7. * indicates statistical significance with p < 0.05 (n = 8).
2
3
4
5
6
7
8
9
In contrast to the results of vascular sprouting in scaffolds embedded with gradients of
elastic modulus, opposite trends are observed in response to immobilized gradients of RGD
concentration. As shown in Figure 11A, when vascular cell spheroids were placed in high
RGD containing regions, significant increases in sprout length were found to occur towards
the gradient (689±77 µm, θ = 90 ◦) as compared to the opposing gradient direction (330±58
◦
◦
µm, θ= 270 ) and the perpendicular directions (448±101 µm , θ =0 , 415±111 µm, θ
◦
= 180 ). Similar finding were observed to occur in scaffold regions of intermediate RGD
concentration with significant increases in sprout length towards the imposed gradient (600
10 ± 55µm, θ = 90◦) as compared to opposing gradient direction (432±50 µm, θ= 270◦). In
11 scaffold regions of low RGD concentration, the average sprout length along the different di-
12 rections was not found to be statistically significant. In scaffold regions of intermediate RGD
13 concentration, the number of sprouts was significantly higher in the direction of the gradi-
14 ent (θ= 90◦) as compared to the perpendicular directions (θ= 0◦, and 180◦), (Figure 11B).
15 Interestingly, in regions of low immobilized RGD concentration, significant increases in the
16 number of sprouts were observed opposing (θ= 270◦) and towards (θ= 90◦)the gradient as
◦
◦
17 compared to the perpendicular directions (θ=0 , 180 ). In summary, these results demon-
26
ACS Paragon Plus Environment

Page 27 of 42
Biomacromolecules
1
2
3
4
5
6
1
strate that the imposed RGD gradient results in increases in sprout length and number when
spheroids are localized to regions of high and intermediate immobilized RGD concentration.
2
7
8
9
A
B
*
*
1000
*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
*
0°
*
20
*
0°
*
800
600
400
200
0
90°
180°
270°
*
*
90°
180°
270°
15
10
5
*
High
RGD
Region
Intermediate
RGD
Region
Low
RGD
Region
0
High
RGD
Region
Intermediate
RGD
Region
Low
RGD
Region
Figure 11: Directional Analysis of Vascular Sprout Parameters in Response to
Gradients of Immobilized RGD Concentration. Quantification of (A) mean sprout
length and (B) number of sprouts in gradient scaffold regions containing high, intermediate
and low immobilized RGD concentration at day 7. (* indicates statistical significance with
p < 0.05 (n = 8).
3
Discussion
33
34
35
4
5
6
7
8
9
A primary factor limiting the volume of tissue that can be engineered is the extent to which
stable blood vessels can be stimulated to form within the implantable scaffold. An ex-
tensive, rapid and stable blood supply is required to the meet mass transport demands
in newly formed engineered tissues. A potential solution is to induce rapid directional
ingrowth of a vascular network guided by spatiotemporal cues embedded within an im-
plantable 3D biodegradable scaffold. Significant progress has been made to elucidate the
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 role of growth factor gradients on neovascularization using microfluidics and biomaterials-
11 based platforms.6,16,31–33 In addition to growth factors, ECM composition, cell adhesion
12 ligands, as well as mechanical properties of the matrix play an important role in blood ves-
13 sel assembly.34,35 Gradient hydrogel scaffolds fabricated using natural, synthetic, or hybrid
14 polymer combinations embedded with spatial variations in cell adhesion ligand concentra-
27
ACS Paragon Plus Environment

Biomacromolecules
Page 28 of 42
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
tion or mechanical properties have been shown to lead to guided vascular cell migration on
2D surfaces.10,11,13,36 Limited studies, however, have attempted to elucidate the role of these
gradients in 3D culture or on vascularized tissue remodeling in vivo. Biomaterial platforms
that enable selective and precise control of matrix gradients and gradient characteristics
would provide significant insight on individual and combinatorial effects of multiple types of
scaffold gradients which are critical for engineering vascularized tissue formation for a variety
of damaged tissues.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In this study, we present a novel gradient hydrogel scaffold fabrication technique to create
proteolytically degradable PEG hydrogel scaffolds with tunable spatial variations in elastic
10 modulus and immobilized RGD concentration. The key feature of this technique is that it
11 allows for precise control of gradient generation due the inherently high rate of polymer-
12 ization of a photopolymerizable ascending reaction front. Unlike conventional crosslinking
13 methods, free-radical frontal photopolymerization enables immediate stabilization of concen-
14 tration gradients before diffusion acts to equilibrate the gradients, resulting in precise control
15 of gradient profiles. In our presented gradient scaffold fabrication approach, polymerization
16 is localized to the reaction front; a thin layer (<1 mm) of viscous prepolymer solution. Since
17 the reaction volume of the propagating front is maintained at low values during crosslinking,
18 the scaffold dimensions are not limited to small volumes as with other gradient fabrication
19 approaches. Furthermore, this approach can be used to produce different types of gradients
20 (immobilized biochemical and mechanical property gradients) with specific gradient charac-
21 teristics including magnitude, slope and shape. In order to demonstrate the feasibility of this
22 technique, we created scaffolds with decoupled variations in elastic modulus and immobilized
23 RGD concentration. Furthermore, modulation of the slope of the immobilized RGD gradi-
24 ent was induced without imposing changes in scaffold modulus and degradation kinetics. In
25 future studies, we plan to create scaffolds with more complex profiles (e.g. sigmoidal), which
26 are more physiologically relevant. In the present study our gradients were characterized on
27 the millimeter scale, which provides a reasonable estimate of linear gradient characteristics
28
ACS Paragon Plus Environment

Page 29 of 42
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1
(magnitude and slope). We acknowledge, however, that alternative methods such as atomic
force microscopy (AFM) will be required to more accurately analyze complex gradient pro-
files, or for evaluating vascular sprouting effects at the micron-length scale. Future efforts
focused on understanding the influence of gradient shapes on 3D cell responses will consider
the use of AFM to quantify micron-scale changes in elastic modulus using Hertz equations.15
Cell adhesion ligands as well as mechanical properties of the ECM have been shown to reg-
ulate neovascularization via mechanosensing pathways.37–39 To investigate the role of these
gradients on 3D vascular sprouting, gradient hydrogel scaffolds were created with decoupled
gradients of elastic modulus and cell adhesive RGD peptide ligands (Figures 5 and 6). Scaf-
2
3
4
5
6
7
8
19
20
21
9
10 folds were designed with gradients in modulus ranging from 500-1500 Pa based on previous
11 values reported to support neovascularization using isotropic scaffolds. For example, en-
12 dothelial cells have been shown to form capillary-like networks on compliant collagen-coated
13 polyacrylamide gels with elastic moduli ranging from 200 to 1000 Pa.40 Studies involving
14 3D cell culture models indicate that neovascularization occurs within compliant hydrogels
15 of collagen, MMP-sensitive PEG hydrogels, and in hyaluronic acid scaffolds with elastic
16 moduli from 175-515 Pa,41 500-3500 Pa,29,35 and 200-800 Pa,42 respectively. These stud-
17 ies indicate that the extent of 3D neovascularization typically increases with decreases in
18 matrix stiffness; however, when matrix stiffness was modulated independent of matrix den-
19 sity in collagen scaffolds, the extent of vascular invasion was found to be higher in stiffer
20 matrices.41,43,44 Nevertheless, matrix stiffness is one of the key material parameters that im-
21 pacts neovascularization. The gradient slope in modulus gradient scaffolds presented in our
22 study is approximately 80 Pa/mm. This value lies within the same order of magnitude of
23 modulus gradients (60 Pa/mm) previously reported to direct 3D neurite outgrowth,18 but
24 is much lower than those reported to guide 2D cell migration (75,000 - 1,000 Pa/mm).13,36
25 One limitation of our study is that we did not consider dynamic changes in modulus as a
26 result of proteolytic degradation. Previous studies have shown that temporal changes in
27 modulus are critial to 3D cell responses, such as invasion, proliferation and differentiation
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
29
ACS Paragon Plus Environment