Step-growth hydrogels crosslinked through grafted polypeptides enable nano- to macroscale synthetic
extracellular matrix design
ARCM ES
by
MASSACHUSETTS (NSTITUTE
OF TECHNOLOLGY
Caroline Marie Chopko
JUN 16 2015
B.S.E. Princeton University (2007)
LIBRARIES
Submitted to the Department of Chemical Engineering
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
February 2015
2014 Massachusetts Institute of Technology. All rights reserved.
Signature of author.................................................
Signature redacted
Departmen of Chemical Engineering
Octobe6r 16, 2014
Certified by................
S ignatu re redacted
.........................
Linda Griffith
S. E. T. I.Professor of
Certified by.......
................
ogical and Mechanical Engineering
Thesis Advisor
Si gnature redacted
auffenburger
V
Aug
Professor of Biological Engineering and Chemical Engineering
Thesis Advisor
A ccepted by ...............................................................
Signature redacted
Patrick S. Doyle
Professor of Chemical Engineering
Chairman, Committee for Graduate Students
1
Step-growth hydrogels crosslinked through grafted polypeptides enable nano- to macroscale
synthetic extracellular matrix design
by
Caroline Marie Chopko
Submitted to the Department of Chemical Engineering on September 26, 2014 in partial
fulfillment of the requirements for the degree of
Doctor of Philosophy in Chemical Engineering
Abstract
Step-growth hydrogels crosslinked through grafted polypeptides are introduced as a powerful platform for
extending the potential of established hydrogel systems, especially for applications in tissue engineering.
Gels crosslinked through grafted polypeptides offer the potential to address many limitations of
established poly(ethylene glycol)-only hydrogel systems, but most notably, gels crosslinked through
synthetic peptides are expected to 1) provide handles to systematically incorporate and modulate
biological, mechanical and chemical signaling, and 2) more closely mimic protein secondary structure
found in the native extracellular matrix.
A specific grafted N-carboxyanhydride polypeptide, poly(y-propargyl-L-glutamate) (PPLG), forms the
foundation of this thesis. PPLG is an especially useful polymer for exploring hydrogel crosslinking
through grafted polypeptides because it 1) can be grafted with nearly perfect efficiency with a wide
variety of functional groups, and 2) maintains a highly stabilized a-helical secondary structure before and
after grafting
Characterization of solution phase behavior of PPLG fully grafted with various side groups demonstrates
the ability to precisely control polymer bulk behavior by systematically tuning the average ratio of
complementary grafting groups, with broad application in pH- and thermo-responsive drug delivery.
Extending these solution-phase studies to gel systems, foundational characterizations first establish a
modular, well-controlled synthetic platform for synthesizing crosslinker-grafted PPLG, easily extended to
a wide variety of covalent crosslinking chemistries. Swelling ratios, fraction polymer incorporation, and
bulk gel stiffness measurements of hydrogels crosslinked through grafted PPLG strongly support both
stochastic substitution of PPLG grafting groups, and significant a-helical secondary structure of graftedPPLG even when crosslinked into a gel. Preliminary studies identify grafted PPLG as supporting both 2D
and 3D cell culture.
Future studies look to expand the scope of these findings to other grafted polypeptide hydrogels with
other grafting strategies and grafting groups. Together, these findings recommend gels crosslinked
through grafted synthetic polypeptides as a platform for investigating and controlling cellular response for
in vitro and in vivo applications.
Thesis Supervisors: Professors Linda Griffith and Douglas Lauffenburger
3
Acknowledgements
I could not have completed this thesis without the support and kindness of so many, both those mentioned
here and the many others who will not be forgotten.
I am fortunate for the many faculty, who have contributed to my success at MIT. First, I would like to
thank my primary thesis advisor, Prof. Linda G. Griffith for being supportive, creative and committed to
doing useful science. Thank you to Professors Paula Hammond, Doug Lauffenburger, Barbara Imperiali
and Alan Grodzinsky. Each of you contributed significantly to the direction of this thesis and in my
development as a scientist and an engineer. I especially appreciate the incredible opportunity and freedom
to explore so many new and exciting questions of how biomaterials can be used to control and
characterize healthy and diseased tissue.
While at MIT, I have had the opportunity to work with so many talented post docs, graduate students, and
undergraduates. First, I would like to thank Amanda Engler who first introduced me to the wonders of
PPLG. Jorge Valdez and Christi Cook, thank you for dealing with so many of my ideas and helping me
sort through the good ones. Thank you to the long line of Imperiali lab chemists who have taken me under
their wings, in order, Matthieu Sainlos, Angelyn Larkin, Andrew Krueger, Carsten Kroll and Kasper
Renggli. Thanks to Eunice Costa for introducing me to layer by layer on beads. Thanks to the protease
team of Miles Miller, Ravindra Kodihalli and Julie Ramseier for opening the door to these fantastic
proteins. Experiments and ideas by Liz Welch contributed directly to this thesis, especially mechanical
measurements by AFM. I couldn't have asked for a better collaborator. Special thanks for Margaret Guo
for computational help. Lauffenburger lab members Shannon Alford, Abby Hill, Aaron Meyer and Simon
Gordonov who regularly brought me back to focusing on applications of cool materials. Hammond lab
members Wei Li, Mohi Quadir, Steven Castleberry, and Wade Wang contributed to work not presented in
this thesis. Thank you to Griffith Lab members including Grinia Nogueria, Abigail Oelker, Jaclyn
Sheppard, Megan O'Grady, Edgar Sanchez, Jaime Rivera, Sam Raredon, and Shelly Payton who have
supported in big and small ways. And thank for Linda Stockdale and Hsinhwa Lee for helping to make
everything come together and for teaching me so much about science and life.
One of the greatest joys of my time at MIT has been working through problems with amazing
undergraduate researchers, especially Erika Lowden, Julie Ramseier, Jenn Ibanez, Jackie Simpson Jenna,
Ricardo, Rebecca, and Anasuya. Thanks for asking the hard questions that drive research forward.
Thanks to the Friday lunch crew, Monday Nights at St. Cs, and the MIT pottery studio. Fantastic friends,
families, roommates, have helped make my time at MIT even more amazing.
Last but most importantly, I would like to thank my family (Mom, Dad, Adam, and Travis). Mom and
Dad, thank you for everything, even proofreading this final thesis. Adam, thanks for being my best friend
in Boston for four years. Travis, thank you for always being there and loving me every day.
4
Table of Contents
List of Figures...............................................................................................................................................
7
1
Introduction and Background..........................................................................................................
10
1.1
M otivation and thesis overview ..............................................................................................
10
1.2
Introduction to polypeptide crosslinked hydrogels .................................................................
12
1.3
Introduction to PPLG grafting groups and functionalities......................................................
13
1.4
Introduction to PPLG crosslinked hydrogel synthetic strategies ............................................
14
1.5
Theory outlining grafted PPLG conform ation and structure .................................................
14
1.5.1
Secondary structure of grafted PPLG .............................................................................
15
1.5.2
M echanical properties of individual a-helices ...............................................................
17
1.5.3
Geom etry of PEG grafted onto PPLG ..............................................................................
19
1.6
M odeling grafted PPLG and PEG crosslinked hydrogel ........................................................
20
1.6.1
M odeling grafted PPLG in solution ...............................................................................
21
1.6.2
Modeling step-growth PEG hydrogels crosslinked through grafted PPLG .................... 22
1.7
PPLG as a non-structural handle for introducing (nanostructured) bio-functionality to hydrogels
....................................................................................................................................................
24
1.8
Conclusions.................................................................................................................................25
1.9
References...................................................................................................................................25
2
Solution-phase behavior of grafted PPLG with multiple grafting groups .....................................
2.1
Introduction.................................................................................................................................
2.2
Materials and M ethods................................................................................................................31
2.3
Results and Discussions ..........................................................................................................
2.4
Conclusion ..................................................................................................................................
2.5
Supporting Inform ation..........................................................................................................
2.6
References...................................................................................................................................
3
29
30
33
38
39
48
Synthesis of acrylate grafted, biofunctional PPLG hydrogels and their application in 2D cell culture..
............................................................................................................................................................
50
3.1
Aqueous crosslinking of PPLG through short acrylates ........................................................
50
3.1.1
Introduction.........................................................................................................................50
3.1.2
M aterials and Methods...................................................................................................
51
3.1.3
Results and Discussion...................................................................................................
53
3.1.4
Conclusions.........................................................................................................................58
3.2
General strategies for organic grafting onto PPLG at low substitution ..................................
58
3.2.1
Introduction.........................................................................................................................58
3.2.2
M aterials and M ethods...................................................................................................
63
3.2.3
Results and Discussion...................................................................................................
66
3.2.4
Conclusions.........................................................................................................................71
3.3
References...................................................................................................................................73
4
Step-growth hydrogels crosslinked through uncharged helical polypeptides for applications in tissue
engineering..................................................................................................................................................75
4.1
Introduction.................................................................................................................................75
4.2
M aterials and M ethods................................................................................................................76
4.3
Results and Discussion................................................................................................................81
4.3.1
PPLG Macrom er Synthesis..............................................................................................
81
4.3.2
Properties of Hydrogels containing PPLG-g-(maleimidexEO2,)...................................
83
4.3.3
Gel point of hydrogels crosslinked through PPLG-g-(maleimideEO2z) ....................... 85
4.3.4
Swelling of hydrogels crosslinked through PPLG-g-(maleimideEO2,)........................88
4.3.5
Mechanical properties of hydrogels crosslinked through PPLG-g-(maleimideEO2,).......90
5
4.3.6
Introducing hydrogels crosslinked through PPLG-g-(maleimidesnorborneneyEO2,).........92
4.3.6
Swelling of hydrogels crosslinked through PPLG-g-(norbomeneyEO2z).......................92
4.3.7
Mechanical properties of hydrogels crosslinked through PPLG-g-(norborneneyEO2,)......94
4.3.8
Swelling and mechanical properties of hydrogels crosslinked through PPLG-g(maleimidenorborneneyEO2,)............................................................................
....... 95
4.3.9
Cell adhesion to 2D hydrogels crosslinked through PPLG-g-(maleimideEO2,)............98
4.3.10
Cell viability in 3D hydrogels crosslinked through PPLG-g-(norborneneyEO2z) .......... 98
4.4
C onclusions.................................................................................................................................99
4.5
References.................................................................................................................................
100
5
Thesis summary and future works ....................................................................................................
5.1
Sum m ary of thesis.....................................................................................................................
5.2
Future w ork ...............................................................................................................................
5.2.1
Alternative crosslinking strategies for grafted PPLG hydrogels.......................................
5.2.2
The influence of local presentation and stiffness on cellular adhesion to PPLG-grafted
peptides ...........................................................................................................................................
5.2.3
Expanding PPLG grafting groups and introducing grafted group clustering....................
5.2.3.2
Grafting on proteins and peptides - bulk and nano- organization .....................................
5.3
R eferences.................................................................................................................................
6
105
105
106
106
109
110
111
112
List of Figures
Figure 1.1: PEG hydrogels formed from A) Free radical crosslinking of di-functionalized PEG chains, B)
Step-growth crosslinking of orthogonally functionalized multi-arm PEG macromers, and C)
Step-growth crosslinking of orthogonally functionalized multi-arm PEG macromers and
10
grafted polypeptides. ...............................................................................................................
Figure 1.2: PPLG grafting where biofunctionality and crosslinkers can be grafted to the PPLG backbone
through an organic phase click reaction, an aqueous phase click reaction and incorporated
13
into the gel through PPLG during gelation.........................................................................
16
Figure 1.3: End to end distance of polypeptide chains modeled.............................................................
Figure 1.4: End to end distance of a PEG chain dependence on molecular weight (MW) where each
20
ethylene glycol repeat unit is 44 g/m ol...............................................................................
21
Figure 1.5: Dim ensions of soluble PPLG ...............................................................................................
Figure 1.6: Schematic of an idealized step-growth polypeptide gel modeled as an interpenetrating face
center cubic crystal structure of the PPLG and PEG crosslinkers having PPLG-g-E0 2 DOP
23
160 at 4% wt/vol polymer with 10k 4-arm PEG thiol at 1.3 wt%..................
Figure 1.7: PPLG grafted with ethylene oxide, crosslinking functionality and bio-functionality. PPLGgrafted biofunctionality might be, A. grafted from a long tether as single groups, B. grafted as
24
clusters or C. grafted mixed functionality. .........................................................................
Figure 2.1: Chemical structure of poly(y-propargyl-L-glutamate) backbones fully substituted with thermo
31
and pH responsive side groups...........................................................................................
33
Figure 2.2: Inventory of thermo- and pH- responsive grafted PPLG polymers.....................................
Figure 2.3: Relationship between cloud point (measured at 3 mg/ml in deionised water) and percent E2
34
substituted onto PPLG backbones functionalized with mEO2 ............................................
Figure 2.4: Influence of temperature on the light transmittance (500 nm, heating 1 'C min-1) of PPLG-97
substituted with 100% E0 2 and PPLG-77 substituted with E0 2 at 19%, 40%, and 100% and
35
remainder with mEG 2 at 3 mg/ml in deionised water .......................................................
Figure 2.5: Relationship between cloud point of grafted PPLG backbones and solution pH measured at I
36
mg/ml in 100mM NaCl, 75mM phosphate buffer...................................................................
Figure 2.6: Circular dichroism spectrum of PPLG-64 measured at 1 mg/ml in 100 mM NaCl, 75 mM
phosphate buffer pH 6.2 at indicated temperatures with various grafting........................... 37
39
Figure 2.S1: Normalized GPC traces of PPLG backbones....................................................................
Figure 2.S2: 'H-NMR spectrum PPLG backbone substituted with mE0 2 and E0 2 or diisoproylamine in
40
[d7] D M F .................................................................................................................................
Figure 2.S3: Influence of temperature on the light transmittance (500 nm, heating 1 'C min-1) of PPLG41
64 fully substituted with mE0 2 polymers in deionized water ............................................
Figure 2.S4: Influence of temperature on the light transmittance (500 nm, heating I 0C min-1) of PPLG
polymers fully substituted with mE0 2 at 3 mg/ml in deionised water ................................ 41
Figure 2.S5: Influence of cycling temperature across 4 heating and cooling cycles from 20 'C to 40 'C on
the light transmittance of PPLG-64 fully substituted with mE02 ................... .. . . .. . .. .. . . .. . . . . 41
Figure 2.S6: Temperature dependent solubility of grafted PPLG with heating and cooling traces and
discussion of implications of stochastic grafting on PPLG thermoresponsiveness............. 42
Figure 2.S7: Influence of pH on the temperature-dependent solubility of PPLG-64 graft 100%
44
diisopropylam ine.....................................................................................................................
Figure 2.S8: Titration of grafted PPLG-64 with diisopropylamine mEO2 ....................... . .. . . .. .. . . .. .. . .. . . . 45
Figure 2.S9: Temperature dependent circular dichroism spectra of PPLG substituted with only mEG 2 or
46
E 02. .......................... .......... . . ...........................................................................................
with
Figure 2.S 10: Representative temperature dependent circular dichroism spectra of PPLG substituted
47
100% or 50% diisopropylam ine..............................................................................................
7
Figure 3.1: Acrylate-PEG= 2-azide synthesis from acryloylcholoride and 2-(2-azidoethoxy)ethanol, with
triethanolamine (TEA). Azide functional side chains are grafted to PPLG in a 2-phase organic
click reaction. Feed ratios of acrylate vary from grafting ratios observed by 'H NMR,
54
indicating significant side reactions in grafting group .......................................................
Figure 3.2: Varied ratios of 4-arm PEG- acrylate 10k (PEG) and DOP 76 PPLG-g-acrylate with
fluorescein (PPLG), each having a molecular weight of 2.5k/acrylate, were crosslinked with
2-arm PEG-thiol 3.4k (PEG thiol). The table shows representative data of the final mass of
the dried polymer gels made with the same wt % polymer and swelling ratios (mass swollen
55
gel/m ass dry gel) of these gels. .........................................................................................
and
2-arm
with
fluorescein
(PPLG)
ratios
DOP
76
PPLG-g-acrylate
3.3:
A.
Gels
having
varied
Figure
PEG thiol-3.4 k (PEG thiol), but constant 12 wt% of reagents, were crosslinked and swollen
in PBS. B. Leached fluorescence from fluorescein-grafted PPLG was calculated as a fraction
of total fluorescence in polymer precursor (n=3 gels). C. A screen (n=1) reporting final mass
fraction of precursor polymer and swelling ratios (mass swollen gel/mass dry gel) of the
57
series of polypeptide crosslinked gels...............................................................................
Figure 3.4: A 4-arm PEG acrylate-IOk (PEG) and DOP 76 PPLG-g-acrylate(9 per) with fluorescein
(PPLG), each having a molecular weight of 2.5k/acrylate, were crosslinked with 2-arm PEG
thiol3.4k (3.4k PEG thiol) or 2-arm PEG thiolIOk (10k PEG thiol). The table shows
representative data of the final mass of the dried polymer gels made with the same wt%
polymer and swelling ratios (mass swollen gel/mass dry gel) of these gels. ...................... 57
Figure 3.5: Distribution of grafting groups per polymer of polymers having a degree of polymerization
(DOP) of 8, 25, 50 or 200 with an average 4 grafting groups per backbone, as modeled with
60
the binomial probability density function...........................................................................
Figure 3.6: Schematic of representative azide-grafted peptides, having carbon linker or a peg linker. ..... 62
Figure 3.7: A. Chemical structure of RGDS having a PEG linker, synthesized through conjugation of Nhydroxysuccinimide-activated carboxylic acid PEG. Also, the Figure 2.Shows a schematic of
4-methypiperidine catalyzed amine deprotection and B. Chemical structure of RGDS peptide
66
having a glycine spacer and a terminal azido lysine. ........................................................
Figure 3.8: A. 'H NMR of PPLG backbone in DMF, B. representative 1H NMR of PPLG partially grafted
with fluorescein, C. representative 'H-NMR of PPLG fully grafted with fluorescein and
PEGI00and purified by dialysis and D. correlation of feed ratio and substitution for a series
of reactions with varied feed ratio and linear fit of calculated substitution from Iluorescein
69
proton integration. ...................................................................................................................
Figure 3.9: A. Calculations of the average molar mass of PPLG-g-RGD and of the RGD grafting group of
a representative polymer and representative brightfield image of hTERT MSCs cultured for
24 hours on PEG hydrogels with PPLG-g-RGD where the concentration of RGD is estimated
70
as 0 .83 m M ..............................................................................................................................
Figure 3.10: Aqueous grafting of thiol terminal peptide to PPLG-g-maleimide. Initial thiol concentration
and complete thiol grafting can be precisely quantified versus a standard curve using
72
E llm an's R eagent. ...................................................................................................................
and
grafting
on
of
-norbomene
and
N
-PEG=1
of
N
-PEGn=
-maleimide
4.1:
In
situ
synthesis
Figure
0
3
3
10
azide-functionalize crosslinker and E0 2 .......................................... . . . .. . .. . .. . . .. . .. . .. . . .. . .. .. . . . . . 82
Figure 4.2: Representative ninhydrin-stained thin layer chromatography plate with N 3-PEGn=1 0-amine
82
standard curve and reaction product....................................................................................
crosslinker
as
Figure 4.3: Correlation of norbornene and maleimide grafting feed ratio and PPLG-grafted
quantified by 1H NMR PPLG backbone and grafting group peak integrals....................... 83
83
Figure 4.4: Representative 'H NMR of PPLG-g-maleimide and norbomene..........................................
Figure 4.5: Schematic of PPLG and 8-arm PEG maleimide-IOk or -40k crosslinking with 4-arm PEG
th iol- Ok. .................................................................................................................................
85
Figure 4.6: Extent of reaction at gel point as modeled with the Flory-Stockmayer theory for gels........... 87
8
Figure 4.7: Swelling ratios of gels crosslinked through PPLG-g-maleimides and control PEG gels
88
crosslinked through 8-arm PEG maleimide. .......................................................................
Figure 4.8: Elastic moduli quantified by AFM indentation of gels crosslinked through PPLG-gmaleimides and control PEG gels crosslinked through 8-arm PEG maleimide ................. 90
Figure 4.9: Distribution of grafting groups per polymer of polymers having a degree of polymerization
(DOP) of 160 with an average 5.6 and 9.8 grafting groups per backbone, as modeled with the
91
binom ial probability density function. ................................................................................
93
Figure 4.10: Phase contrast (20x) images of PPLG-g-norbornene gels ................................................
94
Figure 4.11: Swelling ratios of gels crosslinked through PPLG-g-norbornene ....................................
Figure 4.12: Elastic moduli quantified by AFM indentation of gels crosslinked through PPLG-g95
norbornene ..............................................................................................................................
Figure 4.13: Characterization of gels with single and double crosslinking of three starting polymers,
96
PPLG-g-maleimide and norbornene fully substituted with E02 .........................................
Figure 4.14: Theoretical estimates of the weight fraction having no crosslinkers at given grafting ratios
assuming stochastic modeling and associate maximum gel incorporation. ........................ 97
Figure 4.15: Cell adhesion of hTMSC seeded on PEG and PPLG hydrogels with adhesive peptide......... 98
Figure 4:16: Overall viability as quantified by Live/Dead staining of hTMSC encapsulated in hydrogels
(25 ptL, 15k cells) crosslinked through 4-arm PEG norbornene-IOk or PPLG-g-nor(6.0 per)
99
w ith 4-arm PEG thiol-IOk ..................................................................................................
9
1
Introduction and Background
1.1
Motivation and thesis overview
Decades of research have established poly(ethylene glyol) (PEG) crosslinked hydrogels as useful
synthetic extracellular matrices (ECM) for tissue engineering applications. The field has expanded from
initial studies of gels made from unordered free radical crosslinked PEG macromers, in which crosslinked
PEG chains grow from radical-polymerized hydrocarbon backbones shown below in black (Figure 1. lA),
to include gels made from step-growth multiarm PEG macromers, in which PEG end groups of
complementary functionality react to form a crosslinked network (Figure 1.1 B). Despite the success and
continued growth in the field of end-linked PEG hydrogels for applications in 2D and 3D cell culture
systems, established gels struggle to fully replicate native extracellular matrix control of cellular response.
Limitations of established PEG gel systems might reasonably be improved by 1) having an enhanced
ability to tether and to characterize a wider and more useful range of protein and peptide cues, 2)
optimizing nanoscale organization of these signaling cues, 3) incorporating secondary structure and local
mechanical rigidity inherent in most native matrices 4) extending independent control of gel mechanical
and chemical properties, and 5) systematically expanding bulk gel chemical properties to include more
small molecules or proteoglycans.
A.
B.e
C.
Figure 1.1: Schematic of PEG hydrogels formed from A) Free radical crosslinking of di-functionalized
PEG chains, B) Step-growth crosslinking of orthogonally functionalized multi-arm PEG macromers, and
C) Step-growth crosslinking of orthogonally functionalized multi-arm PEG macromers and grafted
polypeptides.
This thesis extends established step-growth PEG hydrogels into a new gel platform in which one
component of step-growth PEG hydrogels has been substituted by a-helical polypeptides grafted with
multiple crosslinking side chains (Figure 1. lC). Gels crosslinked through grafted polypeptides offer the
potential to address many limitations outlined above for established hydrogels, but most notably gels
crosslinked through synthetic peptides are expected to 1) provide hundreds of handles to systematically
incorporate and modulate biological, mechanical and chemical signaling, and 2) more closely mimic
protein secondary structure found in the native extracellular matrix (ECM).
10
In more detail, the potential of each crosslinking polypeptides to present hundreds of grafted
groups addresses an inherent limitation of the most established step-growth hydrogels, where each
biofunctional group introduced through the gel removes a potential structural crosslink, modulating bulk
mechanical properties and imposing an upper limit on the concentration of pendent functional groups
(Zustiak, Durbal, and Leach 2010). In the case of introducing bioadhesive peptides, researchers have
developed strategies to decouple ligand presentation and bulk gel properties by attaching the peptide
within a modified crosslinker, not to the ends, (Zhu et al. 2006; Deforest, Sims, and Anseth 2010;
Gandavarapu, Azagarsamy, and Anseth 2014), or even introducing ligands as "beads" threaded along
PEG chains (Singh et al. 2013). However, such approaches require significant synthetic effort and limit
modularity. Additionally, the expected utility for hydrogels made of macromers more closely mimicking
protein-like structures has been founded on both the growing appreciation for the role of highly non-linear
mechanical properties of ECM (Wen and Janmey 2013) in controlling cellular response, and also on
recent work suggesting local mechanical properties as providing even more nuanced control of cellular
phenotype (Trappmann et al. 2012).
Recent advances in NCA chemistry have demonstrated synthetic techniques for highly modular
polypeptides readily optimized for broad applications in tissue engineering, including for applications as
hydrogel crosslinkers as demonstrated in this thesis. Since the first NCA polymerized polypeptide in
1906, the field has expanded from an initial focus of synthesizing native polypeptides to more recent
efforts establishing synthetic routes to polypeptides having a wide range of functionalities and structures
(Kricheldorf 2006; Cheng and Deming 2012). In particular, recent research has demonstrated the
immense potential of grafted non-native synthetic polypeptides appropriate as crosslinkers for highly
engineered hydrogels. These polymers are assembled through NCA monomers having reactive side
chains and later functionalized via grafting onto and grafting from the polypeptide backbone. Postpolymerization grafting allows polypeptides to be functionalized with groups previously incompatible
with NCA polymerization. Engler et al (2009) introduced the first example of such systems to use click
functional side groups, thus opening up a broad set of new capabilities for non-native functionalities. The
reader is referred to excellent recent reviews outlining advances in grafted polypeptides (Deng et al. 2014;
Lu et al. 2014; Quadir, Martin, and Hammond 2014). Further, post-polymerization grafting offers the
opportunity to easily extend synthetic polypeptides beyond homopolymers to introduce multiplexed
polypeptide functionality (Tang and Zhang 2011; Rhodes and Deming 2013; Huang et al. 2011; Kramer
and Deming 2012).
A specific grafted NCA polypeptide, poly(y-propargyl-L-glutamate) (PPLG), forms the
foundation of this thesis (Engler, Lee, and Hammond 2009). PPLG is an especially useful polymer for
exploring hydrogel crosslinking through grafted polypeptides because it 1) can be grafted with nearly
11
perfect efficiency by a wide variety of functional groups, and 2) maintains a highly stabilized a-helical
secondary structure before and after grafting (Engler, Lee, and Hammond 2009). For the purposes of this
thesis, PPLG polypeptides can be thought of as a chain of highly organized hooks to which can be added
a wide variety of functional groups.
In addition to the efficient grafting on of PPLG, PPLG's robust a-helical secondary structure
makes it particularly appropriate for initial explorations of step-growth gels crosslinked through grafted
polypetides. Native extracellular matrices are crosslinked through proteins having defined secondary
structure, while the vast majority of synthetic hydrogels have been crosslinked through polymers well
represented as random coils. Crosslinking hydrogels through grafted PPLG offers the opportunity to
systematically explore the effect of defined secondary structure on local cell response and bulk hydrogel
properties. Such understanding is expected to both direct design of future synthetic extracellular matrices
and better define the role of secondary structure in the native ECM. These studies focusing on PPLGcrosslinked hydrogels should
In summary, this thesis presents the synthesis, fabrication, and characterization of PPLG
crosslinked hydrogels and applies these gels as synthetic extracellular matrices in 2D and 3D cell culture
systems. It is expected that general principles established through studies with PPLG might be extended
to inform the design of hydrogels crosslinked through other grafted NCA polypeptides. As such, this
work provides a framework for leveraging advances in NCA polymerization to meet limitations of current
engineered synthetic extracellular matrices. Step-growth hydrogels crosslinked through engineered
polypeptide macromers having defined secondary structure and presenting a wide variety of crosslinking
and bio functionality are expected to extend the level of cellular control achieved through existing
engineered hydrogel systems.
1.2
Introduction to polypeptide crosslinked hydrogels
Covalently crosslinked polypeptide hydrogels have been most often crosslinked through N-
carboxyanhydride (NCA) polymerized homopolymers of native amino acids, including gels crosslinked
through poly(glutamic acid) (Markland et al. 1999; Zhang et al. 2011, Ding et al. 2011), poly(lysine)
(Oliveira et al. 2003), and poly(aspartic acid) (Gyenes et al. 2008). While useful for applications such as
drug delivery, these gels' high charge density and non-specific protein absorption limit their utility for
tissue culture applications traditionally requiring systematic control of biological and chemical cues.
In 2012, the Hammond group built on work presented in Chapters 2 and 3 of this thesis to
introduce step-growth hydrogels covalently crosslinked through neutral, water-soluble polypeptides
(Oelker et al. 2012). Gel crosslinking was demonstrated through anhydrous activation of PPLG fully
grafted with ethylene oxide by a non-specific coupling agent, and, without purification, crosslinking
12
.....
.........
through 4-arm PEG thiol-IOk (Oelker et al. 2012). Chapter 4 of this thesis extends these published results
to introduce a hydrogel crosslinked through a heterofunctional PPLG grafted with both a solubilizing
ethylene oxide bush and various crosslinkers. This novel synthetic strategy leverages both the immense
potential of grafted polypeptides to crosslink highly modular gels with chemistries appropriate for peptide
and even cell encapsulation.
1.3
Introduction to PPLG grafting groups and functionalities
As outlined in a recent review (Quadir, Martin, and Hammond 2014), this pendent alkynes of
PPLG allow grafting on of a wide variety of azide functionalized side chains via copper catalyzed 1,3cyclo addition. Examples of reported grafting groups include long PEG chains (Engler, Lee, and
Hammond 2009), short sugar molecules (Xiao et al. 2010), amines (Engler, Shukla, et al. 2011), sulfonate
ions (Shih et al. 2014), and thermoresponsive ethylene glycol grafting groups (Chopko et al. 2012) . The
robust a-helix before and after grafting of the polypeptide allows for almost perfect grafting efficiency
largely insensitive to the properties of the grafting azide (Engler, Lee, and Hammond 2009).
Unfunctionalized PPLG is only soluble in select organic solvents, which constrains PPLG
grafting reactions to solutions in dimethyl sulfoxide (DMSO) or dimethylformamide (DMF). In initial
organic grafting (Figure 1.2 below), azide terminated grafting groups can be conjugated directly to PPLG
pendent alkynes through copper catalyzed 1,3-cycloaddition. If grafted groups introduced through organic
grafting are sufficiently hydrophilic, PPLG might next be dissolved in an aqueous buffer and conjugated
with additional functionality through a secondary aqueous grafting (Figure 1.2). In aqueous grafting,
azides may be grafted onto remaining PPLG alkyne groups through copper catalyzed 1,3-cycloaddition.
As a second level of grafting, additional crosslinking chemistries orthogonal to 1,3-cycloaddition may be
introduced to PPLG through organic grafting and reacted again during aqueous grafting. Together,
organic and aqueous grafting onto PPLG allow well-controlled grafting on of a wide variety of functional
groups.
0--
1. OrganIc GraftIng
2. Aqueous Grafting
(Optional)
3. Aqueous Gel
(with multirm PEG)
Figure 1.2: Schematic of PPLG grafting where biofunctionality and crosslinkers can be grafted to the
PPLG backbone (shown in blue) through an organic phase click reaction, an aqueous phase click reaction
and during gelation.
13
__AW
This thesis highlights the utility of both organic and aqueous phase grafting in synthesizing PPLG
macromers crosslinked into hydrogels the mimic the biofunctionality of the native ECM. The optimal
PPLG grafting synthetic strategy is dictated by the end application as well as the stability and solubility of
the conjugating groups.
1.4
Introduction to PPLG crosslinked hydrogel synthetic strategies
Synthesizing crosslinking PPLG conjugated with two or more crosslinking groups per PPLG is
expected to allow robust gel formation when polypeptides are reacted with multi-arm crosslinkers. As
such, PPLG can reasonably be crosslinked into a gel through crosslinkers either stochastically grafted
onto polypeptide side chains or crosslinkers directly conjugated to the two polypeptide end groups.
Stochastically governed side chain crosslinker grafting to PPLG readily generates PPLG
crosslinking backbones grafted with a range of average crosslinking groups, as demonstrated in this
thesis. However, heterogeneity in the number of grafted crosslinkers per macromer generate imperfect
gels expected to have decrease gel mechanical properties compared to more ordered, end-crosslinked
systems. Additionally, variable side chain grafted macromere crosslinking complicates direct comparisons
of PPLG with PEG step-growth gels, as PEG gels are most often uniformly functionalized with a wellcontrolled number of crosslinking groups per macromer.
Crosslinking through PPLG end groups offers a promising alternative to stochastic grafting
allowing exactly two functional groups to be grafted to each crosslinking PPLG. However, synthetic
challenges have hindered this approach to date. Unpublished results highlight difficulties encountered in
efficiently conjugating both ends of PPLG, and PPLG with only one crosslinker cannot mechanically
contribute to gel formation. Crosslinkers can be readily conjugated to a single PPLG end when
polymerization is initiated by a dual functional initiator having both an initiating primary amine and a
protected orthogonal grafting group. Introducing a second crosslinker through conjugation with the
polymerizing chain end has been significantly less robust, with an experimentally observed maximum of
only 75% efficiency (unpublished results). As such, crosslinking through PPLG end groups is not
extensively considered in this thesis though it is expected that studies presented in this thesis
characterizing side-chain grafted PPLG crosslinked gels will extend inform future efforts crosslinking
through PPLG end groups.
1.5
Theory outlining grafted PPLG conformation and structure
An underlying assumption motivating the expected utility of crosslinking hydrogels through
grafted PPLG is that the PPLG's distinct secondary structure enables nano- and macroscale hydrogel
properties not already demonstrated by less structured crosslinking systems. Literature and theory guiding
the understanding of grafted-PPLG conformation and structure will be explored in more detail in the
14
following sections, specifically focusing on four areas: 1) solution phase secondary structure of grafted
PPLG, 2) mechanical properties of individual a-helices (especially compared to PEG chains), 3) geometry
of PEG grafted onto PPLG, and 4) modeling and implications of stochastic grafting onto PPLG.
Each of these individual topics will be brought together to develop a spatial model of the soluble
grafted PPLG presented in Section 1.5.3. This solution phase model will inform a model of grafted PPLG
crosslinked into a hydrogel, as presented in Section 1.3.
1.5.1
Secondary structure of grafted PPLG
The geometry of a grafted polypeptide is largely controlled by its secondary conformation, which
has known dependencies on the inherent properties of the polypeptide (degree of polymerization and side
chain functionalization) as well as its environment (solvent and concentration). The following section,
Solutions phase PPLG secondary structure, reviews published studies of similar polypeptides to suggest
that PPLG is primarily a-helical in dilute solutions. A second section, Gel-crosslinked PPLG secondary
structure, extends these solution phase studies to hypothesize how a polypeptides's secondary structure
may change with crosslinking at high concentrations in hydrogels.
Solutions phase PPLG secondary structure
Aqueous studies of dilute polymer by circular dichroism have characterized a grafted PPLG's
secondary structure as primarily a-helical, even after grafting on long PEG chains (Engler, Lee, and
Hammond 2009), short positively charged amines (Engler et al. 2011) and thermoresponsive short
methoxy and hydroxyl terminated PEG linkers (Chopko et al. 2012). Recent research has developed
theory partially explaining the robust a-helical structure demonstrated by PPLG. Systematic screens have
identified integration of a hydrophobic ring structure adjacent to the polypeptide backbone and greater
extension of the charged group from this hydrophobic spacer as stabilizing polypeptide a-helices (Lu et
al. 2011). Similarly the triazole ring from the "click" linker group on PPLG and resulting long side chain
extension may contribute to the dilute grafted polymer's remarkable helical stability.
Assuming an a-helical secondary structure, PPLG is expected to have an extremely regular rodlike conformation dictated by stabilizing hydrogen bonds along the polypeptide backbone. Specifically, ahelices are known to have a translation of 1.5 A (0.15 nm) along the helical axis per amino acids and each
complete turn contains 3.6 amino acids (Corey and Pauling 1953; Langel et al. 2010). From these
assumptions, the dependence of the end to end distance of a perfect a-helix is plotted in Figure 1.3. These
lengths can be compared to random coil (approximated as ranging from [(70-140)* # of residues]" 2
A)
(Tanford, Kawahara, and Lapanje 1966) and fully extended conformations (3.8 A per residue) (Langel et
al. 2010), also seen in Figure 1.3.
15
'
70
1-alphahelix
-
80
60
--
f
lly extended
50 ...----- random coil
S40
g 30
20
$
10
0
0
100
50
150
200
PolypeptideDegree of Polymerization
Figure 1.3: End to end distance of polypeptide chains modeled.
Theoretical approximations of a-helical geometry have been experimentally validated for many
synthetic polypeptides including poly(y-benzyl-L-glutamate) (PBLG). At first approximation, polymer
hydrodynamic radius as a function of molecular weight can be observed through light scattering
measurements of at least two polymers having the same side chains but different degrees of
polymerization. Such studies have been rigorously conducted for PBLG in DMF, a highly helicogenic
solvent, in combination with gel permeation chromatography (GPC) (Temyanko, Russo, and Ricks 2001).
Observed radii of gyration were modeled assuming a given persistence length and compared to
experimental observations, using established calculations for wormlike chains of a given persistence
length (Cotts, Swager, and Zhou 1996), from which any intrinsic PBLG persistence length was estimated
as approaching 240 nm, supporting a-helical structural characterization.
Gel-crosslinked PPLG secondary structure
When considering the expected secondary structure of grafted PPLG crosslinked into hydrogels,
studies of other a-helical polypeptides offer strong precedence that polypeptide's secondary structure in
solution will be well-preserved in the crosslinking hydrogels. Studies of oligopeptide melts demonstrate
the role of concentration in introducing destabilizing kinks to pure a-helical structure. Researchers found
that in concentrated conditions PBLG assumes a less ordered conformation, particularly at low degrees of
polymerization (DOP), e.g. below 20 (Papadopoulos et al. 2004). Further, detailed analysis of even
longer, more stable polymers showed periodic instabilities in a-helical structures, referred to as kinks
(Papadopoulos et al. 2004), resulting in what can be referred to as a broken-rod structure. These kinks,
16
separating sections of perfect a-helices, introduce a particular challenge for estimating a true persistence
length of the helix in non-dilute solutions.
As an intermediate case, perhaps more suggestive of polypeptide structures in gels, the same
PBLG having degrees of polymerizations of 34 to 186 was studied grafted to a rigid poly(norbornene)
polymer at dilute and concentrated grafting (J. Wang et al. 2011). This study used Nuclear Overhauser
Enhancement Spectroscopy (NOESY) experiments to determine that especially at high grafting densities
and high molecular weights PBLG on polynorbornene-g-PBLG adopts an interrupted helical structure,
again represented as va broken rod. However, even in the case of the most concentrated grafting
conditions of long PBLG sidechains (DOP 120), brush polypeptides were expected to be broken into only
two helical segments, on average.
Specifically considering polypeptide grafted into gels, polypeptide secondary structure has not
been directly demonstrated. However, one study exploring hydrogels grafted with poly(lysine) at high and
low pH suggests a crude correlation of shear modulus with polymer volume fraction dependent on
polymer secondary structure (Oliveira et al. 2003). Taken together, these literature observations of related
polypeptide systems suggest that even in a hydrogel, grafted PPLG can be reasonably assumed to
maintain significant a-helical conformation.
The significant a-helical structure of crosslinked PPLG is further supported by studies published
by the Hammond group demonstrating that a gel's bulk modulus could be increased by incorporating ahelical PPLG polypeptides compared to gels from PPLG polypeptides having random coil secondary
structure (Oelker et al. 2012). While this study does not explicitly monitor PPLG polypeptide
conformation in the gel, or rigorously validate comparable crosslinking efficiencies in rod and coil
systems, it strongly suggests the a-helical character of crosslinked grafted PPLG.
1.5.2
Mechanical properties of individual a-helices
Better characterizing the mechanical properties of individual a-helices is useful in developing an
intuition for the stability of the secondary structure when crosslinked into a hydrogel and how this
secondary structure might influence bulk gel properties. In the case of PEG-only gels, most often gel
stiffness is tuned by modulating the number and length of the crosslinking arms and the extent of
crosslinking. Crosslinking through a-helical polypeptide macromers confers more complex nanoscale
rigidity, where this rigidity can be defined both as force per distance chain elongation and bending
modulus observed with externally applied torque. What follows reviews both the single molecule
response of PEG and polypeptides to force and establishes theories for how macromer structure translates
to bulk and local mechanical properties.
17
PEG mechanical properties
Mechanical properties of PEG in aqueous solutions have been extensively investigated, most
recently through characterizing single molecule extension and retraction by atomic force microscopy
(AFM). Detailed explorations of these studies are beyond the scope of this thesis but readers are
especially referred to AFM-based measurements of PEG extension in aqueous solutions (Oesterhelt, Rief,
and Gaub 1999). While PEG can be rigorously modeled as an ideal entropic spring in solvents such as
hexadecane, in water this group quantified significant contributions of both entropic and enthalpic PEG
elasticity. In more detail, in water, individual PEG molecules are stabilized by water bridging and adopt a
trans-trans-gauche helical conformation, which extends to a trans-trans-trans conformation with
elongation in the direction of the force. The observed highly non-linear extension force relationship is
attributed to non-planar supra-structure adopted by PEG polymers in aqueous solutions. Regarding
general stiffness, a single chain of tethered 30k molecular weight PEG polymer was shown to require
only ~5 pN (or 0.005 nN) of force to be extended from its random coil end to end chain length of 20 nm
to a fully extended length of around 240 nm, while extensions beyond require substantially increased
force (Oesterhelt, Rief, and Gaub 1999), approximating chain stiffness as 3x10 5 N m'. Similarly, another
study showed force below the limit of detection to extend PEG having molecular weight of 3400 g/mol
from 2 to 20 nm, with ~ 0.2 nN required to fully extend the PEG chain to around 27 nm (Zegarra et al.
2009). Together, these findings establish a framework of understanding PEG extension for which
individual PEG chains, approximated loosely as greater than 1k, extend 10 times their random coil
conformation with pN force but require exponentially increasing force to approach their fully extended
length. Fully extended conformation is generally approximated as 1 x the random coil formation.
However, it is known that most step- polymerized gels made from crosslinked polymer systems can be
uniaxially stretched only ~lx their swollen dimensions before breaking (Tibbitt et al. 2013) because of
defects in the polymer crosslinking and chain entanglement. Such assumptions have been supported by
studies of recent step-growth PEG hydrogels fabricated without detectable defects such that they
maintained structural integrity and showed no anisotropy under uniaxial elongation, even when stretched
to 5 times their swollen dimensions (Matsunaga et al. 2011).
Alpha-helical polypeptide mechanical properties
In comparison to PEG macromers, which exhibit stiffness most prominently at high extension, ahelical polypeptides exhibit significantly greater stiffness at both low and high extension. One study of
poly(glutamic acid) DOP 40 in water reports forces of 0.04 N m' required to extend the helical
polypeptide at intermediate stiffness, while less than 0.003 N m' was required to extend the random coil
polypepide at pH 8.0 (Zegarra et al. 2009). Therefore, at first approximation, in the range of nM
extension, a-helices in extension are at least 1 000x stiffer than PEG chains.
18
Alpha-helices deform not just with extension but also if torqued. Under small externally applied
torque, the helix deforms so that its chain contour adopts the arc of a circle and the torque required for
bending the rod through a given angle grows linearly with that angle. At a critical torque, the secondary
structure of the molecule is locally disrupted, producing a kink, or small length of the backbone with a
much softer bending modulus. In a particularly helpful analogy, researchers compared this process to the
bending of a drinking straw (Chakrabarti and Levine 2005). The force required to bend an a-helical
polypeptide is dependent on the relative persistence lengths of the ordered and disordered states, as well
as on chain cooperativity parameters, which are dictated by side chain grafting (Chakrabarti and Levine
2005). However, the highly cited moledular dynamics simulations of several representative polypeptides
suggest that the bending and twist elasticities of a-helices with neutral side chains are primarily'from
distortions of hydrogen-bonding along the backbone, making the polypeptides at small deformations wellrepresented by an elastic and isotropic rod (Choe and Sun 2005). Using a model of a 78 amino acid
polypeptide of poly(alanine) tethered at one end and with bending force applied to the other, these authors
posit a minimal force of 20 pN required to deform the helix and suggest that the bent polypeptide
maintains a-helical structure even with 50 pN force (Choe and Sun 2005).
These solution based studies can be extended to consider the influence of nanoscale rigidity on
the mechanical properties of step-growth hydrogels made only of PEG or of gels with both PEG and
polypeptides.
1.5.3
Geometry of PEG grafted onto PPLG
This section continues to develop the geometry of grafted PPLG by estimating the diameter of
PEG-grafted PPLG a-helices. PPLG without grafting is insoluble in many common solvents, but has been
demonstrated as soluble in DMSO, DMF, and a mixture of chloroform and trifluoracetic acid. Charged or
hydrophilic side chains confer aqueous solubility. This thesis looks primarily to PEG as a solubilizing
agent for hydrogel crosslinking, and a means of linking biomolecules.
First, considering the minimum radius of an aqueous soluble brush, as shown in Chapter 2,
complete grafting with 2-(2-azidoethoxy)ethanol (E02 ) was identified as sufficient to fully solubilize
PPLG in aqueous solutions where published research estimates a hydrated E0 2 brush as extending -0.8
nm (Wang, Kreuzer, and Grunze 2000).
Longer PEG chains, tethering grafted crosslinking functionality and bioactive molecules, present
a second relevant radius. Various models of understanding of the geometry of tethered PEG chains can be
visualized, as shown in Figure 1.4, where each ethylene glycol repeat unit has a molecular weight of 44
g/mol. The fully extended chain length assumes perfect trans conformation of the repeat units, and has the
longest end to end distance of length =Na where N is the number of repeat units and a is the length per
19
repeat unit, assumed as 3.5 A (Jeppesen et al. 2001). The reported random coil dimensions was
approximated modeling PEG as a freely jointed chain, where length =NO-
a, and where 0.64 is an
experimentally derived exponent for a PEG system, which is close to the Flory exponent (v), where v=
0.59 for the average dimension of asymptotically large polymer chains in a good solvent (Jeppesen et al.
2001).
Within the bounds of fully extended and random coil conformations, extensive research has
approximated real dimensions for given geometries and concentrations. For example, PEG chains in
solutions are known to adopt a more oval than perfectly spherical conformation as modeled by a freely
jointed chain (Lee et al. 2008). When tethered to a planar surface, PEG chains have been shown to adopt a
mushroom conformation (Rixman, Dean, and Ortiz 2003). Different studies about PEG-tethered biotin
and surface-tethered streptavidin have shown high frequencies of PEG binding events well beyond the
length scale predicted by random coil assumptions, where bindings are attributed to stochastic extensions
of the highly mobile tethered PEG chains (Moore and Kuhl 2006). At short PEG lengths, this binding
distance can approach the fully extended conformation (Jeppesen et al. 2001).
90
80
70
-
Fully Extended
-
Random Coil I
060
1050
40
230
Actual PEG
'U 20
10
0
0
2000
6000
4000
PEG mw
8000
10000
Figure 1.4: End to end distance of a PEG chain dependence on molecular weight (MW) where each
ethylene glycol repeat unit is 44 g/mol. Fully extended and random coil represent the longest and shortest
end to end distances respectively.
1.6
Modeling grafted PPLG and PEG crosslinked hydrogel
Consider as an example a grafted PPLG presented in Chapter 4 of this thesis, where PPLG having
a degree of polymerization of 160 is fully grafted with 2-(2-azidoethoxy)ethanol (E02) and on average
four longer PEG chains with a terminal crosslinking group (assuming here that each crosslinker is
conjugated through 10 ethylene glycol repeat units). This polymer can be termed PPLG-g(crosslinker 4EO2 156 ). Section 1.6.1 discusses in detail the expected structure of this example grafted
20
. ............
PPLG in solution. Section 1.6.2 outlines a model structure of a hydrogels formed from this polypeptide
crosslinked with PEG.
1.6.1
Modeling grafted PPLG in solution
Assuming that grafted PPLG adopts an a-helical secondary structure as outlined above, its
primary dimensions are radius and length. Approximations of both dimensions will be presented in this
section to generate a solution phase spatial model of grafted PPLG. The contour length of grafted PPLG is
defined by the polymer's degree of polymerization (Section 1.5.1). Dimensions of the radius of the
grafted PPLG depend on the character of the specific grafting groups, as demonstrated by the following
consideration of PPLG grafted with an E2 brush and with a PEGn= 10 longer linker (Figure 1.5).
The radius of the polypeptide is the sum of the helix, the glutamic acid side chain and triazole
ring, and the grafted functionalities. The a-helix internal diameter is estimated as 2.7 A as determined by
well-defined crystal structure (Langel et al. 2010). The overall radius is summed using a rough estimate of
the equivalent of five carbon bonds, comprising the glutamic acid extension, (5*1.26A) with 3.8 A length
of the 1,4-disubstituted 1,2,3- triazoles (Angell and Burgess 2007), totaling around 1.0 nm. The E02
brush is estimated as extending an additional 0.8 nm (R. L. C. Wang, Jtrgen Kreuzer, and Grunze 2000).
Together, these assumptions approximate the radius of PPLG fully grafted with an EQ brush as
approximately 2.1 nm (Figure 1.5C).
Considering the radius of the grafted crosslinkers, a PEGn=1O chain has a fully extended length of
3.6 nm as discussed above, suggesting a maximum radius for PPLG grafted crosslinkers as 5 nm. If, for
example, a peptide having 8 amino acids were grafted to the PEGn= 10 chain rather than a small
crosslinking molecule, the fully extended peptide would be expected to contribute another 3 nm to the
radius of grafted PPLG.
A.
B.
/
+0.8 nm E0 2
0
0
R-N
R(NJLPC
00
2.1 nm radius brush
NH,
0.3 nm helix
00
N.
rrmtoE0 2
N
N
helix
2rr to EO 2
0C.
.3nm
3.6 nm fully extended EON,
N
+ 2.8 nm 8 amino acids
7.7 nm peptide
3.6 side chains
rotation
HNper
X
Figure 1.5: A) chemical structures of a single monomer of PPLG grafted with E02 and PPLG grafted with
PEGn= 10 B) Model of these side chains on an PPLG backbone, shown 3 rotations and C) calculations of
PPLG radii for various grafting groups.
21
Figure 1.5 graphically summarizes these assumptions regarding the radial dimensions of grafted PPLG,
where PPLG grafted with E0 2 can be modeled with a radius of around 2.1 nm with longer grafted PEG
sidechains, crosslinkers, or peptides extending farther from the helical backbone. The length of an a-helix
having a DOP of 160 is estimated as 24 nm as discussed above.
PPLG crosslinking macromers used throughout this thesis use the following terminology to
describe the various grafting substitutions: (i) mono-functionality macromers are designated by PPLG-g(crosslinkerlEO2z) or PPLG-g-(crosslinker2,EO2,), where the number of crosslinkerlor crosslinker 2
functional groups (x) and inert, water solubilizing 2-(2-azidoethoxy)ethanol chains (z) grafted per PPLG
molecule were systematically varied to cover a range of functional groups/PPLG while ensuring PPLG
chains were fully grafted such that x + z = DP (ii) di-functional macromers grafted with both
crosslinkerland crosslinker2 functional groups, along with inert solubilizing E02 chains, are designated
by PPLG-g-(crosslinkerlcrosslinker2yEO2z) such that x + y + z = DP.
1.6.2
Modeling step-growth PEG hydrogels crosslinked through grafted PPLG
The solution phase model of PPLG structure introduced in the previous sections forms the
foundation for modeling a hydrogel crosslinked through this polypeptide. Specifically, we consider as an
example the polypeptide gel discussed in Chapter 4 of this thesis.
)
This example hydrogel is crosslinked from a precursor solution of 4 wt% PPLG-g-(crosslinker 4EO2 156
(with geometry as outlined in the previous section) and 1.3 wt% 4-arm PEG thiol- 10k in an aqueous
buffer. Here we also introduce a naming convention for star PEG polymer, n-arm PEG endgroup- mk, in
which PEG macromer having n arms each functionalized with the designated endgroup has a total
molecular weight of m thousand g/mol.
The molecular weight of the grafted PPLG can be calculated from the degree of polymerization
using NMR spectroscopy and the molecular weights of the grafting groups (eg. 160*289 per E0 2 grafted
repeat unit + crosslinker and PEG extensions=-50k/PPLG) and the molecular weight for 4-arm PEG
thiol-IOk is reported for each lot by the manufacturer, here 1 lk/4-arm PEG. From these molecular
weights can be calculated molarities of the PPLG and PEG macromers as 0.8 and 1.12 mM, respectively.
Using the precursor solution concentrations as an estimate for concentrations of the polymer in the gels
assumes minimal swelling as common for this gel system.
This thesis proposes the utility of visualizing a heterogeneous, two-component, hydrogel structure
having comparable macromer molarities as taking the conformation of an interpenetrating face center
cubic (FCC) crystal lattice. In cubic crystal theory, each constituent, here an individual crosslinking
polymer macromer, is represented as a sphere having a radius r, tight-packed into a three dimensional
22
structure. FCC was chosen from established cubic crystal lattices as it has the highest packing factor of
0.74, where the packing factor is the fraction of volume in a crystal structure that is occupied by
constituent particles. Modeling gel macromers as spheres in an interpenetrating FCC lattice is not
intended to be more rigorous than the close packed tetrahedral model presented by Flory, but rather to
provide a useful intuitive framework for visualizing gel structure. The FCC unit cell for this particular
polymer system is shown in Figure 1.6, with grafted PPLG represented as the blue sphere while green
spheres represent a second interpenetrating FCC structure of the 4-arm PEG crosslinkers.
If the polymers were evenly dispersed in an FCC cubic structure, having a 0.74 packing factor,
they would have unit length of 14 and 13 nm for PPLG and PEG, respectively. At first approximation,
this allows the mixture of polymers to be modeled as an interpenetrating FCC crystal structure with unit
lengths of 13-14 nm (Figure 1.6).
RPEG=10nm
Front
Face
DPPLG n=11
=4 nm
LPPLG_160
24 nm
SPEG=14 nm
SPPLG=l3
nm
Figure 1.6: Schematic of an idealized step-growth polypeptide gel modeled as an interpenetrating face
center cubic crystal structure of the PPLG and PEG crosslinkers. Gel is that explored in Chapter 4 of this
thesis having PPLG-g-(crosslinkerxEO2y) at 4% wt/vol polymer with 10k 4-arm PEG thiol at 1.3 wt%.
Isolating only the front face of the cube helps to visualize an ideal structure of polymers within
the gel (Figure 1.6). In this idealized structure, PPLG a-helices are shown in blue, having the rod like
structure developed above, randomly oriented on points of the blue FCC lattice. For simplicity, only one
4-arm PEG crosslinker is shown, representing PEG polymers centered on the green FCC lattice.
Estimated dimensions of the crosslinked PEG macromers could vary from radii of 2 nm, the random
length of an arm to 20 nm, the fully extended length. Recent small-angle neutron scattering (SANS)
measurements of 4-arm PEG amine molecules suggest that both in solution and in gels, 4-arm PEGs are
well modeled as shown in Figure 1.6, as spheres with hydrodynamic radii expanding to fill the available
23
space (Matsunaga et al. 2009). As a final approximation, a-helical PPLG cylinders modeled with the
radius of the E02 groups, are expected to fill 21% of the total gel volume.
PPLG as a non-structural handle for introducing (nanostructured) bio-functionality to
1.7
hydrogels
The utility of grafted polypeptides, including PPLG, as a structural component of crosslinked
hydrogels, is expected to extend to their applications as highly controlled handles presenting a wide
variety of peptides, proteins, and proteoglycans with nanoscale structural control.
Specifically considering PPLG, its highly defined stochastic substitution, organized secondary
structure, and remarkable synthetic robustness enables PPLG to function as a modular adapter integrating
the growing menu of biofunctional conjugating chemistries with hydrogel crosslinking chemistries. As
shown in Figure 1.7, PPLG might be grafted with functional groups supporting both covalent crosslinking
into a bulk PEG hydrogel and grafting on of biofunctionality. Such an approach of using PPLG as a
modular adapter for hydrogel biofunctionalization significantly streamlines functional gel synthesis by
eliminating constraints that biofunctional groups be directly functionalized with chemistries compatible
with gel crosslinking.
A.
B.
C.
Figure 1.7: PPLG (blue rod) grafted with ethylene oxide, crosslinking functionality and bio-functionality
(yellow rectangle and green circle). PPLG-grafted biofunctionality might be, A. grafted from a long
tether as single groups, B. grafted as clusters or C. grafted mixed functionality.
Further, grafted PPLG represents an intriguing handle for introducing into hydrogels clustered
functionality of single or multiple types of functionalities. Average PPLG functionality is expected to be
well-controlled by varying feed ratios of PPLG and grafting group as is demonstrated in this thesis for
crosslinking functionality. The size of PPLG-biofunctional clusters might be controlled by PPLG DOP
and the length of the linker. Additional background, motivation, and specific examples introducing PPLG
as a tool for nanoscale organization and biofunctional clustering in synthetic ECMs are included in
Section 5.2.3.2.
24
1.8
Conclusions
The literature review and basic modeling presented in Sections 1.2-1.7 develop a theoretical
framework for conceptualizing the crosslinking and structure of the step-growth PPLG-crosslinked
hydrogels presented in this thesis. These new hydrogels are suggested to be complementary to, but
distinct from, both established PEG gels systems crosslinked through only PEG macromers and
established polypeptide gel systems crosslinked through charged homopolymers. Significantly, PPLG
crosslinked gels will be demonstrated to have bulk properties influenced by the nanoscale structure of the
crosslinking a-helical polypeptides. PPLG grafted with short PEG molecules and crosslinkers is presented
as a biologically inert background gel to which a wide range of biofunctional groups can be incorporated
through organic and aqueous grafting. In this chapter, crosslinking PPLG has been introduced as a
structural gel crosslinker, but, as will be introduced in Section 3.1.3.2, grafted PPLG might also be
substituted as a small fraction of a bulk PEG hydrogel, functioning as a handle for more straightforward
integration of a variety of chemistries. Finally, hydrogels crosslinked through grafted polypeptides such
as those outlined here and demonstrated in the following chapters represent the natural intersection of
recent innovations in both grafted peptides NCA polymerization and step-growth hydrogels. Hydrogels
presented in this thesis, integrating these two polymer techniques, are expected to enable well-controlled,
modular hydrogels useful for engineering highly controlled synthetic extracellular matrices.
1.9
References
Angell, Yu L., and Kevin Burgess. 2007. "Peptidomimetics via Copper-Catalyzed Azide-Alkyne
Cycloadditions." ChemicalSociety Reviews 36 (10): 1674-89. doi:10.1039/b701444a.
Chakrabarti, Buddhapriya, and Alex Levine. 2005. "Nonlinear Elasticity of an a-Helical Polypeptide."
PhysicalReview E 71 (3): 031905. doi:10.1 103/PhysRevE.71.031905.
Cheng, Jianjun, and Timothy J. Deming. 2012. "Synthesis of Polypeptides by Ring-Opening
Polymerization of a -Amino Acid N-Carboxyanhydrides." Topics in CurrentChemistry 310 (June
2011): 1-26. doi:10.1007/128_2011_173.
Choe, Seungho, and Sean X Sun. 2005. "The Elasticity of Alpha-Helices." The Journalof Chemical
Physics 122 (24): 244912. doi:10.1063/1.1940048.
Chopko, Caroline M., Erika L. Lowden, Amanda C. Engler, Linda G. Griffith, and Paula T. Hammond.
2012. "Dual Responsiveness of a Tunable Thermo-Sensitive Polypeptide." ACS Macro Letters 1 (6):
727-31. doi:10.1021/mz300088w.
Corey, R. B., and L. Pauling. 1953. "Fundamental Dimensions of Polypeptide Chains." Proceedings of
the Royal Society B: BiologicalSciences 141 (902): 10-20. doi:10.1098/rspb.1953.0011.
Cotts, Patricia M., Timothy M. Swager, and Qin Zhou. 1996. "Equilibrium Flexibility of a Rigid Linear
Conjugated Polymer." Macromolecules 29 (23): 7323-28. doi: 10.102 1/ma9602583.
Deforest, Cole, Evan Sims, and Kristi S. Anseth. 2010. "Peptide-Functionalized Click Hydrogels with
Independently Tunable Mechanics and Chemical Functionality for 3D Cell Culture." Chemistry of
Materials 22 (21): 4783-90. doi:10.1021/cmO1l391y.
Deng, Chao, Jintian Wu, Ru Cheng, Fenghua Meng, Harm Anton Klok, and Zhiyuan Zhong. 2014.
"Functional Polypeptide and Hybrid Materials: Precision Synthesis via A-Amino Acid NCarboxyanhydride Polymerization and Emerging Biomedical Applications." Progressin Polymer
Science 39 (2). Elsevier Ltd: 330-64. doi: 10.101 6/j.progpolymsci.2013.10.008.
25
Engler, Amanda C., Hyung-il Lee, and Paula T. Hammond. 2009. "Highly Efficient 'Grafting onto' a
Polypeptide Backbone Using Click Chemistry." Angewandte Chemie (InternationalEd. in English)
48 (49): 9334-38. doi:10.1002/anie.200904070.
Engler, Amanda C., Anita Shukla, Sravanthi Puranam, Hilda G. Buss, Nina Jreige, and Paula T.
Hammond. 2011. "Effects of Side Group Functionality and Molecular Weight on the Activity of
Synthetic Antimicrobial Polypeptides." Biomacromolecules 12 (5): 1666-74.
doi: 10.102 1/bm2000583.
Gandavarapu, Navakanth R., Malar Azagarsamy, and Kristi S. Anseth. 2014. "Photo-Click Living
Strategy for Controlled, Reversible Exchange of Biochemical Ligands." Advanced Materials26:
2521-26. doi:10.1002/adma.201304847.
Gyenes, Tamis, Viktoria Torma, Benjamin Gyarmati, and Mikl6s Zrinyi. 2008. "Synthesis and Swelling
Properties of Novel pH-Sensitive Poly(Aspartic Acid) Gels." Acta Biomaterialia4: 733-44.
doi: 10. 1016/j.actbio.2007.12.004.
Huang, Yugang, Yonghong Zeng, Jianwen Yang, Zhaohua Zeng, Fangming Zhu, and Xudong Chen.
2011. "Facile Functionalization of Polypeptides by Thiol-Yne Photochemistry for Biomimetic
Materials Synthesis." Chemical Communications (Cambridge, England) 47 (26): 7509-11.
doi:10.1039/clcc12177d.
Jeppesen, Claus, Joyce Y. Wong, Tonya L. Kuhl, Jacob N. Israelachvili, Nasreen Mullah, Samuel
Zalipsky, and Carlos M. Marques. 2001. "Impact of Polymer Tether Length on Multiple LigandReceptor Bond Formation." Science (New York, N. Y) 293 (5529): 465-68.
doi: 10.1 126/science.293.5529.465.
Kramer, Jessica R., and Timothy J. Deming. 2012. "Preparation of Multifunctional and Multireactive
Polypeptides via Methionine Alkylation." Biomacromolecules 13: 1719-1723.
doi: 10.1021/bm300807b.
Kricheldorf, Hans R. 2006. "Polypeptides and 100 Years of Chemistry of Alpha-Amino Acid NCarboxyanhydrides." Angewandte Chemie (InternationalEd. in English) 45 (35): 5752-84.
doi: 10. 1002/anie.200600693.
Langel, Ulo, Tamas Bartfai, Benjamin F. Cravatt, Nils Arne Elofsson, and Astrid Graslund. 2010.
Introduction to Peptides and Proteins. Boca Raton, FL: Taylor and Francis Group.
Lee, Hwankyu, Richard M. Venable, Alexander D. Mackerell, and Richard W. Pastor. 2008. "Molecular
Dynamics Studies of Polyethylene Oxide and Polyethylene Glycol: Hydrodynamic Radius and
Shape Anisotropy." BiophysicalJournal95 (4). Elsevier: 1590-99.
doi: 10.1529/biophysj. 108.133025.
Lu, Hua, Jing Wang, Yugang Bai, Jason W Lang, Shiyong Liu, Yao Lin, and Jianjun Cheng. 2011. "Ionic
Polypeptides with Unusual Helical Stability." Nature Communications 2 (January). Nature
Publishing Group: 206. doi: 10.103 8/ncomms 1209.
Lu, Hua, Jing Wang, Ziyuan Song, Lichen Yin, Yanfeng Zhang, Haoyu Tang, Chunlai Tu, Yao Lin, and
Jianjun Cheng. 2014. "Recent Advances in Amino Acid N-Carboxyanhydrides and Synthetic
Polypeptides: Chemistry, Self-Assembly and Biological Applications." Chemical Communications
(Cambridge, England) 50 (2): 139-55. doi:10.1039/c3cc46317f.
Markland, Peter, Yuehua Zhang, Gordon L. Amidon, and Victor C. Yang. 1999. "A pH- and Ionic
Strength-Responsive Polypeptide Hydrogel: Synthesis, Characterization, and Preliminary Protein
Release Studies." Journal ofBiomedicalMaterialsResearch 47 (4): 595-602.
doi: 10.1002/(SICI) 1097-4636(19991215)47:4<595::AID-JBM 1 7>3.0.CO;2-I.
Matsunaga, Takuro, Hanako Asai, Yuki Akagi, Takamasa Sakai, Ung I Chung, and Mitsuhiro
Shibayama. 2011. "SANS Studies on Tetra-PEG Gel under Uniaxial Deformation." Macromolecules
44: 1203-10. doi:10.1021/ma102658e.
Matsunaga, Takuro, Takamasa Sakai, Yuki Akagi, Ung-il Chung, and Mitsuhiro Shibayama. 2009.
"Structure Characterization of Tetra-PEG Gel by Small-Angle Neutron Scattering."
Macromolecules42 (4): 1344-51. doi: 10.102 1/ma802280n.
26
-
Moore, Nathan W., and Tonya L. Kuhl. 2006. "The Role of Flexible Tethers in Multiple Ligand-Receptor
Bond Formation between Curved Surfaces." BiophysicalJournal91 (5). Elsevier: 1675-87.
doi:10.1529/biophysj.105.079871.
Oelker, Abigail M., Shannon M. Morey, G Griffith, Paula T. Hammond, and Linda G. Griffith. 2012.
"Helix versus Coil Polypeptide Macromers: Gel Networks with Decoupled Stiffness and
Permeability." Soft Matter 8 (42): 10887. doi: 10.1039/c2sm26487k.
Oesterhelt, F., M. Rief, and H. E. Gaub. 1999. "Single Molecule Force Spectroscopy by AFM Indicates
Helical Structure of Poly(ethylene-Glycol) in Water." New JournalofPhysics 1 (January): 6-6.
doi: 10.1088/1367-2630/1/1/006.
Oliveira, Eder D., Samuel G. Hirsch, Richard J. Spontak, and Stevin H. Gehrke. 2003. "Influence of
Polymer Conformation on the Shear Modulus and Morphology of Polyallylamine and Poly(a-L
Lysine) Hydrogels." Macromolecules 36 (16): 6189-6201. doi: 10.102 1/ma021468k.
Quadir, Mohiuddin, Mackenzie Martin, and Paula T. Hammond. 2014. "Clickable Synthetic Polypeptides
- Routes to New Highly Adaptive Biomaterials." Chemistry ofMaterials 26: 461-76.
doi: 10.1021/cm4023997.
Papadopoulos, P., G. Floudas, H.-A. Klok, I. Schnell, and T. Pakula. 2004. "Self-Assembly and
Dynamics of Poly(gamma-Benzyl-L-Glutamate) Peptides." Biomacromolecules 5 (1): 81-91.
doi:10.1021/bm034291q.
Rhodes, Allison J., and Timothy J. Deming. 2013. "Soluble, Clickable Polypeptides from AzideContaining N -Carboxyanhydride Monomers." A CS Macro Letters 2 (5): 351-54.
doi:10.1021/mz4001089.
Rixman, M. A., D. Dean, and C. Ortiz. 2003. "Nanoscale Intermolecular Interactions between Human
Serum Albumin and Low Grafting Density Surfaces of Poly(ethylene Oxide)." Langmuir 19 (22):
9357-72. doi:10.1021/1a0340571.
Shih, Ke-ying, Tai-shen Hsiao, Shiang-lin Deng, and Jin-long Hong. 2014. "Water-Soluble Poly(yPropargyl- L -Glutamate) Containing Pendant Sulfonate Ions and Terminal Fluorophore:
Aggregation-Enhanced Emission and Secondary Structure." Macromolecules 47 (12): 4037-47.
doi: 10.1021/ma500886v.
Singh, Anirudha, Jianan Zhan, Zhaoyang Ye, and Jennifer H. Elisseeff. 2013. "Modular Multifunctional
Poly(ethylene Glycol) Hydrogels for Stem Cell Differentiation." Advanced FunctionalMaterials 23
(5): 575-82. doi:10.1002/adfm.201201902.
Tanford, Charles., Kazuo Kawahara, and Savo Lapanje. 1966. "Proteins in 6-M Guanidine
Hydrochloride. Demonstration of Random Coil Behavior." The JournalofBiological Chemistry 241
(8): 1921-23. PMID: 5947952.
Tang, Haoyu, and Donghui Zhang. 2011. "Multi-Functionalization of Helical Block Copoly(a-Peptide)s
by Orthogonal Chemistry." Polymer Chemistry 2: 1542. doi:10.1039/clpyOO015b.
Temyanko, Elena, Paul S. Russo, and Holly Ricks. 2001. "Study of Rodlike Homopolypeptides by Gel
Permeation Chromatography with Light Scattering Detection: Validity of Universal Calibration and
Stiffness Assessment." Macromolecules 34 (3): 582-86. doi:10.1021/maOO1351 1.
Tibbitt, Mark W., April M. Kloxin, Lisa a. Sawicki, and Kristi S. Anseth. 2013. "Mechanical Properties
and Degradation of Chain and Step-Polymerized Photodegradable Hydrogels." Macromolecules 46
(7): 2785-92. doi:10.1021/ma302522x.
Trappmann, Britta, Julien E. Gautrot, John T. Connelly, Daniel G. T. Strange, Yuan Li, Michelle L.
Oyen, Martien Cohen Stuart, et al. 2012. "Extracellular-Matrix Tethering Regulates Stem-Cell
Fate." Nature Materials 11 (7). Nature Publishing Group: 642-49. doi:10.1038/nmat3339.
Wang, Jing, Hua Lu, Yuan Ren, Yanfeng Zhang, Martha Morton, Jianjun Cheng, and Yao Lin. 2011.
"Interrupted Helical Structure of Grafted Polypeptides in Brush-like Macromolecules."
Macromolecules 44: 8699-8708. doi: 10.102 1/ma201833b.
Wang, Richard L., H. JArgen Kreuzer, and Michael Grunze. 2000. "The Interaction of Oligo(ethylene
Oxide) with Water: A Quantum Mechanical Study." Physical Chemistry Chemical Physics 2 (16):
3613-22. doi:10.1039/b002593n.
27
Wen, Qi, and Paul Janmey. 2013. "Effects of Non-Linearity on Cell-ECM Interactions." Experimental
Cell Research 319 (16). Elsevier: 2481-89. doi: 10.1016/j.yexcr.2013.05.017.
Xiao, Chunsheng, Changwen Zhao, Pan He, Zhaohui Tang, Xuesi Chen, and Xiabin Jing. 2010. "Facile
Synthesis of Glycopolypeptides by Combination of Ring-Opening Polymerization of an AlkyneSubstituted N-Carboxyanhydride and Click 'Glycosylation'." MacromolecularRapid
Communications 31 (11): 991-97. doi: 10.1002/marc.200900821.
Zegarra, Fabio C., Gian N. Peralta, Alberto M. Coronado, and Yi Qin Gao. 2009. "Free Energies and
Forces in Helix-Coil Transition of Homopolypeptides under Stretching." Physical Chemistry
Chemical Physics: PCCP 11: 4019-24. doi: 10. 1039/b820021a.
Zhang, Zhe, Li Chen, Mingxiao Deng, Yunyan Bai, Xuesi Chen, and Xiabin Jing. 2011. "Biodegradable
Thermo- and pH-Responsive Hydrogels for Oral Drug Delivery." JournalofPolymer Science Part
A: Polymer Chemistry 49 (13): 2941-51. doi:10.1002/pola.24730.
Zhu, Junmin, Jeffrey A. Beamish, Chad Tang, Kandice Kottke-Marchant, and Roger E. Marchant. 2006.
"Extracellular Matrix-like Cell-Adhesive Hydrogels from RGD-Containing Poly(ethylene Glycol)
Diacrylate."Macromolecules 39 (4): 1305-7. doi:10.1021/ma052333s.
Zustiak, Silviya P., Rohan Durbal, and Jennie B. Leach. 2010. "Influence of Cell-Adhesive Peptide
Ligands on Poly(ethylene Glycol) Hydrogel Physical, Mechanical and Transport Properties." Acta
Biomaterialia6 (9). Acta Materialia Inc. 3404-14. doi: 10.101 6/j.actbio.2010.03.040.
28
2
Solution-phase behavior of grafted PPLG with multiple grafting groups
This chapter demonstrates that the temperature- and pH- dependent solubility of poly(y-
propargyl-L-glutamate) (PPLG) functionalized through a copper-catalyzed 1,3-cycloaddition reaction
between an alkyne and an azide can be tuned with precision over a broad range of conditions by varying
the ratio of substitution of short oligo(ethylene glycol) and diisopropylamine side groups.
By characterizing the solution phase behavior of grafted PPLG, this chapter establishes several
foundational findings directing subsequent exploration of PPLG hydrogels. Most significantly, this
chapter introduces grafted 2-(2-azidoethoxy)ethanol (E0 2) as sufficient to solubilize PPLG, extending
previous studies which had achieved PPLG aqueous solubility by grafting long PEG polymers (Engler,
Lee, and Hammond 2009) or charged groups (Engler, Bonner, et al. 2011 a). For solubilizing PPLG
crosslinked into hydrogels, grafting short E02 groups is favored over established techniques as this group
is expected not to sterically hinder access to grafted biocues or to mediate significant non-specific protein
binding. Not presented here, PPLG grafted with a single ethylene oxide group (EO) was not water
)
soluble. For the purpose of solubilizing PPLG crosslinkers, 1 -azido-2-(2-methoxyethoxy) ethane (mEO 2
was initially expected to be more useful than grafted E02 because the distinct 'H NMR signature of the
mEO 2 methyl peaks aid straightforward quantification of mEO 2 grafting ratios. However, temperature
dependent solubility was observed for PPLG grafted with mEO 2, as reported in this chapter. Grafting E02
and the more hydrophilic 2-(2-(2-azidoethoxy)ethoxy)ethanol (E03) were used for the remainder of this
thesis.
Furthermore, solution phase studies presented in this chapter strongly support the stochastic
nature of PPLG grafting, which varies both the average ratio of substitution and the average polymer
properties. This observation suggests both that PPLG grafting can be well-modeled as stochastic and that
bulk properties of crosslinking gels might be systematically tuned by varying the ratio of substitution of
crosslinking groups.
Finally, extensive characterization of grafted PPLG by circular dichoism confirms the robust
stability of PPLG's a-helical secondary structure, grafted with E2 and with destabilizing amines. The
unusually stable secondary conformation of mEO 2 and E0 2 grafted PPLG is expected to confer bulk and
local properties to PPLG-crosslinked hydrogels not demonstrated in gels crosslinked only through
poly(ethelene glycol), a random coil.
29
The following has been previously published:
Caroline M. Chopko, Erika L. Lowden, Amanda C. Engler, Linda G. Griffith, and Paula T.
Hammond. "Dual responsiveness of a tunable thermosensitive polypeptide." ACS macro letters 1,
no. 6 (2012): 727-73 1.
2.1
Introduction
While multi-stimuli responsive polymers have been widely researched, the medical application of
polymers that respond quickly and reversibly to changes in both environmental pH and temperature has
been especially investigated (Schmaljohann 2006; Dimitrov et al. 2007). Efforts to fabricate multiplexed
tunable polymer systems have built on single stimuli responsive design strategies in seeking to adjust a
macromolecular properties of a polymer by colocalizing different types of responsive elements in a
random (Yamamoto, Pietrasik, and Matyjaszewski 2008), block (Schilli et al. 2004), or graft (Durand and
Hourdet 1999) copolymer.
While poly(N-isopropylacrylamide) (PNIPAm) is the most established thermoresponsive
polymer, polymers based on short oligo(ethylene glycol) side chains have developed as an alternate nonfouling biocompatible thermoresponsive platform (Lutz 2008). Han and colleagues established that the
solubility and cloud point of comb-like methacrylate polymers can be modified by controlling the length
of the oligo(ethyleneglycol) side chains (Han, Hagiwara, and Ishizone 2003). The observed cloud point of
the homopolymer could be systematically varied from 28 'C to 90 'C by adjusting the ratio of a 2-(2methyloxyethyloxy)ethyl methacrylate to poly(ethylene glycol) methyl ether methacrylate in a random
copolymer (Lutz and Hoth 2006). More recent efforts have expanded methacrylate thermoresponsive
platforms by the incorporation of additional functional groups. Strategies have focused on introducing
small fractions of comonomers with reactive side chains that can be grafted post-polymerization,
including alkynes (Leung et al. 2011; Jung et al. 2011) and allyl pendant functionalities. Hydroxyl
terminated oligo(ethylene glycol) methacrylate monomers have been incorporated, with 2-hydroxyethyl
methacrylate, both as a functional handle and as a modifier to tune the copolymer's lower critical solution
temperature (LCST) properties (Laloyaux et al. 2010). Short ester oligo(ethylene glycol) side chains have
also been used to confer thermoresponsiveness to polypeptides with defined a-helical secondary structure
(Yu et al. 1999; Chen, Wang, and Li 2011).
We have recently introduced the N-carboxyanhydride (NCA) polymerization of poly(y-propargyl
-L-glutamate) (PPLG) as a particularly attractive platform for exploring multiplexed functionality, as it
allows highly efficient post-polymerization grafting through copper catalyzed "click reaction" to an ahelical polypeptide backbone (Engler, Lee, and Hammond 2009). The PPLG backbone has since been
demonstrated as the basis for readily functionalized, biocompatible, biodegradable polypeptides with
30
either thermal (Ding et al. 2011) and pH (Engler, Bonner, et al. 201 ib) responsiveness. This chapter
demonstrates that these two properties can be combined through the utility of PPLG grafted with a
combination of short oligo ethylene glycol side chains that yield a thermo-responsive highly tunable
polypeptide and tertiary amine groups that confer pH sensitivity in a biologically relevant range. The
potential for facile, multiplexed conjugation of responsive functional groups is demonstrated through the
generation of a family of polymers which can be adjusted to undergo solubilisation at very specific and
relevant pH and temperature.
The strategy employed in this study is illustrated in Figure 2.1. A series of PPLG backbones is
fully substituted with varied ratios of 2-(2-azidoethoxy)ethanol (EG 2) and I-azido-2-(2-methoxyethoxy)
ethane (mEG 2). The highly robust click chemistry offers the potential to multiplex optimized thermal and
pH sensitivity with orthogonal functionality including those that confer bioactivity (Parrish, Breitenkamp,
and Emrick 2005; Tang and Zhang 2011; Xiao et al. 2010). This chapter establishes PPLG grafted with
both mEO 2 and N-(2-azidoethyl)-N-isopropylpropan-2-amine (diisopropylamine) as one example of a pH
and temperature responsive polymer platform. More broadly, it establishes the systematic, robust, and
facile grafting of PPLG with multiple side groups conferring complementary aggregated functionality.
While this chapter focuses on controlling the responsiveness of aqueous PPLG solutions, the flexibility of
mono or bi functional amine-initiated NCA polymerization of the polymer backbone suggests the facile
deployment of the modular functionalized polypeptides in such applications as block copolymers (Engler,
Lee, and Hammond 2009), surface coatings, drug delivery systems or responsive hydrogels (Dimitrov et
al. 2007).
0
0
R
H2 "Click"
N
H
NH 2
RIN
R -N 3
0
00
0
R=(CH 2)5CH 3
N3 "
R.
0"O"CH
mE0 2
N3
O
N3
N
OH
Diisoproplyamine
E0 2
Figure 2.1: Post polymerization poly(y-propargyl-L-glutamate) backbones are fully substituted with
thermo and pH responsive side groups.
2.2
Materials and Methods
L-(+)-Glutamic acid 99% minimum was purchased from EMD Chemicals (Gibbstown, NJ). 2-(2-
azidoethoxy)ethanol (E02 ) (Sinha, Sahoo, and Kumar 2009), and N-(2-azidoethyl)-N-isopropylpropan-231
amine (diisopropylamine) (Engler, Shukla,. et al. 2011) were synthesized as previously reported. All other
chemicals were purchased from Sigma-Aldrich (St. Louis, MO). All materials were used as received.
1
H NMR spectra were recorded on Bruker 400 MHz FT-NMR spectrometers. Gel permeation
chromatography (GPC) measurements were carried out using a Waters Breeze 1525 HPLC system
equipped with two Polypore columns operated at 75 'C, series 2414 refractive index detector, series 1525
binary HPLC pump, and 717 plus autosampler. DMF with 0.01 M LiBr was the eluent for analysis, and
samples were dissolved at 4-6 mg/mL in DMF. Data collection and processing was performed using
Waters' Breeze chromatography software version 3.30. The average molecular weight of the sample was
calibrated against narrow molecular weight poly(methyl methacrylate) (PMMA) standards.
Temperature responsive solubility was measured by a Cary 500i UV-Vis-NIR Dual-Beam
Spectrophotometer automated Multi-Cuvette Sampling Accessory with Temperature Controller. The
polymer solutions in cuvettes with a 10 mm pathlength at concentrations of 0.1 to 3 mg/ml were
equilibrated for at least 5 minutes, heated from 20 'C to 70 'C at a rate of 1 'C/min, held at 70 'C for 3
minutes, and returned to 20 'C. Reported temperatures correspond to the block temperature. A measured
by a temperature probe, after ~2 degrees of a heating or cooling cycle, the block temperature consistently
led the solution temperature by 1.6 'C +0.2 'C. The reported cloud point was defined as the temperature
corresponding to 50% of the range of transmittance.
Circular dichroism spectroscopy of polymer solutions was performed using an Aviv model 202
CD spectrometer. Samples were prepared at a concentration of 0.5-1.1 mg mL- in Milli-Q water.
Measurements were sampled every nm with a 3-5 s average time over the range of 195-260 nm
(bandwidth 1% 1.0 nm) at temperatures
0.1 C the reported temperature ranging from 25 'C to 50 'C.
Measurements were taken using a cell with a 1 mm path length.
Titration of diisopropylamine mEO 2 grafted PPLG was performed on 3 ml solution at 2.5 mM
amine solution in 125 mM NaCl adjusted to pH 3. The solution was titrated with 5 pL aliquots of 0.1 M
NaOH, and the pH was measuring after each addition. Titration of PPLG graft 11% diisopropylamine was
performed at 0.25 mM amine with 0.01 M NaOH.
Poly(y-propargyl-L-glutamate)(PPLG). PPLG was synthesized as previously reported (Engler,
Lee, and Hammond 2009).
mEO2. In brief, 2.75 g (42 mmol) of sodium azide was added to an aqueous solution of 2 g (10.6
mmol) 1 -bromo-2-(2-methoxyethoxy) ethane. The solution was heated at 75 'C for 15 hours, then treated
with 5% NaOH solution and extracted 4 times with 10 mL of diethylether. The combined organic layer
was dried over Na 2SO 4 and concentrated to give a clear viscous liquid. Typical yield is 46%. 'H NMR
32
13
C NMR (CDC1 3
,
(CDC 3, 6ppm) 3.37 (s, 3H), 3.38 (t, 2H, J = 5.24 Hz), 3.54 (m, 2H), 3.63 (m, 4H).
6ppm) 72.14, 70.82, 70.22, 59.32, 50.83. HRMS (ESI-TOF) calcd for C 5H1 1 O 2 N3Na : 168.0743
([M+Na]), found 168.0750.
Azide-functionalized PPLG. A typical procedure includes grafting onto the PPLG backbone azide
terminated side chain at the target ratio feed ratio of alkyne/azide/CuBr/N,N,N',N',N"pentamethyldiethylenetriamine (PMDETA) equal to 1/1.2/0.1/0.1. PPLG (0.025 g, 0.150 mmol alkyne
repeat units), azide (0.165 mmol), and PMDETA (3.1 gL, 0.015 mmol) were all dissolved in DMF (3
mL). Azide terminated side chains used in this study include mEO 2, E0 2 and diisopropylamine. The CuBr
catalyst (0.002 g, 0.015 mmol) was added to the degassed solution, and the reaction solution was stirred at
room temperature. After more than 2 hours, the reaction solution was precipitated in 40 mL cold
diethylether, dissolved in 10 mL distilled water, and incubated for 30 min with 5 mg Dowex M4195
sulfate copper chelating resin. The beads were removed by filtration and the polymer solution was
dialyzed against water acidified by HCl (pH <4) for 24 hours and against distilled water for 12 hours. The
polymer structure and degree of substitution was verified by 'H-NMR. Typical yield is 75%.
2.3
Results and Discussions
Four PPLG backbones were synthesized by NCA polymerization as recently reported (Engler,
Lee, and Hammond 2009). These backbones have DPs (degree of polymerization) ranging from 44 to 97,
as calculated by NMR (Table 1). Varied ratios of E0 2 and mEO 2 grafting groups were introduced to tune
the polymer's temperature responsiveness. Ratios of E0 2 and mEO 2 grafting groups, calculated through
'H NMR (Figure 2.S2a), closely match that of the feed ratios (Figure 2.2).
Poly
mer
PPLG
-44
PPLG
-64k
PPLG
PD1'
1-13
Feed Ratio
Substitution of
EO:
0, 20, 40%
NMR-based
Substitution
of E0 2
0. 21. 38%
16300
1.33
0, 20, 40%
0. 16, 39%
77
20300
123
0, 20,40%
0, 19, 40%
97
28100
1.14
0. 25, 50,
DP by
NMR
44
Mn'
9500
64
- 77
PPLG
0,24,51,
100%
100%
ODMF GPC with 0,1% lithium bromide with PMMA standards,
5
Additionally substituted with various ratios diisopropylamine.
IH-NMR-observed percent substitution of E02 in backbone with remaining fully substituted with mEO.
-97
Figure 2.2: Inventory of grafted PPLG polymers
By visual inspection, all polymers were soluble at room temperature and only the polymer fully
substituted with E0 2 did not exhibit an LCST, and became turbid when heated in deionized water.
The effects of the backbone degree of polymerization and the ratio of methoxy and hydroxyl terminated
grafted side chains on the temperature responsiveness of the PPLG-based platform was characterized by
temperature controlled turbidity measurements of the 13 polymers outlined in Figure 2.3. All polymers
33
were characterized at a constant 3 mg/mi in deionized water, as their thermal responsiveness is
concentration dependent (Figure 2.S3) (Cheng et al. 2011; Engler, Shukla, et al. 2011). Figure 2.3 shows
observed cloud points, with reported temperatures corresponding to that of solutions having 50% of the
range of transmittance.
When comparing polymers of similar grafting ratios, all polymers maintain the expected trend of
decreasing solubility with increasing molecular weight (Figure 2.S4) (Cheng et al. 2011; Engler, Shukla,
et al. 2011). For each backbone length, increasing the fraction of more hydrophilic E0 2 side chains
predictably increases the cloud point of the fully functionalized polymer (Figure 2.3 and Figure 2.4). Even
the longest tested polymer, PPLG-97, functionalized fully with E0 2 remained soluble at all tested
temperatures further suggests that the mEO2 functionality specifically confers observed
thermoresponsiveness (Figure 2.4). The observed thermo-responsive behavior is reversible under
experimental conditions, with the PPLG-64 100% mEG 2 showing consistent measured turbidity values
across 4 heating and cooling cycles (Figure 2.S5). Further, all sample heating and cooling transmittance
curves showed minimal hysteresis, with cloud points varying by less than 3 'C (Figure 2.S6a).The robust
dependence of cloud point on grafting group composition across all four backbone lengths demonstrates
the ability to systematically tune a given backbone's temperature responsiveness at biologically relevant
temperatures by varying only the ratio of hydroxyl and methoxy terminated short PEG grafting groups.
This approach modulates temperature sensitivity while preserving much of the polymer's overall
structural and chemical identity.
65
60
-
PPLG-44
55
-
A PPLG-64
+, 50
-
U
o
mPPLG-77
PPLG-97
045
S40
A
40
jo35
30
25
0
10
30
20
40
50
60
% Substitution E02
Figure 2.3: Relationship between cloud point (measured at 3 mg/ml in deionised water) and percent E02
substituted onto PPLG backbones functionalized with mEO 2 for polypeptides having DOPs of n, PPLG-n.
34
____ - - - -
. , 11
Cr--- -, .
......
..I.."
.......
...
100
80
-
E02
E02
*40% E02
6A
*100% E02
0%
a
A 19%
60
E
-
A
4020
20
30
40
50
60
70
Temperature (*C)
Figure 2.4: Influence of temperature on the light transmittance (500 nm, heating 1 'C min-') of PPLG-97
substituted with 100% E02 and PPLG-77 substituted with E02 at 19%, 40%, and 100% and remainder
with mEO 2 at 3 mg/ml in deionised water.
All cloud points reported in Figure 2.3 were generated from heating curves; for most systems, the
transmittance decreases from 95% to 5% of the initial value took place over a range of less than 8 'C,
with some transitions as sharp as 2'C. The sharpness of these transitions is unusual for a synthetic
polymer, which typically would exhibit broader transitions due to a spread of molecular weights and side
chain functionalization efficiencies; we attribute these sharper transitions to narrow molecular weights
and well-defined grafting efficiencies. The exceptions included three polymers, PPLG-97 50%, PPLG-44
20%, and PPLG-44 40%, which exhibited less defined transitions spanning around 15' C. The less
discrete temperature responsiveness of polymers functionalized with mixed grafting groups supports the
anticipated result of the stochastic functionalization of the PPLG backbones. When all else is constant,
grafting is expected to yield a stochastic, binomial distribution, and one would dictate that short polymers
and those with equal substitution of the grafting groups would generate a population with the highest
variance and most gradual temperature response (Figure 2. S6).
For application in biological systems, the effects of salt must also be considered. Salts are broadly
understood to lower the cloud point of polymers by disrupting the surrounding hydration structure in a
process known as "salting out." As an initial characterization, the temperature response of the PPLG-64
polymers at 3 mg/ml in PBS was compared to that in deionized water. Predictably, all three polymers
demonstrate cloud points in physiological buffer that are lower than those in pure water, though the
magnitude of this decrease seems dependent on the individual polymer's composition, ranging from 3 'C
to 9 'C (Table SI).
A dual responsive system with temperature and pH responsiveness was investigated. PPLG was
grafted with mEO 2 and diisopropylamine, a group conferring PPLG the ability to undergo a solubility
phase transition with decreasing degree of amine ionization (Engler, Bonner, et al. 201 lb). Four grafted
35
.
........
...
......
........
..
PPLG-64 polymers were synthesized including polymers with feed ratios of 100% diisopropylamine,
50% diisopropylamine and 50% E0 2, 50% diisopropylamine and 50% mEO 2, and 10% diisopropylamine
and 90% mEO 2, with calculated grafted values closely matching feed ratios (Figure 2.S2b).
All polymers grafted with diisopropylamine were shown to buffer between pH 6.1 and pH 7.2.
PPLG grafted with only diisopropylamine buffered slightly lower than polymers having partial
diisopropylamine grafting, while polymers grafted with only mEO 2 showed no buffering above pH 4
(Figure 2.S8). 100% diisopropylamine grafted PPLG was insoluble above pH 6.3 while polymers with
only partial diisopropylamine grafting precipitated around pH 6.7.
Temperature controlled turbidity measurements of PPLG-64 grafted with 100% diisopropylamine
demonstrated robust thermal responsiveness with linear dependence on solution pH between pH 5.0 and
pH 6.2, having cloud points ranging from 66 'C to 38 'C (Figure 2.S7 and Figure 2.5). This temperature
responsiveness is not surprising considering the polymer's structural similarities to other known amine
containing temperature responsive polymers (Jung et al. 2011).
A
75
50% amine 50% mEO2
*10% amine 90% mEO2
A 50% amine 50% E02
100% amine
* 100% mEO2
70
65
ao60
-
4J
C
55
o
50 04545 A
A
A
-0
U
35
30
25
5.5
pH
6
6.5
7
Figure 2.5: Relationship between cloud point of grafted PPLG backbones and solution pH measured at 1
mg/ml in 100mM NaCl, 75mM phosphate buffer.
Cloud point measurements of diisopropylamine grafted PPLG polymers between pH 5.6 and pH
6.6 confirm additive properties of dual grafting (Figure 2.5). In solutions having pH values below the
amine's buffering region, grafted diisopropylamine increased PPLG's solubility compared to PPLG
grafted with only mEO 2. Specifically, at pH 5.6, PPLG graft 100% mEO 2 had the lowest tested cloud
point and PPLG graft 50% diisopropylamine and 50% mEO 2 was more soluble than PPLG graft 10%
diisopropylamine and 90% mEO 2 (Figure 2.5). However, with increased solution pH, grafted tertiary
amines become less protonated and more hydrophobic, leading to a reverse order of cloud points at pH
6.6, where polymers grafted with diisopropylamine exhibit lower cloud points than polymers grafted with
only mEO 2. Finally, PPLG graft 50% amine and 50% E0 2 demonstrated both thermal responsiveness and
36
.. .............
pH sensitivity but had the highest tested cloud points, confirming both grafted diisopropylamine's
contribution to thermal sensitivity and the additive bulk properties of dual PPLG functionalization.
The effect of grafting multiple groups on the secondary structure of the grafted PPLG was
= 208
explored using circular dichroism (CD), specifically by monitoring the relative absorbances at k
and 222 nm, characteristic minima of a-helical secondary structures. In both distilled water and buffered
salt conditions, CD spectra of polymers functionalized only with combinations of mEO 2 and E0 2
consistently maintained minima at 208 nm and 222 nm when heated though their cloud points (Figure
2.S9). Decrease in the heated solution's signal intensity is attributed to observed precipitation, and ahelical structure is only confirmed for the soluble fraction. At 250 C in all tested buffers ranging from pH
5.6 to pH 6.6, spectra of PPLG graft 100% and 50% diisopropylamine were also characteristically ahelical. However, as these polymers were heated to their cloud point, their CD absorbance at 208 nm
increased relative to 222 nm suggesting a shift from a pure a-helical to a partially random coil secondary
structure (Figure 2.6A and Figure 2.S10). At the same pH conditions, PPLG graft 10% diisopropylamine
and 90% mEO 2 maintained minimum at 208 and 222 nm as the temperature of the polymer solutions were
raised at least 7' C past their cloud point, (Figure 2.6B) suggesting that carefully selected dual grafting
designs can also direct polymer secondary structure.
B
A
60
E
025*
60
40
30* C
20
40*C
-0oA
-20
A 35* C
40* C
35* C
42
15
215
2
255~2
2
452
- 1155
''
C
30*C
0 40
00
Co
0250
IV
C
i
-
-_
x-40
-60
Wavelength (nm)
Wavelength (nm)
Figure 2.6: Circular dichroism spectra of PPLG-64 measured at I mg/ml in 100mM NaCl, 75mM
phosphate buffer pH 6.2 at indicated temperatures. A)50% mEO 2 and 50% diisoproplyamine) B)90%
mEO 2 and 10% diisoproplyamine.
For applications in drug delivery and nanomedicine, the dual pH and temperature sensitivity of
functionalized PPLG can be optimized over a broad and biologically relevant range of values to amplify
or tailor polymer responsiveness to feed different biological microenvironments simply by adjusting the
ratio of mEO 2 to diisopropylamine grafting groups or by introducing additional grafting groups. E0 2
might be substituted to lower the observed cloud point of the polypeptide solution while grafting on
37
alternative amines such as diethylamine can tune the polymer's buffering capacity and related pH
responsiveness (Engler, Bonner, et al. 201 lb).
2.4
Conclusion
In conclusion, we have demonstrated the synthesis of a library of reversibly thermoresponsive a-
helical polypeptides with readily tunable temperature dependent solubility at biologically relevant
conditions. In both distilled water and PBS, the cloud point of a given polypeptide is increased by
increasing the substitution of short hydroxyl terminated grafted PEG oligomers on PPLG backbones
substituted with short methoxy terminated PEG side chains. Furthermore, by incorporating the
diisopropylamine side group, we demonstrated a dual responsive system where the solubility and
secondary structure can be tuned. The highly efficient 1,3 cycloaddition grafting chemistry allows the
potential to incorporate a wide variety of functional groups including those with demonstrated thermo and
pH responsiveness.
38
2.5
Supporting Information
1
PPLG-97
----PPLG-77
0.8
Z
4d
- PPLG-64
-
0.6
L-
0.4
'A\
0.2
--
0
z
-
N
PPLG-44
0
12
16
14
Time (min)
Figure 2.SI: Normalized GPC traces of PPLG backbones
39
18
M
Figure 2.S2a: 1H-NMR spectrum PPLG backbone substituted with 75%
1
and 25% EQ 2 in [d7] DMF.
'H NMR-based Substitution
Feed Ratio Substitution
Azide
nEO 2
Azide
Azide 2
1
Azide 2
-
100%
diisopropylamine
-
diisopropylamine
50%
MEO 2
50%
diisopropylamine
50%
mEO 2
diisopropylamine
90%
mEO 2
11%
diisopropylamine
89%
MEO 2
diisopropylanine
50%
E02
53%
diisopropylamine
47%
E02
diisopropylamine
50%
10%
50%
-
-
100%
Figure 2.S2b: 'H-NMR spectrum PPLG backbone substituted with 50% mEO 2 and 50% diisoproylamine in D 2 0
and list of diisopropylamine functionalized PPLG-64 polymers synthesized for chapter.
40
. ..........
. ..
*
90
80
C
4-A
4-,
42
70
60
40
50
36A
E
40
Ln
C
30
I-
.
100
0
38
1334
A
-2 304P_
20
10
0
0
2
1
3
4
polymer concentration
50
40
30
20
(mg/mI)
70
60
Temperature (*C)
.
Figure 2.S3: Influence of temperature on the light transmittance (500 nm, heating 1 *C min-') of PPLG-64 fully
substituted with mE0 2 polymers in deionized water at concentrations of. 0.5 mg/ml, * 1 mg/ml, A 2 mg/ml, and
3 mg/ml. (inset: effect of concentration on cloud point)
100
U
E
U,
90
80
70
60
50
40
30
20
10
0
40
U
36
32
A
0
28
24
0
U
40
60
80
100
polymer DP
20
30
40
60
50
70
Temperature (*C)
100
40
C
4-J
E
C
U,
Co
LH
*
90
80
70
60
5
0
*
*
Figure 2.S4: Influence of temperature on the light transmittance (500 rn, heating 1 'C min-1) of PPLG polymers
having degree of polymerization of .44, A 64, .77, and 097 fully substituted with mEO 2 at 3 mg/ml in deionised
water.
5
50
40
30
20
10
0
*
0
1
p
2
3
4
Number of heating/cooling cycle
Figure 2.S5: Influence of cycling temperature across 4 heating and cooling cycles from 20 OC to 40 *C on the light
transmittance of PPLG-64 fully substituted with mEO2 (500 rnm, heating 10 'C min-' measurement after 3 min at
target temperature) at 3 mg/ml in deionized water.
41
100
90
80
70
60
N
a)
0
-A
-
50
.1
40
30
-A
C
20
1020-A
0
A
40
30
20
60
50
70
Temperature (*C)
Figure 2.S6a: Influence of temperature on the light transmittance (500 urn, heating 1 0 C min-) of representative
PPLG polymers at 3 mg/ml in deionised water. Figure 2.Shows the heating (filled shapes) and cooling (open shapes)
traces of four polymers having alkyne groups grafted with the specified percentage E0 2 and the remaining groups
with mEO 2 : o 0% E0 2 PPLG-64, A 39% E02 PPLG-64, m 100% E0 2 PPLG-97. The reported temperature is that of
the UV Vis heating block which, as measured by an external temperature probe, lags the temperature of the polymer
solution by 1.6 *C 0.2 *C, partially contributing to observed hysteresis. The dotted lines represent the linear fit of
the heating traces, as reported in Figure 2.S6b.
45
40
-
o
-
35
- C
30
25
)4--
0
A PPLG-64
0 PPLG-77
15
>.9
10
*PPLG-97
*
A
- 20
0
PPLG-44
.
A
5
0
20
40
60
%Substitution E0 2
.
Figure 2.S6b: Effect of the percent E0 2 groups grafted onto mEO 2 grafted PPLG backbones on the absolute value
of the slope of the linear region of the percent transmittance traces used to generate the cloud point values reported
in Figure 1 (500 nm, heating 1 0 C min-', 3 mg min~1 in deionized water). A representative linear region is traced with
a dotted line for the two thernoresponsive polymers in Figure 2.S6a. A greater slope indicates more discreet
thermosensitivity. All four PPLG backbones show more gradual temperature responsiveness with increasing
substitution of E0 2
42
.
41,
-
--------------
0.3
S0.25
0.2
0.15
/
0.1
/
/
9 0.05
60
40
Number of fhmbtonai groups atached
Nmr20
80
100
Figure 2.S6c: Theoretical model of the stochastic functionalization of a monodisperse PPLG backbone, degree of
polymerization = 100, with incomplete grafting of a single functional group, where 50 percent substitution shows
the greatest variance. In the experimental system of PPLG grafted with mEO 2 and E0 2 , this increased polydispersity
of functionalized backbones approaching equal substitution of the two grafted groups, results in more gradual
temperature responsiveness.
%EQ 2 by
NMR
0
Cloud
point in
water
(*C)
31
16
39
Cloud
PBS
(*C)
28
Difference
in cloud
point (*C)
3
39
34
5
51
42
9
point in
Table Sl: Comparison of cloud points measured in distilled water and PBS of PPLG-64 at 3 mg/mi substituted with
0
mEO 2 and the indicated percent E02 (heating 1 C min-1).
43
-
90
80 70
60
85
50
-
A
65
E 40
cL 45
-4
c 30
A
2 25
20
5
10
10
0
5.5pH6
6.5
ApH
20
30
40
50
70
60
Temperature (*C)
Figure 2.S7: Influence of pH on the temperature-dependent light transmittance (500 nm, heating 1 'C min-) of
PPLG-64 graft 100% diisopropylamine, 1 mg/mi in 100mM NaCl, 75mM phosphate buffer OpH 5.0, epH 5.6. mpH
6.0, and ApH6.2.
44
-
-
-
I
Y(-,
- -
-
-
.......
...
- .. -
-
a.
0.008
0.007
A
0.006
0A
0.004
50% amine 50% mE02
50% amine 50% E02
100% amine
-
* 100% mE02
-
0.005
.
0.002
-
0.003
0.001
0.000
3
7
5
9
pH
b.
0.7
10% amine 90% mE02
-
0.6 0.
0.4
o
0.3
0.2
-
E
-
0.5
-
..
0.1
0.0
3
I
I
5
7
9
pH
Figure 2.S8: Titration of grafted PPLG-64 with sodium hydroxide. Legend identifiers correspond to polymers
detailed in Table S1. Titration of diisopropylamine mEO2 grafted PPLG was performed on 3 ml solution in 125 mM
NaCl adjusted to pH 3. Figure 2.S8a shows solution of 2.5 mM amine titrated with 5 pL aliquots of 0.1 M NaOH.
Figure 2.S8b shows titration of PPLG graft 11% diisopropylamine performed at 0.25 mM amine with 5 pL aliquots
of 0.01 M NaOH.
45
.
.........
..
a.
80
0
0 25 *C
30 *C
I.
60 li
E
U
-0
E
-0
20k
A- 0*
20
9A
0
50
2
V
2f!
215S
-215
-40
x
N
-60
Wavelength (nm)
b.
Temp (*C)
0
30
35
40
45
-
-5
-10
-15
-20
E
vv-
A
-25
m
S
0
-30
x CL-35
-40
1 208 nm
222 nm
-45
Figure 2.S9a and Figure 2.S9b: Representative temperature dependent circular dichroism spectra of PPLG
substituted with only mEO 2 or E02. This spectra is of PPLG-97 graft mEO 2 in distilled water at 1 mg/ml. Figure
2.S9a shows representative screens of spectra recorded for a manual temperature screen from 25 'C to 50 *C,
allowing the solution to equilibrate at least 5 minutes between measurements. Figure 2.S9b shows the consistent
relative magnitude of alpha helical characteristic minima at 208 nm and 222 nm, the absolute magnitude of which
begins to decrease around 27 'C, the cloud point reported for this polymer in Figure 1. The decreased signal
intensity is attributed to decreased effective concentration caused by thermo-induced polymers precipitation.
46
a.
60.00
50.00
E
4
.
40.00 r
E
20.00
10.00
0
0.00
{.00
U
46
-20.00
0251.00
215.00
-10.0p9,
A A A
'-
168
C
74* C
30.00
U
25* C
600 C
AA
-30.00
-40.00
Wavelength (nm)
b.
Temp (*C)
0.00
25
-5.00
00
35.00
45.00
55.00
65.00
222 nm
Eo
C
-10.00
0 208 nm
-15.00
-
w
-20.00
-25.00
Xu
eA
-30.00
-35.00
Figure 2.SlOa and Figure 2.SM0b: Representative temperature dependent circular dichroism spectra of
PPLG substituted with 100% or 50% diisopropylamine. This spectra is of PPLG-64 graft 50%
diisopropylamine and 50% mEO 2 in 100 mM NaCl, 75mM phosphate buffer pH 5.6 at 1 mg/ml. Figure
2.S 1 Oa shows representative screens of spectra recorded for a manual temperature screen from 25 'C to
74 'C, allowing the solution to equilibrate at least 2 minutes between measurements. Figure 2.S 1 Ob shows
the decrease in magnitude of alpha helical characteristic minima at 222 nm relative to 208 nm, the
absolute magnitude of which begins to decrease around 60 'C, the cloud point reported for this polymer
in Figure 4. The decreased signal intensity and associated increase in intensity at 208 nm is attributed to a
transition of the polymer from purely a-helical to a mixture of a-helical and random coil secondary
structure.
47
2.6
References
Chen, Chongyi, Zhaohui Wang, and Zhibo Li. 2011. "Thermoresponsive Polypeptides from Pegylated
Poly-L-Glutamates." Biomacromolecules 12 (8): 2859-63. doi:10.1021/bm200849m.
Cheng, Yen-ju, Lung-chang Hung, Fong-yi Cao, Wei-shun Kao, Chih-yu Chang, and Chain-shu Hsu.
2011. "Alternating Copolymers Incorporating cyclopentadithiophene Unit and Organic Dyes for
Photovoltaic Applications." JournalofPolymer Science PartA: Polymer Chemistry 49 (8): 179 11801. doi:10.1002/pola.24604.
Dimitrov, Ivaylo, Barbara Trzebicka, Axel H. E. MUller, Andrzej Dworak, and Christo B. Tsvetanov.
2007. "Thermosensitive Water-Soluble Copolymers with Doubly Responsive Reversibly Interacting
Entities." Progressin Polymer Science 32 (11): 1275-1343.
doi: 10.101 6/j.progpolymsci.2007.07.00 1.
Ding, J., X. Zhuang, C. Xiao, and Y. Cheng. 2011. "Preparation of Photo-Cross-Linked pH-Responsive
Polypeptide Nanogels as Potential Carriers for Controlled Drug Delivery." JournalofMaterials ...
21: 11383-91. doi:10.1039/cljmlO391a.
Durand, Alain, and Dominique Hourdet. 1999. "Synthesis and Thermoassociative Properties in Aqueous
Solution of Graft Copolymers Containing poly(N-Isopropylacrylamide) Side Chains." Polymer 40
(17): 4941-51. doi:10.1016/S0032-3861(98)00698-3.
Engler, Amanda C., Daniel K. Bonner, Hilda G. Buss, Eva Y. Cheung, and Paula T. Hammond. 201 Ia.
"The Synthetic Tuning of Clickable pH Responsive Cationic Polypeptides and", 5627-37.
doi:10. 1039/clsm05O64h.
Engler, Amanda C., Hyung-il Lee, and Paula T. Hammond. 2009. "Highly Efficient 'Grafting onto' a
Polypeptide Backbone Using Click Chemistry." Angewandte Chemie (InternationalEd. in English)
48 (49): 9334-38. doi:10.1002/anie.200904070.
Engler, Amanda C., Anita Shukla, Sravanthi Puranam, Hilda G. Buss, Nina Jreige, and Paula T.
Hammond. 2011. "Effects of Side Group Functionality and Molecular Weight on the Activity of
Synthetic Antimicrobial Polypeptides." Biomacromolecules 12 (5): 1666-74.
doi:10. 1021/bm2000583.
Engler, Amanda C., Daniel K. Bonner, Hilda G. Buss, Eva Y. Cheung, and Paula T. Hammond. 2011 b.
"The Synthetic Tuning of Clickable pH Responsive Cationic Polypeptides and Block
Copolypeptides." Soft Matter 7 (12): 5627. doi:10.1039/c lsm05O64h.
Han, Seok, Mamoru Hagiwara, and Takashi Ishizone. 2003. "Synthesis of Thermally Sensitive WaterSoluble Polymethacrylates by Living Anionic Polymerizations of Oligo(ethylene Glycol) Methyl
Ether Methacrylates." Macromolecules 36 (22): 8312-19. doi: 10.102 1/ma0347971.
Jung, Seo-hyun, Hye-young Song, Youngil Lee, Han Mo Jeong, and Hyung-il Lee. 2011. "Novel
Thermoresponsive Polymers Tunable by pH." Macromolecules 44 (6): 1628-34.
doi:10.1021/ma102751p.
Laloyaux, Xavier, Emilie Fautr6, Thomas Blin, Viswas Purohit, Jerome Leprince, Thierry Jouenne, Alain
M Jonas, and Karine Glinel. 2010. "Temperature-Responsive Polymer Brushes Switching from
Bactericidal to Cell-Repellent." Advanced Materials22 (44): 5024-28.
doi: 10. 1002/adma.201002538.
Leung, Melissa K. M., Georgina K. Such, Angus P. R. Johnston, Dhee P. Biswas, Zhiyuan Zhu, Yan Yan,
Jean-Frangois Lutz, and Frank Caruso. 2011. "Assembly and Degradation of Low-Fouling ClickFunctionalized Poly(ethylene Glycol)-Based Multilayer Films and Capsules." Smalll 7 (8): 1075-85.
doi: 10. 1002/smll.201002258.
Lutz, Jean-Frangois. 2008. "Polymerization of Oligo(ethylene Glycol) (meth)acrylates: Toward New
Generations of Smart Biocompatible Materials." JournalofPolymer Science PartA: Polymer
Chemistry 46 (11): 3459-70. doi:10.1002/pola.22706.
Lutz, Jean-Frangois, and Ann Hoth. 2006. "Preparation of Ideal PEG Analogues with a Tunable
Thermosensitivity by Controlled Radical Copolymerization of 2-(2-Methoxyethoxy)ethyl
48
Methacrylate and Oligo(ethylene Glycol) Methacrylate." Macromolecules 39 (2): 893-96.
doi:10.1021/ma0517042.
Parrish, Bryan, Rebecca B. Breitenkamp, and Todd Emrick. 2005. "PEG- and Peptide-Grafted Aliphatic
Polyesters by Click Chemistry." Journalof the American Chemical Society 127 (20): 7404-10.
doi: 10. 1021/ja0503 1On.
Schilli, Christine M., Mingfu Zhang, Ezio Rizzardo, San H. Thang, Bill Y. K. Chong, Katarina Edwards,
Goran Karlsson, and Axel H. E. MUller. 2004. "A New Double-Responsive Block Copolymer
Synthesized via RAFT Polymerization: Poly( N -Isopropylacrylamide)- B Lock -Poly(acrylic
Acid)." Macromolecules 37 (21): 7861-66. doi:10.1021/ma035838w.
Schmaljohann, Dirk. 2006. "Thermo- and pH-Responsive Polymers in Drug Delivery." Advanced Drug
Delivery Reviews 58 (15): 1655-70. doi:10.1016/j.addr.2006.09.020.
Sinha, J., Sahoo, R., & Kumar, A. (2009). Processable, Regioregular, and "Click"able Monomer and
Polymers Based on 3,4-Propylenedioxythiophene with Tunable Solubility. Macromolecules, 42(6),
2015-2022. doi:10.1021/ma802289j
Tang, Haoyu, and Donghui Zhang. 2011. "Multi-Functionalization of Helical Block Copoly (a-Peptide) S
by Orthogonal Chemistry." Polymer Chemistry 2: 1542-51. doi:10.1039/cIpyOOOI5b.
Xiao, Chunsheng, Changwen Zhao, Pan He, Zhaohui Tang, Xuesi Chen, and Xiabin Jing. 2010. "Facile
Synthesis of Glycopolypeptides by Combination of Ring-Opening Polymerization of an AlkyneSubstituted N-Carboxyanhydride and Click 'Glycosylation'." MacromolecularRapid
Communications 31 (11): 991-97. doi: 10.1 002/marc.20090082 1.
Yamamoto, Shin-ichi, Joanna Pietrasik, and Krzysztof Matyjaszewski. 2008. "Temperature- and pHResponsive Dense Copolymer Brushes Prepared by ATRP." Macromolecules 41 (19): 7013-20.
doi:10.1021/ma8011366.
Yu, Miaoer, Andrew P. Nowak, Timothy J. Deming, and Darrin J. Pochan. 1999. "Methylated Mono- and
Diethyleneglycol Functionalized Polylysines: Nonionic, a -Helical, Water-Soluble Polypeptides."
Journal ofthe American Chemical Society 121 (51): 12210-11. doi:10.1021/ja993637v.
49
3
Synthesis of acrylate grafted, biofunctional PPLG hydrogels and their application in 2D cell
culture
This chapter contains two complementary sections which present, 1) the fabrication and preliminary
characterization of aqueous step-growth hydrogels crosslinked through PPLG grafted with acrylates and
2) the synthesis, characterization, and bioactivity of hydrogels functionalized with PPLG grafted with
bioactive peptides. Together, these sections introduce the underlying theory and practice enabling broad
future application of step-growth polypeptide hydrogels, especially when used as a synthetic extracellular
matrix. These preliminary studies form the foundation for the more comprehensive characterization
demonstrated in Chapter 4.
3.1
Aqueous crosslinking of PPLG through short acrylates
3.1.1
Introduction
This section presents a foundation for fabricating neutral polypeptide-cross linked hydrogels in
which PPLG polypeptides are grafted with acrylate sidechains and solubilizing 2-(2-azidoethoxy)ethanol
(E0 2). Limited precedence establishes the theory of hydrogels crosslinked with neutral polypeptides or
with crosslinkers being non-uniformily substituted with crosslinking functionality. The preliminary
screens presented in this section establish baseline characterization of these materials, which will be
extended in the following chapter.
Acrylate crosslinking
Various factors make acrylates an appropriate grafting group to initially explore polypeptide
hydrogel crosslinking. First, polypeptide crosslinking through acrylates closely mirrors established
strategies for crosslinking PEG-only hydrogels. Adoption of grafted polypeptide hydrogels by the tissue
engineering community is anticipated to be enhanced through straightforward integration of polypeptide
crosslinking with established crosslinking strategies. The acrylate group is one of the first and most often
used crosslinkers for PEG hydrogels, with the potential to crosslink through both step-growth (Phelps et
al. 2011; Zheng Shu et al. 2004; Hennink and van Nostrum 2002) and chain-growth (Nemir and West
2010; Hem and Hubbell 1998) polymerization, and as such, is an obvious first choice to crosslink
polypeptide hydrogels. Second, published synthesis of azido-propyl acrylate, a very similar molecule to
the proposed acrylate-PEG.=2 -azide. suggests a synthetic approach using inexpensive starting material,
enabling future large-scale synthesis. In comparison, commercially available PEG polymers (molecular
weight 3.4 and 5k) having acrylate and azide end groups, are both expensive ($300 per 100 mg) and
challenging to synthesize. A significant limitation of crosslinking through grafted acrylate for some
applications is the slow rate of gel formation and incomplete reaction through Michael-type addition with
thiols compared to other Michael-type conjugation strategies such as reaction with vinyl sulfones and
50
maleimides (Zheng Shu et al. 2004; Vanderhooft, Mann, and Prestwich 2007; Phelps et al. 2011).
Incomplete and inconsistent reactivity across a series of crosslinking polymers hinders robust
characterization of how the macromer structure impacts bulk gel properties. Extensions to crosslinking
through maleimides will be explored in Chapter 4 of this thesis.
E0 2 grafting
In this chapter, PPLG polypeptides are conferred aqueous solubility through grafting on of E02.
The previous chapter introduces this short grafting group as sufficient to confer aqueous solubility on
even long PPLG polypeptides. From studies of E0 2 chains grafted at high density as self-assembled
monolayers (SAMs), these short PEG units are expected to hydrate the polypeptide through hydrogen
bonding of water to the E02 side chains (Prime and Whitesides 1993). Additionally, this length PEG
chain grafted at high density is expected to limit nonspecific protein absorption (Banerjee, Pangule, and
Kane 2011).
3.1.2
Materials and Methods
L-(+)-Glutamic acid 99% minimum was purchased from EMD Chemicals (Gibbstown, NJ). 4-
arm PEG acrylate-IOk, 2-arm PEG thiol-3.4k,and 2-arm PEG thiol-IOk were purchased from Laysan Bio
(Arab, Alabama). 6-Carboxy-fluorescein-TEG azide was purchased from Berry & Associates (Dexter,
MI). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Culture Well Gaskets
(9mm diameter) were purchased from Electron Microscopy Sciences (PN 70465-8R).
Poly(y-propargyl-L-glutamate)(PPLG). PPLG was synthesized as previously reported (Chopko
et al. 2012; Engler, Lee, and Hammond 2009). The polymer had a degree of polymerization of 76 by
proton NMR ('H NMR), as determined by comparing heptylamine initiator protons to those on the PPLG
backbone, and a poly dispersity index (PDI) of 1.2 as determined by gel permeation chromatography
(GPC) against poly(methyl methacrylate) standards (Engler, Lee, and Hammond 2009).
)
Acrylate-PEGn=2-azide. A typical procedure includes purging 2-(2-azidoethoxy) ethanol (E0 2
(500 mg, 3.814 mmol) and triethylamine (0.798 mL, 5.725 mmol) with argon before anhydrously adding
dicholoromethane (6 ml) and chilling on ice for 15 minutes. Acryloylcholoride (401 pL, 4.96 mmol) was
added dropwise over 30 minutes with magnetic stirring. The reaction was removed from ice and reacted
for 8 hours. The product was gravity filtered to remove triethylamine hydrochloride, washed with 3x the
reaction volume 1 M HCl, water, 1 M NaOH, and water, dried with anhydrous magnesium sulfate and
then concentrated. Typical yield is 50%. Mequinol (12 mg) was then added to the product. The product
was a yellow liquid, which was stable for 4 hours under argon protected from light before
autopolymerizing into a lumpy solid. 'H NMR (CDCl 3, 6 ppm) 3.37 (2H, t), 3.66 (2H, in), 3.73 (2H, m),
51
4.31 (2H, m), 5.82 (1H, d),6.14 (LH, q), 6.39 (1H, d). The polymerized acrylate side chains were detected
by1 H NMR and had peaks from 0.5 to 2.5 ppm.
PPLG-g-acrylate.A typical procedure included grafting azide terminated side chains onto the
PPLG backbone in a two-stage reaction at an overall target ratio feed ratio of
alkyne/azide/CuBr/PMDETA equal to 1/1.2/0.1/0.1. To target 12% grafting of acrylate functionality and
1% fluorescein label, PPLG (0.05 g, 0.30 mmol alkyne repeat units), acrylate-PEGn= 2-azide (6.6 mg,
0.0360 mmol), carboxy-fluorescein-TEG azide (100 pL 0.03 M solution in DMF, 0.003 mmol) and
PMDETA (6.2pL, 0.03mmol) were all dissolved in dimethylformamide (DMF) (1.25 mL). The CuBr
catalyst (0.002 g, 0.015 mmol) was added to the degassed solution, and the reaction solution was stirred at
room temperature. After 1.5 hour, E0 2 (0.047 g, 0.359 mmol) was added quickly under a blanket of
argon. After another 1 hour, the reaction solution was precipitated in 40 mL cold diethylether, dissolved
in 10 mL distilled water, and incubated for 30 min with 5 mg Dowex@ M4195 sulfate copper chelating
resin. The beads were removed by filtration and the polymer solution was dialyzed for 12 hours against
water acidified by HCl (pH 5). and against distilled water for 6 hours. The polymer structure and degree of
substitution were verified by 1H NMR in [d7] DMF by comparing integration of the PPLG grafting peak
at 5.23 ppm (2H) to characteristic protons of the grafted acrylate at 5.94 (1H, d), 6.14 (1H, q), and 6.32
(1H, d) ppm.
Gelformation. Polymers were dissolved at 10 wt% in ultrapure water less than 1 hour before use.
Acrylate polymers, 4-arm PEG acrylate- 10k or PPLG-g-(acrylategfluoresceino.8 EO2 59), were mixed with
sufficient water to dilute the 3M triethanolamine (pH 7.65) to 300 mM in the final precursor solution. The
solution of acrylate, buffer and excess water was vortexed for 10 seconds before adding the thiol
crosslinker and again vortexing. The concentrations and nature of polymer were varied for individual gels
as reported in the results. The precursor solution (50 tL) was pipetted into 9 mm diameter Culture Well
Gaskets, and placed on a glass slide. The slide was transferred to a humidified chamber and allowed to gel
for 2 hours.
Fluorescence-basedquantification of incorporatedPPLG. Gels were transferred to sealed 1.5 mL
tubes containing 1 ml phosphate buffered saline (PBS) and were swelled at 37 'C for 12 hours. A control
polymer suspension was made by adding 50pL gel solution to the mixture, in which water was added in
place of the 2-arm PEG-SH crosslinker. The percent of incorporated PPLG in the swollen gels was
quantified by comparing the fluorescence of the supernatant against a dilution series of the uncrosslinked
gels, corrected for additional fluorescent-grafted PPLG incorporated in the precursor solution of gels with
higher percent PPLG.
52
3.1.3
Results and Discussion
3.1.3.1 Synthesis of PPLG-g-acrylates
Synthesis of acrylate-PEG= 2-azide followed the published synthesis of azidoethyl acrylate (N.
Narayana Reddy et al. 2010) with significant modifications to limit autopolymerized acrylate impurities.
Acrylates are known to autopolymerize through free radical polymerization. This side reaction is
especially undesirable when forming the crosslinking azide as it results in clusters of azides which, upon
grafting to the PPLG, crosslink the grafted polypeptides. During the work documented by this thesis,
undesired acrylate polymerization introduced significant challenges at each stage of the synthetic process
including during 1) acrylate azide synthesis, 2) the organic click reaction, 3) the purification of grafted
PPLG, and 4) the storage of the grafted PPLG.
Therefore, the time of reaction in the acrylate-PEGn= 2-azide synthesis was optimized to limit the
formation of the polymerized product while maximizing product yield. UV exposure was limited in all
synthesis steps, aqueous washes were chilled, and free radical inhibitors were incorporated. Optimized
techniques gave purified product, free of polymerized acrylate as observed by 'H NMR. However, even
when protected from UV light, with inhibitor, under vacuum, and chilled, this pure product only remained
stable for hours before autopolymerizing into a viscous gel, and as such, was grafted to PPLG
immediately after synthesis.
A series of PPLG backbones were grafted with both acrylates and E0 2 in a two-stage grafting
.
approach (Figure 3. 1B). PPLG was grafted with azides for one hour and then reacted with excess E0 2
Grafted PPLG was dialyzed at pH 5, kept chilled, protected from light, and stabilized with a radical
inhibitor to limit acrylate crosslinking. The extent of acrylate substitution in the lyophilized, filtered
product was quantified by 'H NMR, comparing the integrated acrylate peaks at 5.9-6.4 ppm to those of
the PPLG side chain. As shown in Figure 3.1 C, the percent of acrylate grafting, as observed by 'H NMR,
varied significantly from intended feed ratios, with all reactions having grafting ratios below the targeted
substitution. This inconsistency, which had not been observed in previous grafting reactions, is
attributable primarily to radical side reactions of the grafting acrylates, resulting in less soluble products
which precipitated during dialysis or were lost during filtration. The likelihood of side reactions increased
with increased acrylate grafting ratios, and even with optimized synthesis strategies, polymers having a
degree of polymerization of 76 and a feed ratio greater than 20% acrylates were unable to be purified as
non-crosslinked products. Additionally, discrepancies between feed ratios and grafted acrylates, as
characterized by 'H NMR, might also be attributed to imprecise quantification of the doublet and quartet
'H NMR peaks especially at low grafting ratios. These results emphasize the utility of grafting groups
53
having distinct 'H NMR signatures when using 'H NMR to quantify PPLG substitution and will be
discussed more in Section 3.2.
N3
O
H
B
R
CI6H
H2
"Click" R
-0"
R'-N 3
I 0
CHCH
8Rhrs
ITEA
R=C2513
N 3-PEG 3 - fluorescein
N 3-PEG2 - acrylate
0
NH2
C
76Exmebahs
Varied acrylate feed ratio in E02
00Feed
\
N.N
N
in DCM
I D1
O
N3,,^',O,-OH
R'
%
NMR
%
A
5
2.2
12
7
15
5
20
12
1.3
gelled
Figure 3.1: A. Acrylate-PEG= 2-azide is synthesized from acrylocholoride and 2-(2-azidoethoxy)ethanol,
with triethanolamine (TEA). B. Azide functional side chains are grafted to PPLG in a 2-phase organic
click reaction. C. Feed ratios of acrylate vary from grafting ratios observed by 'H NMR, suggesting
significant side reactions through grafted acryates.
To demonstrate the potential for polypeptide hydrogels to crosslink into step-growth hydrogels,
PPLG was synthesized having a DOP 76, with, on average, 9 acrylates grafted per backbone, 0.8
fluorescein grafted per backbone and solubilized by fully grafting with E02. This polymer will be
referenced with the naming convention, PPLG-g-(acrylatefluoresceinyEO2,), or PPLG-g(acrylate9 fluoresceino.8EO2 66 ). A series of step-growth acrylate-thiol crosslinked hydrogels were
fabricated with constant and equal molar feed ratios of acrylates and thiols, where acrylates, presented
from either 4-arm PEG acrylate- 10k or PPLG-g-(acrylategfluoresceino.8 E026 6 ), were crosslinked in 300
mM triethanolamine, pH 7.65, with 2-arm PEG thiol-3.4k. Triethanolamine is known to increase the rate
of the addition reaction by aiding proton exchange (Zhu 2010). The 4-arm PEG acrylate- 10k was selected
for having the same molecular weight per acrylate as the PPLG-g-(acrylategfluoresceino. 8 E0266 ). This
same mass per functional group allowed the ratio of acrylate crosslinker from PEG and PPLG to be varied
without also changing the overall polymer weight % in the gels, as bulk properties of PEG gels are known
to depend on the percent polymer in the precursor solution (Lutolf and Hubbell 2003). Additionally, 1%
fluorescein was grafted to the PPLG backbone during the organic click reaction, simplifying initial gel
characterization by labeling the PPLG with on average 0.8 fluorescein groups per macromer. This
grafting density was chosen to achieve maximum fluorescence while limiting multiple fluorescein groups
per PPLG, which would lead to fluorescein self-quenching.
54
.....
.......
3.1.3.2 Gelation of PPLG mixed with 4-arm PEG via dithiol PEG crosslinking
Gel precursor solutions (50 gL) having a total 12 wt% polymer with varied ratios of 4-arm
acrylate- 10k and PPLG-g-(acrylategfluoresceino 8 EO266 ) in 300 mM triethanolamine pH 7.65 were
vortexed, transferred to 9mm Teflon molds on glass slides, and allowed to gel for one hour in a
humidified chamber at room temperature. Gels were transferred to 1 ml PBS and swollen at 37 'C for 24
hours. Gelling time was not rigorously quantified but was around 35-50 minutes for all gels presented in
this chapter. Figure 3.2 shows gels after swelling.
Macromers (varied ratios)
0
PEG
a'
0'
PEG thiol
PPLG
% PEG 100 99
% PPLG 0
1
90
10
75
25
50
50
0
100
*5 wt% PEG thiol, 7 wt% (PEG+PPLG)
Figure 3.2: Varied ratios of 4-arm PEG- acrylate 10k (PEG) and PPLG-g-(acrylategfluoresceino 8EO266)
with fluorescein (PPLG), each having a molecular weight of 2.5k/acrylate, were crosslinked with 2-arm
PEG-thiol 3.4k (PEG thiol). The table shows representative data of the final mass of the dried polymer
gels made with the same wt % polymer and swelling ratios (mass swollen gel/mass dry gel) of these gels.
The gels that formed at all conditions had sufficient mechanical rigidity to allow easy
manipulation. However, those gels formed with 100% PPLG-g-(acrylategfluorescein. 8 EO2 66) seemed
qualitatively more brittle than PEG-only gels, resulting in the breaking off of gel fragments during
handling and the smaller diameter of the 0% PEG gel seen in Fig 3.2. Gel swelling, calculated as the mass
of the fully hydrated gel divided by the dried gel fraction, was consistent across the series of gels despite
the different structure of the crosslinking acrylates. Measurements of total incorporated polymer lacked
sensitivity and were not repeated due to material constraints. However, combined with the additional
evidence of the visible increase in orange fluorescence incorporated with increasing the fraction PPLG
and the constant faint orange of all PBS swelling solutions, these results suggest PPLG was wellincorporated into the gels.
55
3.1.3.3 Influence of acrylate:dithiol ratios on gelation properties of PPLG
A second series of gels (50 piL each) were fabricated with the same PPLG-g(acrylategfluoresceino.8E0266 ) and 2-arm PEG thiol-3.4k at a constant 12 wt%, with varying intended
precursor solution ratios of thiol to acrylate functional groups. For gels having ratios 0.8:1 to 1.2:1,
PPLG-g-(acrylategfluoresceino. 8 EO266 ) was varied from 6.6 to 7.7 wt% polymer while 2-arm PEG thiol3.4k was varied from 5.4 to 4.3% polymer (Figure 3.3). Fluorescence quantification of the swelling
supernatant suggests incomplete fluorescein-grafted PPLG gel incorporation at 1.1:1 and 1.2:1 intended
ratios of acrylate to thiol with nearly complete incorporation at conditions intended to have excess thiol.
Incomplete crosslinking is supported by the decreased mass of the lyophilized polymer at higher ratios of
acrylate and greater swelling of the 12 wt% polymer gels with higher ratios of PPLG-g(acrylategfluoresceino.8 EO266 ) (Figure 3.3C).
This screen can be considered within the context of two published ratio screens in Michael-type
step-growth hydrogel. Gels formed between thiolated hyaluronan and PEG diacrylate required two to
three fold excess thiol:acrylate to achieve near-complete acrylate consumption (Zheng Shu et al. 2004).
This compares to step-growth hydrogels made from dithiol peptides crosslinked with multiarm PEG vinyl
sulfone, a more reactive Michael-type receptor than an acrylate (Lutolf and Hubbell 2003). In a similar
ratio screen, the authors identify an optimal ratio near 1:1.1 vinyl sulfone to thiol as producing nearly
complete crosslinking and the stiffest gel with the least swelling (Lutolf and Hubbell 2003).
In addition to providing insight into the crosslinking of this particular system, these results
demonstrate a new technique of using pre-conjugated fluorescent dyes to quantify the extent of the
incorporation of individual species into crosslinked polypeptide hydrogels. Most frequently, percent
polymer hydrogel incorporation is calculated by mass loss, a technique that does not detect incomplete
incorporation of the different polymer species, or GPC of the supernatant, therefore requiring significant
protocol optimization (Zheng Shu et al. 2004). Fluorescence monitoring of polypeptides with low
fluorescent substitution is presented here as a useful tool for monitoring gel formation with limited
polymer precursor. Of note, the model dye presented here, fluorescein, has absorbance overlapping with
absorbance of the thiol reporting molecule, 5,5'-dithiobis-(2-nitrobenzoic acid) (Ellman's reagent),
thereby limiting concurrent quantification of unreacted thiols through reaction with Ellman's reagent.
Future experiments could readily enable both assays by grafting with fluorophores having nonoverlapping absorbances.
56
...
......
.. ........
..........
.
B-
_
Poly
PE G th iol
PPLG
.
Varied intended ratio
acrylate-thiol
0.8:1
0.91
1:2
1.1:1
1.2:1
A.
C-
0.25
o ass yield (dried gel S
0.2
A 0.15
(acrylatefthiol)
mass/original
polymer mass)
Oi
0.91
15
0.9:1
08
0-83
17
19
0.82
0.81
21
25
0.1
r
m 0.05
.0:1
.1:l
t
0
1.21
0.&1
U.
1.0:1
0.9:1
1.1:1
1.2:1
Ratio PPLG Acrylate: PEG Thiol
Swelling and final mass of polymers show incomplete
incorporation of polymer at ratios >1 and higher
swelling at greater ratio acrylate.
Quantified fluorescence leached into the supernatant
suggests incomplete PPLG incorporation at high ratio
PPLG-g-acrylate
Figure 3.3: A. Gels having varied ratios PPLG-g-(acrylategfluoresceino. 8EO266) and 2-arm PEG thiol-3.4 k
(PEG thiol), but constant 12 wt% of reagents, were crosslinked and swollen in PBS. B. Leached
fluorescence from fluorescein-grafted PPLG was calculated as a fraction of total fluorescence in polymer
precursor (n=3 gels). C. A screen (n=1) reporting final mass fraction of precursor polymer and swelling
ratios (mass swollen gel/mass dry gel) of the series of polypeptide crosslinked gels.
In a final experiment PPLG-g-(acrylategfluoresceino. 8E02 66) was again crosslinked with 2-arm PEG thiol3.4 k and also with 2-arm PEG thiol-10 k (Figure 3.4) to explore how thiol crosslinker length influences
bulk gel properties. In these gels, the wt% of thiol crosslinkers were 4.8 and 7.9 wt% of the precursor
solution for the 3.4k and 10k PEG dithiol crosslinkers respectively where acrylate wt % of PEG or PPLG
0
PEG
P% Prrey
PEG
0
0
0
100
OR
PPLG
+
OR
1
o
PPL
%PEG 100
%PPLG 0
ratio
1
21
1|
0.7
22
o0s
30 _ _
4.
0
M
Swelling
ma
%finalof poly
fiewd
o
120
0
100
Figure 3.4: A 4-arm PEG acrylate-10k (PEG) and PPLG-g-(acrylategfluoresceinO. 8EO266) with fluorescein
(PPLG), each having a molecular weight of 2.5k/acrylate, were crosslinked with 2-arm PEG thiol 3.4k
(3.4k PEG thiol) or 2-arm PEG thiol 10k (10k PEG thiol). The table shows representative data of the final
mass of the dried polymer gels made with the same wt% polymer and swelling ratios (mass swollen
gel/mass dry gel) of these gels.
57
polymers was 7.2 and 4.1 wt%. Swelling ratios of the gels depended greatly on the thiol crosslinker
(Figure 3.4), but not on the acrylate, supporting results in Figure 3.2, that swelling of this gel system
seemed independent of acrylate functional group.
3.1.4
Conclusions
Section 3.1 reports hydrogels that crosslinked through the Michael-type addition of PPLG grafted
acrylate with multiarm PEG thiol macromers, and in doing so, demonstrates the first reported hydrogels
crosslinked through grafted polypeptides. These studies validate that the E0 2 grafting group is sufficient
to solubilize PPLG for hydrogel applications. Additionally, systematic screens such as those presented in
Figure 3.2 using fluorescein as a model functional group, establish grafted PPLG as a highly controlled
handle able to crosslink functionality into a step-growth hydrogel across a wide range of concentrations.
More broadly, these screens show the potential to readily integrate grafted PPLG into an existing stepgrowth PEG hydrogels by grafting PPLG with appropriate complementary crosslinking chemistries.
The observed swelling dependence of grafted polypeptide hydrogels on gel stoichiometric ratios
(Figure 3.3) and on PEG crosslinker length (Figure 3.4) indicate significant similarities between
hydrogels crosslinked through PEG and those crosslinked through grafted PPLG. However, empirical
observations of the relative rigidity and elasticity of PEG and polypeptide hydrogels hint at fundamental
differences in the character of the two crosslinked gel systems.
While these finding suggests the utility of hydrogels crosslinked through PPLG grafted acrylates,
they also led to more systematic studies presented in Chapter 4, which explores gels crosslinked through
thiol addition to maleimides rather than to acrylates. Systematic explorations of gels crosslinked through
the highly efficient reaction of thiols with maleimides at constant polymer wt% and constant amount and
wt% of the crosslinking PEG thiol are used in these studies to more rigorously characterize relationships
between precursor macromer properties and final gel properties for gels crosslinked through grafted
polypeptides.
3.2
General strategies for organic grafting onto PPLG at low substitution
3.2.1
Introduction
Beyond introducing crosslinking groups such as grafted acrylate, a second challenge in adopting
polypeptide hydrogels for tissue engineering applications is incorporating biofunctional groups. This
section introduces organic grafting of azide functionalized peptides as a vehicle to investigate
systematically both the synthesis and characterization of grafted biofunctionality.
The short peptide, RGD, was selected as a model biofunctional group. This sequence of amino
acids is present in a variety of extracellular matrix proteins and can be bound by the acp3 integrin (Hynes
and Naba 2012), promoting cell adhesion and survival. As outlined in a recent review (Colombo and
58
Bianchi 2010), RGD has been conjugated extensively through click chemistries, but grafting to synthetic
grafted polypeptides has not been previously reported. More notably, strategies presented for grafting
RGD can be readily extended to grafting a variety of peptides as dictated by the end application, where
grafted peptide can either introduce biofunctionality or crosslinkers (Section 5.2.1.2).
Sections 3.2.1.1 and 3.2.1.2 explore the general theory of grafting PPLG at low levels of grafting
group substitution. Section 3.2.1.1 specifically introduces a model exploring how stochastic grafting to
PPLG is expected to result in well-defined polydisperse populations of grafted PPLG. Section 3.2.1.2
introduces strategies for choosing optimal tethers for grafting crosslinkers and biofunctional groups to
PPLG at low substitution.
Experimental results presented in Section 3.2.3 first introduce the solid phase peptide synthesis of
RGD-PEG,
12 -azide
and highlight challenges inherent in quantifying the substitution of peptides grafted
onto PPLG. Incorporating peptides engineered to have a distinct 'H NMR signature is recommended to
aid grafting quantification, but fails to provide precise quantification at low substitution. Similarly,
section 3.2.3.3 introduces a label-free approach to quantifying PPLG grafting but again fails to precisely
quantify PPLG grafting at less than a 5% substitution. Section 3.2.3.4 demonstrates the bioactivity of
PPLG grafted RGD peptide by demonstration that cell adhesion to hydrogels is dependent on the
concentration of grafted RGD.
Together, these experimental findings demonstrate successful PPLG organic grafting of a
biofunctional peptide while highlighting challenges in precisely quantifying peptide grafting, especially at
low substitutions. Results suggest the utility of considering complementary strategies for grafting
biofunctional groups as dictated by application. As will be introduced in the conclusion of this section, in
applications where a precise quantification of PPLG grafted peptides at low substitution is required, one
might consider the alternative approach of aqueous grafting onto PPLG.
3.2.1.1 Stochastic modeling of PPLG substitution of limiting functional groups
A significant feature of PPLG grafting is its near stochastic nature as supported by this thesis.
Appreciation of both the attributes and limitations inherent to this grafting strategy inform the utility of
grafted PPLG, particularly when considering the low target substitution of grafting groups as crosslinkers
and peptides. Perfectly ordered polymers, such as proteins, represent the extremes of both perfectly
engineered design and limited synthetic modularity and ease. Synthetic polypeptides inherently sacrifice
the precise control afforded in biological polymerization for higher throughput and more robust synthesis.
However, PPLG's robust grafting chemistries and defined secondary structure recover some control of
fully defined systems by allowing grafting well-modeled by a perfectly stochastic system. As such, in this
59
thesis we propose that PPLG substitution can be well modeled with the binomial probability density
function, for which, at a given value x and given pair of parameters n and p is:
y= f(x
In, p)= (n
,,...,(x)
Jpq(n-
where q = 1 - p. In this equation the result y, is the probability of observing x successes in n independent
trials, where the probability of success in any given trial is p, and where x is constrained to integer values.
This function is readily calculated through the Matlab function, binopdf(x,n,p).
The utility of this modeling approach is demonstrated by considering the case in which a target of
4 functional groups (e.g. crosslinkers of peptides) are intended to graft to each PPLG polypeptide, where
the polypeptide has a DOP of 8, 25, 50 or 200 (Figure 3.5).
0.3
c
0
(A
.0
DOP=8
0.25
- DOP=25
0.2
M DOP=200
.E0.15
0
0.1
U-
0.05
0
0
1
2
4
3
5
6
7
8
>8
Number substitution per macromer (target=4)
Figure 3.5: Distribution of grafting groups per polymer of polymers having a degree of polymerization
(DOP) of 8, 25, 50 or 200 with an average 4 grafting groups per backbone, as modeled with the binomial
probability density function.
As seen in Figure 3.5, for low intended grafting densities, increasing the degree of polymerization
decreases the fraction of macromers having the target grafting density, while increasing the fractional
population that has both no grafting groups and high numbers of grafting groups. The utility of this
modeling is demonstrated by considering the example of grafted crosslinkers. Targeted grafting ratios and
PPLG DOP might be optimized considering the constraints that PPLG backbones with less than two
crosslinking grafting group are unable to elastically contribute to network formation and not all
crosslinkers on macromers of high grafting densities are expected to be sterically available for
60
crosslinking. This simplified model assumes a mono-disperse starting population (e.g. PDI=1) but could
be extended to account for known polymer heterogeneity.
The utility of this stochastic modeling in characterizing crosslinked polypeptides is demonstrated
by applying this approach to analyzing published data presenting grafting of DNA to PPLG, the single
published example of incomplete PPLG grafting with an azide functionlized grafting group (Chen et al.
2012). As reported by the authors, DNA was clicked onto the PPLG backbone DOP 116, where one
azide-DNA was introduced per every 98 PPLG repeat alkynes, or on average 1.14 DNA per backbone.
Ungrafted PPLG was precipitated from an aqueous suspension and any ungrafted azide-DNA was
removed by dialysis. The concentration of purified fraction is not reported. The dialyzed grafted PPLG
was separated into discrete bands via polyacrylamide gel electrophoresis. Whereas the authors tentatively
attribute these discrete bands to a combination of incomplete click reaction between PPLG and
polydispersity of the grafting PPLG backbone, the above consideration of the stochastic nature of PPLG
grafting offers an alternative explanation. Modeling this grafting with the binomial probability density
function introduced above predicts a fractional distribution of reacted PPLG polymers as 0.25, 0.35, 0.23
and 0.1 having 0, 1, 2 and 3 grafted DNA chains respectively. This stochastic framework elegantly
explains the reported insoluble fraction of ungrafted PPLG and even the relative intensities of discrete
bands reported in supplemental Figure 2.S2, where bands of decreased mobility have increasing numbers
of grafted DNA. As an application for the grafted PPLG, complementary DNA was grafted to a second
batch of PPLG, and the two polymers were mixed, promoting the self-assembly of non-covalent PPLGDNA hydrogels. Systematic understanding of PPLG grafting and of crosslinking presented in this thesis is
expected to aid the design and the development of this and other crosslinking systems by better
characterization of the distribution of crosslinkers as determined by grafting feed ratios.
PPLG grafting group distribution is also relevant to the grafting of adhesive peptides for tissue
engineering applications, where clustering inherent in stochastic substitution could both limit and enhance
biological activity of the tethered ligands. Concentrated grafted peptides on a single backbone may limit
the accessibility of the grafted ligands. However, clustered RGD has also been shown to enhance cell
adhesion in both 2D (Maheshwari et al. 2000) and 3D (Lam and Segura 2013) cell culture systems, where
increased adhesion may be attributed to both locally engaging multiple integrins and the activity
enhancement by locally concentrating the adhesive ligands (Kiessling, Gestwicki, and Strong 2006).
Stochastic grafting on PPLG is expected to allow for future rational design of tissue engineering
platforms enhanced by engineered nanoscale organization and is explored in more detail in Section 5.2.4.
61
3.2.1.2 Geometry of grafted peptide
This section presents information pertaining to the synthesis of peptides, having azide linkers,
synthesized exclusively through solid phase peptide synthesis. It does not include extensive
considerations of peptides or proteins produced through bacterial expression systems and conjugated with
an azide, or even synthesized with engineered expression systems allowing incorporation of azidecontaining unnatural amino acids. Although both solid phase synthesis and bacterial expression may be
tractable, grafting of large proteins seems more appropriate through the second aqueous grafting, as
shown in Figure 1.2, rather than through organic grafting as considered here. Aqueous grafting limits
protein exposure to the potentially destabilizing pure organic solvent required to solubilize ungrafted
PPLG.
Focusing on peptide conjugated azides synthesized through solid phase synthesis, the structure
and length of the azide linker is dictated both by synthetic feasibility and application. Consider three
examples of linker geometries presented below in Figure 3.6. Figure 3.6A presents the option of
introducing azide functionality through peptides synthesized with the unnatural amino acid, azidolysine,
where the linker length could be further extended with a peptide spacer. Figure 3.6B shows an alternative
strategy where an azide-terminal PEG is conjugated to the deprotected epsilon amine of lysine during
solid phase peptide synthesis. A final approach also suggested in Figure 3.6B involves the grafting of a
longer, polydisperse, azide-terminated PEG side chain, where the PEG chain is conjugated to the peptide
during solid phase peptide synthesis or an off resin after peptide purification. Figure 3.6B shows PEG
conjugation through a lysine but PEG conjugates could also be synthesized by reacting them with the
terminal amine or through a cysteine reacting with azide-PEG-maleimide. Extensive experimentation
grafting polydisperse PEG chains, which is not included in this thesis, demonstrated numerous practical
challenges to sourcing pure di-functional PEG, maintaining solubility during the organic click reaction,
and characterizing conjugated products. As such, this section focuses exclusively on discrete
commercially available PEG linkers, which are available having up to 27 repeat units.
0
A
B
N3^
'
NH2
PEG-, linker ~3 nm fully extended
<
P-.6 nm
HN
OOO
HN
PEPTIDE
HN
0N
2
(Alternative azide linker might be PEG5k 1around 40 nm long)
Figure 3.6: Schematic of representative azide-grafted peptides, having carbon linker (A) or a peg linker
(B).
62
Optimal azide linker length is dictated by application. Considering the specific application of
tethering integrin-engaging ligands, grafting only through azidolysine would likely render the RGD
sterically inaccessible particularly with the additional grafting of the solubilizing EQ brush. The
literature offers some perspective in determining optimal peptide tether lengths for adhesive peptides, but
inconsistencies in published results across gel and cell platforms offer only loose guidelines (Kuhlman et
al. 2007; Wilson, Liliensiek, and Murphy 2012). In an effort to balance synthetic tractability with
maintaining bioactivity, this thesis introduces the model of grafting a peptide system of RGD grafted to
PPLG with a discrete PEGn=8 tether, as shown in Figure 3.6B. This length tether is thought to extend
beyond the PEG brush, but avoids the expense or challenging synthesis of polydisperse PEG linkers.
3.2.2
Materials and Methods
Azido-dPEG 4-NHS ester, Azido-dPEG 8 -NHS ester, and Azido-dPEG12-NHS ester were
purchased from Quanta Biodesign Ltd. (Plain City, Ohio). N 3 PEG11,
ooo
0 was synthesized following
previously reported synthesis (Engler, Lee, and Hammond 2009). Fmoc-Lys(Mtt)-OH and Fmocazidolysine-OH were purchased from Anaspec (Fremont, CA). Carboxy-fluorescein-TEG azide was
purchased from Berry & Associates (Dexter, MI). The control peptide, C-RGDS-NH 2 was synthesized
using standard solid phase peptide synthesis techniques (Sainlos and Imperiali 2007). O-(2-Azidoethyl)O-[2-(diglycolyl-amino)ethyl]heptaethylene glycol and O-(2-Aminoethyl)-O'-(2-azidoethyl)nonaethylene
glycol and other chemicals were purchased from Sigma.
Synthesis of N-linker-RGDS. N3 -PEG=4 12-RGDS: Resin-bound peptide,Fmoc-K(Mtt)-RGDS, was
synthesized with an amidated C terminus using standard manual Fmoc chemistry (Sainlos and Imperiali
2007) on Fmoc-PAL-PEG-PS resin (Applied Biosystems, Life Technologies). Synthesis was performed
in a 3m1 syringe with filter frit (New England Peptides).The methyltrityl(Mtt) protected amine side chain
of the terminal lysine group was selectively cleaved by 3, 10 minute washes with 2% (vol/vol)
trifluoroacetic acid (TFA) in dichloromethane with 1% thioanisole (TIS). Deprotected peptides were
washed with dichloromethane (3x) and a solution of 10% (vol/vol) diisopropylethylamine in DMF (1
minute).
The information below outlines three interchangeable synthetic strategies used to conjugate PEG
azide to the resin-bound peptide. Molarity of peptide is overestimated as 0.18 mmol/g resin from reported
resin loading, not accounting for the weight of the conjugated peptide or for imperfect compiling.
Strategy 1 was used for specific peptides reported in this thesis. The PEG linker length can be
systematically varied, with synthesis being limited by diffusion of conjugating PEG azide into the peptide
resin. In unpublished findings, by using a more porous resin such as TentaGel resin, the complications for
conjugating larger polymers were limited, but were not necessary for conjugating the short discrete PEG
63
linkers presented in this thesis. Three successful methods for conjugating the PEG linkers are given
below.
1.
Conjugation through N-hydroxysuccinimide-activatedcarboxylic acid PEG. In a representative
synthesis, resin (0.25g, 0.045 mmol) was washed with a mixed solvent of dimethylformamide and
dichloromethane, ratio 9:1 (3x, 2 ml). NHS-PEG 12-azide (67 mg, 0.09 mmol) was dissolved in the
same mixed solvent (1 mL) with N,N-diisopropylethylamine (39 pL, 0.225 mmol) and drawn into
the resin-containing syringe. The reaction was vigorously mixed overnight on a vortexer. Note: A
similar procedure gave complete grafting of NHS-PEG 4-azide and NHS-PEG 8 -azide. Typical
reactions showed complete conjugation of NHS-PEG 4-azide and NHS-PEG8 -azide and greater
than 80% conjugation of NHS-PEG12-azide as quantified by HPLC of intermediate products.
2.
Conjugationthrough carboxylic acidPEG. In a representative synthesis, resin (0.25g, 0.045
mmol) was washed with N-Methylpyrrolidone (3x, 2mL). O-(2-Azidoethyl)-O-[2-(diglycolylamino)ethyl]heptaethylene glycol (0.037 g, 0.068 mmol) was incubated with Nmethylpyrrolidone (2 ml), N,N-diisopropylethylamine (62 pL, 0.360 mmol), and (benzotriazol-1yl-oxytripyrrolidinophosphonium hexafluorophosphate) (PyBOP) (0.070 g, 0.135 mmol) for 5
minutes before being added to the resin. The reaction was mixed overnight on a vortexer.
3.
Amine to carboxylic acid then reaction with amine PEG. In a representative synthesis, resin (0.25
g, 0.045 mmol) was washed with DMF (3x, 2 mL). Diglycolic anhydride (0.052 g, 0.450 mmol)
and N,N-diisopropylethylamine (62 ptL, 0.360 mmol) were mixed with dimethylformamide
(2mL) and immediately added to the resin. After five minutes, the resin was washed with
dimethylformamide (3x, 2 mL). O-(2-Aminoethyl)-O'-(2-azidoethyl)nonaethylene glycol (0.036
g, 0.068 mmol) was incubated with N-Methylpyrrolidone (2 ml), N,N-diisopropylethylamine (62
pL, 0.360 mmol), and (benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate)
(PyBOP) (0.070 g, 0.135 mmol) for five minutes before being added to the resin. The reaction
was vigorously mixed overnight on a vortexer.
If desired for final applications, the terminal Fmoc protecting group could be removed with 4methylpiperidine (20% in DMF, 3x 10 minutes). Using standard procedures, PEG conjugated peptides
were washed, cleaved from resin, and purified on a reverse phase C18 column using 0.1% trifluoroacetic
acid with a gradient of acetonitrile and water (Sainlos and Imperiali 2007). Greater than 95% purity of the
peptides was confirmed by analytical HPLC and mass spectrometry.
Fmoc-azidolysine-GGG-RGDS: Resin-bound peptide, Fmoc-azidolysine-GGGRGDS-NH 2 was
synthesized with an amidated C-terminus using standard manual Fmoc-based chemistry (Sainlos and
Imperiali 2007) on Fmoc-PAL-PEG-PS (Applied Biosystems, Life Technologies). Synthesis was
performed in a 3 ml syringe with filter frit (New England Peptides).
64
Synthesis ofPPLG-g-fluorescein. PPLG-g-(fluoresceinyEO2z) was synthesized with minimal
modifications as above in this chapter's section, "Azide-functionalized PPLG." In a two-stage reaction,
PPLG having a degree of polymerization of 120 was grafted first with 6-carboxy-fluorescein-TEG azide
at a target substitution of 2 to 8 fluorescein per backbone and half the reaction was precipitated in
diethylether before being characterized by 'H NMR in d7-DMF. The remaining portion was grafted with
excess N 3-PEGmw,,ooo, dialyzed against 3 changes of water for 12 hours, lyophilized and characterized by
'H NMR in d-DMF.
Synthesis ofPPLG-g-(RGD,,acrylatefluoresceinEO2,).Four polymers, PPLG-g(RGDwacrylatexfluoresceinyEO2z), having varied substitution of RGD were synthesized with minimal
modifications as above in this chapter's section, "Azide-functionalized PPLG." In a series of four, twostage reactions, PPLG having a degree of polymerization of 70 was grafted with 6-Carboxy-fluoresceinTEG azide (0.01 equivalent to PPLG alkyne), Fmoc-K(PEGn=8-azide)-RGDSP-NH 2 (0, 0.01, 0.05 or 0.1
equivalent to PPLG alkyne), and acrylate-PEG= 2-azide (0.12 equivalent to PPLG alkyne) for 1.5 hours
.
before being grafted with excess E0 2
Cell attachment to PPLG-g-(RGD3 .sacrylate
8 .4fluoresceino.7EO25 4) in PEG hydrogels. Hydrogels
with 12 wt% polymer and varying concentrations of grafted-PPLG were synthesized from base gels of 4arm PEG acrylate-IOk and 2-arm PEG thiol-3.4k having equal molarity acrylate and thiol crosslinking
groups in precursor solutions. For a representative 120 ptL control gel, crosslinked without grafted PPLG,
4-arm PEG acrylate-IOk (8.4 mg) and 2-arm PEG thiol-3.4k (6.0 mg) were dissolved in 300 mM
triethanolamine, pH 7.65, to a total volume of 120 ptL. For gels containing RGD-grafted or control-PPLG,
pure buffer was substituted with grafted PPLG dissolved in buffer. In a representative gel targeting a bulk
concentration of 0.83 mM RGD-functional groups presented from PPLG-g)
(RGD 3.5acrylate8 4fluoresceino 7E02 54), solutions of PPLG-g-(RGD 3.5 acrylate8 4fluoresceino. 7EO2 54
(estimated MW 7,080 per RGD with grafted group ratios as established in Figure 3.9A) and PPLG-g(RGDoacrylate8 ,4fluoresceino 7EO2sg) (similar grafted PPLG with E0 2 substituted for grafted RGD) were
prepared (5 mg polymer per 0.6 ml buffer). PPLG containing precursor solutions were made as those of
control gels, with additional 8.8 pL PPLG solutions replacing 8.8 pL buffer in gel precursor. Positive
control gels were made with precursor solutions containing 1mM C-RGDS-NH 2. At these concentrations,
PPLG grafted acrylate-PEGn= 2-azide was less than 1% of that introduced by the 4-arm PEG acrylate and
was not considered significant in modifying bulk hydrogel properties.
The gel precursor solution (25 pL) was quickly added to wells of a 96 well plate. The plate was
centrifuged at 1000 rpm for 1 hour allowing gels to crosslink without the presence of a meniscus. Gels
were swollen overnight at 4C in 150 ptL PBS. Fluorescent plate reader measurements (X(ex) 485 nm,
X(em) 530 nm), calibrated against a standard curve, confirmed comparable conjugation of fluoresce in
65
grafted PPLG backbone in wells having PPLG grafted with RGD and control PPLG. One hour before cell
seeding, gels were incubated in 150 ptL cell culture media (DMEM supplemented with 10% Atlantic Bio
FBS, 1% Non-essential amino acids, 1% sodium pyruvate, 1% L-glutamine, and 1% PenStrep).
hTERT immortalized mesenchymal stem cells (hTMSCs) passage 15 were seeded in 100 IlL full
serum media at 3k cells/well and cultured for 24 hours. Cell attachment was observed using brightfield
microscopy (lOx).
3.2.3
Results and Discussion
3.2.3.1 Synthesizing azide-conjugated peptides for organic phase grafting-on
Organic phase peptide grafting onto PPLG alkynes requires synthesis of an azide-conjugated
peptide, being conjugated through either a polymer or peptide linker.
A.
HN
(o
N3
H
NH20%
HN
NH2
B.
H
N
N
N-+'N
10 mN
NH 2
4-methylpiperidine HN NHH
H
D
H
H0
OH
0
N
0OHN
HH
HN
NH 2
Figure 3.7: A. Chemical structure of RGDS having a PEG linker, synthesized through conjugation of Nhydroxysuccinimide-activated carboxylic acid PEG. Also, the Figure 2.Shows a schematic of 4methypiperidine catalyzed amine deprotection and B. Chemical structure of RGDS peptide having a
glycine spacer and a terminal azido lysine.
The structure of the RGD peptide conjugated through a PEG linker, as synthesized through
conjugation with N-hydroxysuccinimide-activated carboxylic acid PEG is as outlined above (Figure
3.7A). This peptide is shown before and after the removal of the Fmoc amine protecting group. Figure
3.7B shows a second linker strategy of having a three-glycine spacer between the RGD sequence and an
azido lysine. Future applications requiring longer peptide linkers may consider incorporating repeats of
the self-avoiding, hydrophilic and protease resistant peptide, (AlaGly) 3ProGluGly (McGrath and Fournier
1992; Wong Po Foo et al. 2009).
3.2.3.2 Organic phase peptide grafting and quantifying substitution by 'H NMR of peptide groups
RGD peptides with terminal Fmoc groups (Figure 3.7) were grafted through copper catalyzed
1,3-cyclo addition in DMF using reaction conditions previously established in the Hammond lab (Engler,
Lee, and Hammond 2009). Incorporating the Fmoc-amine protecting group in the grafted peptide has
utility for two reasons: 1) it allows 'H NMR quantification of the grafted peptide as its aromatic rings
have 'H NMR peaks distinct from the PPLG backbone and 2) it leverages well-established solid phase
66
peptide synthesis strategies to temporarily protect peptide amines from cross reacting during the organic
click reaction.
PPLG organic grafting of Fmoc-RGD-PEGn= 8-azide (Figure 3.7B) demonstrates both the
potential and limitations of how the Fmoc-based chemistry allows estimation of peptide PPLG grafting
density by 'H NMR. While none of the amino acids or PEG of the grafting peptide have 'H NMR peaks
distinct from the PPLG backbone, the ratio of the integral of the 'H NMR peaks from the four Fmoc
protons between 7.3-7.5 ppm to that of the grafted PPLG peak (2H, 5.3 ppm) can be used to approximate
peptide grafting. Proton NMR analysis suggests increasing peptide grafting with increasing peptide feed
ratios, although feed ratios and observed grafting are imperfectly correlated (where 1%, 5% and 10%
grafting peptides feed ratios had 'H NMR calibrated grafting efficiencies of 1.7%, 3.7%, and 7.4%
respectively).
This screen highlights systematic challenges identified in precisely quantifying Organic Phase
peptide grafting especially at low peptide substitution, for which a small error is a large percentage of the
total signal. Small inconsistencies between feed ratios and observed grafting could be reasonably
attributed to limited precision in both controlling feed ratios and in quantifying peptide grafting through
'H NMR. In more detail, feed ratios are calculated assuming an experimental mass and molar mass of the
grafting peptide, but the small amount of grafting peptide (<I mg for all three reactions) coupled with
variable non covalent association of peptide salts limit precision of both measurements. Routine 'H NMR
characterization is inherently imperfect but might be further optimized by increasing the number of scans
(here only 64), increasing relaxation time, increasing sample concentration and further optimizing solvent
conditions, although preliminary screens failed to identify conditions that would significantly improve
sensitivity. A more promising approach would be to explore alternative signature grafting with groups
having a distinct, singlet 'H NMR peaks rather than the multiple peaks of the Fmoc group.
Two final commentaries regarding peptide g rafting should be noted. First, because the Fmoc
protecting group may not be appropriate for certain applications, as a proof of concept, the group was
readily removed from the grafted peptide by incubating the grafted PPLG in 20% 4-methylpiperidine in
DMF for 10 minutes. Preliminary 'H NMR characterization of the purified product showed complete
amine deprotection and no obvious side reactions. Second, experience grafting other peptides not
included in this thesis highlights an alternative synthesis approach of grafting in DMSO, rather than
DMF, as DMSO was found to better solubilize some PPLG grafted peptides during the grafting reaction.
However, 1,3-cyclo addition in DMSO has slower grafting kinetics and requires heating (Xiao et al.
2010), even with optimized phase transfer agents (Chen et al. 2012).
67
3.2.3.3 Quantifying substitution by 1H NMR of partially grafted PPLG backbone
An alternative strategy was explored to develop a more broadly applicable approach appropriate
for quantifying grafting of a variety of peptides regardless of their peptide sequence or characteristic
proton spectra. Rather than monitoring the appearance of a new 'H peak from the grafting peptide,
grafting was quantified by characterizing the ratio of protons associated with a grafted and ungrafted
PPLG backbone. Accuracy of the feed ratios of the grafting species was improved over the peptide
system demonstrated in Section 3.2.3.2 by switching to the model compound carboxy-fluorescein-TEG
azide, which like peptides, is grafted through a PEG linker. The precise molar mass of this pure
commercial produce limited the variabilities suspected in peptide grafting feed ratios.
In Figure 3.8 A-C, representative 'H NMR of PPLG (Figure 3.8A) grafted with carboxyfluorescein-TEG-azide in a two stage reaction, where the partially grafted PPLG (Figure 3.8B) was then
grafted with excess PEG 2000azide and dialyzed against water (Figure 3.8C). As shown in Figure 3.8 D,
integration of the seven aromatic fluorescein protons around 6.75 ppm in d6-DMSO correlates with the
grafting feed ratio of 2 to 8 dyes per backbone. In contrast, label free grafting quantification dependent
only on PPLG backbone protons correlates well with grafting feed ratios only for reactions targeting
substitutions greater than 5%.
When combined, these results 1) confirm the general assumption of complete peptide grafting, 2)
highlight the utility of intermediate 'H NMR quantification as a label free-approach to quantifying PPLG
side chain grafting and 3) confirm challenges of absolute quantification of grafted peptides at less than
5% substitution.
68
A.
b~J
Add azide-Peg-5-Fam
with varied amounts of
10 azide-fuorescein
R
b
9
10
B.
8
5
6
7
2
3
4
1
ppm
Add exc ess of
azide -P EG
q(e
1
R10
C.
7f
9
6
7
5
4
3
2
1
ppm
3
2
1
PPm
II
I
I
n
~
q
'
Jk
Dialyze
f
9
10
8
7
6
5
4
HO
D.
20
18
x 16
14
* calculated from 1Hintegration
of b/(k+b)-as in 8
from iHintegration
1
caiculated
of (f/7/k -as in C
M
12
_10
t. 0
S6
Ca
4
z-0.S9x+ 1.21
S - 0.93
U2
LI
0
2
8
7
6
5
4
3
Intended Feed Substitution - number dye per 100 alkynes
9
10
Figure 3.8: A. 'H NMR of PPLG backbone in d-DMF, B. representative 'H NMR of PPLG partially
grafted with fluorescein, C. representative 'H-NMR of PPLG fully grafted with fluorescein and PEGiooo
and purified by dialysis and D. correlation of feed ratio and substitution for a series of reactions with
varied feed ratio and linear fit of calculated substitution from fluorescein proton integration.
69
)
3.2.3.4 Bioactivity of the PPLG-g-(RGD 3.sacrylate. 4fluorescein. 7EO257
The adhesion of hTERT immortalized mesenchymal stem cells (hTMSCs) to PEG hydrogels
crosslinked with PPLG was investigated as a metric to access the bioactivity and accessibility of the
PPLG-grafted RGD peptides. This section presents only preliminary studies broadly establishing PPLGg-acrylate with E0 2 as non-cell adhesive, while additional grafting of azide conjugated RGD peptides to
crosslinked PPLG enables cell attachment. Chapter 4 presents approaches allowing for more robust
characterization of cell response in a complimentary system that is able to more precisely quantify
grafted-peptides concentrations.
It is well-established that cells have limited adhesion to unfunctionlized PEG gels, but adhesion
can be improved through incorporation of short peptides such as RGD (Zhu 2010). The minimal
concentration of RGD needed to promote cell adhesion varies depending on the specific RGD sequence,
the adherent cell type, the character of the background gel, and the local presentation of the adhesive
ligand. Unpublished studies of a variety of bulk hydrogels tethered with RGDS peptides suggest minimal
hTMSC attachment beginning at 0.1 mM bulk incorporated RGD and robust adhesion to gels with at least
0.75 mM incorporated RGD.
The molecular weight per PPLG-grafted RGD was calculated from feed ratios of PPLG grafting
groups as outlined in Figure 3.9A. A similar calculation could be conducted based on substitution as
quantified by 'H NMR.
hTMSCs were seeded on PEG hydrogels having 12 wt% polymer from crosslinked 4-arm PEG
acrylate 10k and 2-arm PEG thiol 3.4k and varying ratios of PPLG with and without grafted RGD.
Twenty four hours after seeding, gels without RGD did not support hTMSC cell attachment or spreading.
Gels with 0.83 mM RGD grafted to PPLG or introduced through a terminal cysteine supported robust cell
adhesion and spreading as show in Figure 3.9B.
A.
Numwer
Mnomer
Ftuomsteh-PE(
Aayb*-PW
mWWnith
Goup
gup
feed
(gIM11
M2
facion
352
E02as
____
IM
00
0.22
.
owbntion
B.
of each of each
group
woup
DeWeeof
povmerzafiw perPPG (g/ffQ
5s
1
0.7
0
SA
Ms
O
57.4
7,3M
M-a 3.s 4,
IPPIG
W/jMWl
24,M
7,2M
Figure 3.9: A. Calculations of the average molar mass of PPLG-g-RGD and of the RGD grafting group of
a representative polymer grafted with fluorescein-PEG3 (Carboxy-fluorescein-TEG azide), acrylatePEG2 (acrylate-PEG= 2-azide), E0 2 and Fmoc-K(PEG= 1ON3)RGDS-NH 2 or PPLG-g(RGD 3.5 acrylate8. 4fluoresceino 7E02 57). B. Representative brightfield image of hTERT MSCs cultured for
24 hours on PEG hydrogels with PPLG-g-(RGD 3 5acrylate8 .4fluoresceino. 7E02 57) where the bulk
concentration of RGD is estimated as 0.83 mM.
70
Finally, published literature suggests that peptides tethered to PEG hydrogels require long PEG
tethers (>n=40) to robustly support cell adhesion (Wilson, Liliensiek, and Murphy 2012). Figure 3.9 and
related studies suggest RGD grafted through PEGn=8 linker is sufficiently accessible to promote ligandspecific cell adhesion. These results motivated additional cell studies presented in Chapter 4, investigating
hTMSC response to various gels grafted with RGD through only a PEG,=10 linker.
3.2.4
Conclusions
Section 3.2 broadly establishes strategies for grafting and characterizing azide-conjugated
peptides synthesized through solid phase peptide synthesis. It introduces several synthetic strategies to
incorporate azides through synthesis with azido lysine and a glycine spacer or through azide-terminated
discrete PEG linkers. These peptides are readily grafted on PPLG at quantitative yield as determined by
'H NMR. Attributes and limitations of characterizing the grafted product through 'H NMR is extensively
discussed, highlighting improved sensitivity of quantifying grafting of peptides containing aromatic
protons such as those in the Fmoc protecting group or fluorescein as compared to quantifying grafting of
peptide without aromatic labels. Finally, the biological activity of PPLG-grafted RGD was confirmed as
indicated by hTERT immortalized mesenchymal stem cells adhesion to 2D hydrogels crosslinked with
RGD grafted PPLG, demonstrating that the RGD PEGn=8 linker allowed sufficient extension of the
grafted peptide from the E0 2 brush for cell surface proteins to access the grafted RGD, even with the
additional steric constraints of crosslinking PEG chains also grafted to PPLG.
Beyond synthesis, a challenge inherent to grafting peptides onto PPLG is characterizing the
average substitution of the grafted group. Prior to this thesis, small molecule grafting to PPLG was
characterized exclusively through 'H NMR. However, the majority of amino acids have 'H NMR spectra
overlapping with the signal from the PPLG backbone, complicating characterization. This thesis presents
two broadly applicable strategies for adopting 'H NMR techniques to quantifying peptide grafting, either
through introducing groups having proton spectra distinct from those of the PPLG backbone (Section
3.2.3.2 and 3.2.3.3) or by quantifying the ratio of PPLG protons conjugated to ungrafted alkynes or
triazole groups (Section 3.2.3.3). Both techniques enable robust characterization for peptides grafted at
greater than 5% substitution as might be desired for introducing clustered adhesive ligands (Section 5.2.4)
or hydrogel crosslinking peptides (Section 5.2.1.2).
These studies also demonstrate the limitations of precise quantification of PPLG grafted peptides,
especially at low substitution. The utility of precise characterization of peptide grafting becomes apparent
when considering the case of PPLG having a degree of polymerization of 160 with an intended average of
4 adhesive peptides grafted per polypeptide backbone, or 2.5% substitution. Variability of only +/- 0.5%
71
observed substitution characterizes polymers as having 20% greater or less molar mass per grafting
group. Future studies may look to use PPLG grafted peptides to systematically explore how variables
such as linker length and nano-scale organization of single or multiple peptides affect cell response. It is
expected that the effects of these more subtle variables will be obscured by large variations in cell
response to variable adhesive peptide concentrations.
Applications requiring precisely calibrated peptide concentrations across multiples samples might
better employ aqueous grafting rather than organic grafting as presented in this chapter. Chapter 4
introduces crosslinking PPLG hydrogels through PPLG grafted maleimides, but this same grafted
macromer additionally allows an alternative approach to peptide conjugation through aqueous grafting as
presented in Figure 3.10.
1 minute
Average
1 per
PPLG
PBS
Complete SH
Consumption of by
Eliman's Assay
PPLG-g-maleimide
Peptide-SH
(Molar concentration SH
by Eliman's Assay)
1minute
Average
3 per
PPLG
PBS
Complete SH
Consumption of by
Eliman's Assay
PPLG-g-maleimide
Peptide-SH
(Molar concentration SH
by Ellman's Assay)
Figure 3.10: Aqueous grafting of thiol terminal peptide to PPLG-g-maleimide. Initial thiol concentration
and complete thiol grafting can be precisely quantified versus a standard curve using Ellman's Reagent.
The reaction scheme presented in Figure 3.10 takes advantage of the extremely efficient addition of
peptide conjugated thiols to maleimide grafted PPLG and easy commercial access to cysteine terminated
peptides. Ellman's Reagent can be used to precisely quantify the initial concentration of thiol-peptides
against a cysteine standard curve. Complete consumption of the peptides through aqueous grafting can be
verified by an Ellman's Reagent assay of the product. As such, aqueous grafting immediately before
hydrogel crosslinking should be considered as an alternative approach to organic grafting presented here
for applications requiring rigorous control of the concentration of grafted peptide in a sample.
72
(
3.3
References
Banerjee, Indrani, Ravindra C. Pangule, and Ravi S. Kane. 2011. "Antifouling Coatings: Recent
Developments in the Design of Surfaces That Prevent Fouling by Proteins, Bacteria, and Marine
Organisms." Advanced Materials, no. 23: 690-718. doi: 10.1002/adma.201001215.
Chen, Ping, Chuang Li, Dongsheng Liu, and Zhibo Li. 2012. "DNA-Grafted Polypeptide Molecular
Bottlebrush Prepared via Ring-Opening Polymerization and Click Chemistry." Macromolecules 45
(24): 9579-84. doi:10.1021/ma302233m.
Chopko, Caroline M., Erika L. Lowden, Amanda C. Engler, Linda G. Griffith, and Paula T. Hammond.
2012. "Dual Responsiveness of a Tunable Thermosensitive Polypeptide." A CS Macro Letters 1:
727-31. doi:10.1021/mz300088w.
Colombo, Matteo, and Aldo Bianchi. 2010. "Click Chemistry for the Synthesis of RGD-Containing
Integrin Ligands." Molecules (Basel, Switzerland) 15 (1): 178-97. doi: 10.3390/molecules15010178.
Engler, Amanda C., Hyung-il Lee, and Paula T. Hammond. 2009. "Highly Efficient 'Grafting onto' a
Polypeptide Backbone Using Click Chemistry." Angewandte Chemie (InternationalEd. in English)
48 (49): 9334-38. doi:10.1002/anie.200904070.
Hennink, W E, and C. F. van Nostrum. 2002. "Novel Crosslinking Methods to Design Hydrogels."
Advanced Drug Delivery Reviews 54 (1): 13-36. doi: 10.10 16/SO 1 69-409X(01)00240-X.
Hem, Diane L., and Jeffery Hubbell. 1998. "Incorporation of Adhesion Peptides into Nonadhesive
Hydrogels Useful for Tissue Resurfacing." JournalofBiomedical MaterialsResearch 39 (2): 26676. doi: 10.1002/(SICI)1097-4636(199802)39:2<266::AID-JBM14>3.0.CO;2-B.
Hynes, Richard 0., and Alexandra Naba. 2012. "Overview of the Matrisome--an Inventory of
Extracellular Matrix Constituents and Functions." Cold Spring HarborPerspectives in Biology 4
(1): a004903. doi: 10.11 01/cshperspect.a004903.
Kiessling, Laura L., Jason E. Gestwicki, and Laura E. Strong. 2006. "Synthetic Multivalent Ligands as
Probes of Signal Transduction." Angewandte Chemie (InternationalEd. in English) 45 (15): 234868. doi:10.1002/anie.200502794.
Kuhlman, William, Ikuo Taniguchi, Linda G. Griffith, and Anne M. Mayes. 2007. "Interplay between
PEO Tether Length and Ligand Spacing Governs Cell Spreading on RGD-Modified PMMA-G-PEO
Comb Copolymers." Biomacromolecules 8 (10): 3206-13. doi: 10.102 1/bm070237o.
Lam, Jonathan, and Tatiana Segura. 2013. "The Modulation of MSC Integrin Expression by RGD
Presentation." Biomaterials34 (16). Elsevier Ltd: 3938-47. doi: 10.101 6/j.biomaterials.2013.01.091.
Lutolf, Matthias P., and Jeffery A. Hubbell. 2003. "Synthesis and Physicochemical Characterization of
End-Linked Poly(ethylene Glycol)-Co-Peptide Hydrogels Formed by Michael-Type Addition."
Biomacromolecules 4 (3): 713-22. doi: 10.102 1/bm025744e.
Maheshwari, Gargi, Gillian Brown, Douglas Lauffenburger, Allen Wells, and Linda G. Griffith. 2000.
"Cell Adhesion and Motility Depend on Nanoscale RGD Clustering." Journalof Cell Science 113
Pt I (May): 1677-86. PMID:10769199.
McGrath, Kevin P., and Maurille J. Fournier, Thomas L. Mason David A. Tirrell. 1992. "Genetically
Directed Syntheses of New Polymeric Materials. Expression of Artificial Genes Encoding Proteins
with Repeating-(AlaGly) 3ProGluGly-Elements." Journalofthe American Chemical Society 114:
727-33. doi:10.1021/ja00028a048.
N. Narayana Reddy, Y. Murali Mohan, K. Varaprasad, S. Ravindra, K. Vimala, and K. Mohana Raju.
2010. "Surface Treatment of Plasticized Poly(vinyl Chloride) to Prevent Plasticizer Migration."
JournalofApplied Polymer Science 115: 1589-97. doi:10. 1002/app.
Nemir, Stephanie, and Jennifer L. West. 2010. "Synthetic Materials in the Study of Cell Response to
Substrate Rigidity." Annals ofBiomedical Engineering38 (1): 2-20. doi:10.1007/s10439-009-981 11.
Phelps, Edward A., Nduka 0. Enemchukwu, Vincent F. Fiore, Jay C. Sy, Niren Murthy, Todd Sulchek,
Thomas H. Barker, and Andres J. Garcia. 2011. "Maleimide Cross-Linked Bioactive PEG Hydrogel
Exhibits Improved Reaction Kinetics and Cross-Linking for Cell Encapsulation and in Situ
73
Delivery." Advanced Materials (DeerfieldBeach, Fla.) 24 (1): 64-70, 2.
doi: 10.1002/adma.201103574.
Prime, Kevin L., and George M. Whitesides. 1993. "Adsorption of Proteins onto Surfaces Containing
End-Attached Oligo(ethylene Oxide): A Model System Using Self-Assembled Monolayers."
Journalof the American Chemical Society 115 (23): 10714-21. doi:10.1021/ja00076a032.
Sainlos, Matthieu, and Barbara Imperiali. 2007. "Tools for Investigating Peptide-Protein Interactions:
Peptide Incorporation of Environment-Sensitive Fluorophores via on-Resin Derivatization." Nature
Protocols2 (12): 3201-9. doi:10.1038/nprot.2007.442.
Vanderhooft, Janssen L., Brenda K. Mann, and Glenn D. Prestwich. 2007. "Synthesis and
Characterization of Novel Thiol-Reactive Poly(ethylene Glycol) Cross-Linkers for ExtracellularMatrix-Mimetic Biomaterials." Biomacromolecules 8 (9): 2883-89. doi: 10.102 1/bm0703 564.
Wilson, Michell J., Sara J. Liliensiek, Christopher J. Murphy, William L. Murphy and Paul F.
Nealey. 2012. "Hydrogels with Well-Defined Peptide-Hydrogel Spacing and Concentration: Impact
on Epithelial Cell Behavior." Soft Matter 8: 390-98. doi:10.1039/c 1sm06589k.
Wong Po Foo, Cheryl T. S., Ji Seok Lee, Widya Mulyasasmita, Andreina Parisi-Amon, and Sarah C.
Heilshorn. 2009. "Two-Component Protein-Engineered Physical Hydrogels for Cell Encapsulation."
Proceedingsof the NationalAcademy of Sciences of the United States ofAmerica 106 (52): 2206772. doi:10.1073/pnas.0904851106.
Xiao, Chunsheng, Changwen Zhao, Pan He, Zhaohui Tang, Xuesi Chen, and Xiabin Jing. 2010. "Facile
Synthesis of Glycopolypeptides by Combination of Ring-Opening Polymerization of an AlkyneSubstituted N-Carboxyanhydride and Click 'Glycosylation'." MacromolecularRapid
Communications 31 (11): 991-97. doi: 10.1002/marc.20090082 1.
Zheng Shu, Xiao, Yanchun Liu, Fabio S. Palumbo, Yi Luo, and Glenn D. Prestwich. 2004. "In Situ
Crosslinkable Hyaluronan Hydrogels for Tissue Engineering." Biomaterials 25 (7-8): 1339-48.
doi: 10.1016/j.biomaterials.2003.08.014.
Zhu, Junmin. 2010. "Bioactive Modification of Poly(ethylene Glycol) Hydrogels for Tissue Engineering."
Biomaterials31 (17). Elsevier Ltd: 4639-56. doi: 10.101 6/j.biomaterials.2010.02.044.
74
4
Step-growth hydrogels crosslinked through uncharged helical polypeptides for applications
in tissue engineering
4.1
Introduction
Decades of research have established poly(ethylene glyol) (PEG) crosslinked hydrogels as a
useful synthetic extracellular matrix (ECM) for tissue engineering applications (Zhu 2010; Kharkar,
Kiick, and Kloxin 2013). The field has expanded from initial studies of gels made from unordered free
radical crosslinked PEG macromers, in which crosslinked PEG chains extend from radical-polymerized
hydrocarbon backbones (Nguyen and West 2002), to include gels made from step-growth crosslinking, in
which endgroups of multiarm PEG macromers having complementary functionality react to form a
crosslinked network (Elbert et al. 2001; Lutolf and Hubbell 2003). Despite proven utility and continual
advances in the field of step-growth PEG hydrogels, established gels only partly replicate the native
extracellular matrix (ECM) in terms of eliciting desired cellular responses. This work systematically
extends established step-growth PEG hydrogels into a new gel platform in which one component of stepgrowth PEG hydrogels is substituted by a-helical polypeptides grafted with multiple crosslinking
sidechains and a solubilizing short ethylene glycol brush. Polypeptide gels crosslinked with macromers
having defined secondary structure and readily functionalized side chains are expected to extend existing
PEG-only systems by 1) providing hundreds of handles on each crosslinker to systematically incorporate
and modulate desired biological and chemical functionality and 2) forming bulk gels possessing a more
diverse range of mechanical properties than standard PEG gels, moving them into a realm mimicking
those of the native ECM.
Despite the continued interest in both grafted synthetic polypeptides (Yan and Pochan 2010; He
et al. 2012; Rhodes and Deming 2013; Deng et al. 2014; Feng et al. 2011; Lu et al. 2014) and in stepgrowth PEG hydrogels, the literature offers limited precedence for step-growth hydrogels crosslinked
through polypeptides. Unlike the proposed, relatively inert, PEG-grafted peptide gels, established
polypeptide gels rely almost exclusively on crosslinking charged native polypeptides such as
poly(glutamic acid) (Markland et al. 1999; Zhang et al. 2011), poly(lysine) (Oliveira et al. 2003), and
poly(aspartic acid) (Gyenes et al. 2008), with only recent work introducing nanogels crosslinked through
photodimerization of polypeptide-side chain cinnamyloxy groups (Ding et al. 2011). Hydrogels formed
from un-natural polypeptides have consisted almost exclusively of non-covalently crosslinked gels (Yan
and Pochan 2010; P. Chen et al. 2012; Cheng et al. 2013), rather than covalent crosslinking established
here.
One specific N-carboxyanhydride polymerized polypeptide, poly(y-propargyl-L-glutamate)
(PPLG), was previously established in the Hammond group (Engler, Lee, and Hammond 2009) and forms
the foundation of this work. As outlined in a recent review (Quadir, Martin, and Hammond 2014), this
75
polypeptide's pendent alkynes allow grafting on of a wide variety of azide functionalized side chains via
copper catalyzed 1,3-cyclo addition. Examples of reported grafting groups include long PEG chains
(Engler, Lee, and Hammond 2009), short sugar molecules (Xiao et al. 2010), amines (Engler, Shukla, et
al. 2011), sulfonate ions (Tang and Zhang 2010), and thermoresponsive ethylene glycol grafting groups
(Chopko et al. 2012) . The robust a-helix before and after grafting of the polypeptide allows for almost
perfect grafting efficiency largely insensitive to the properties of the grafting azide (Engler, Lee, and
Hammond 2009). In 2012, the Hammond group introduced step-growth hydrogels covalently crosslinked
through neutral, water-soluble polypeptides, and demonstrated that the gel's bulk modulus could be
increased by incorporating a-helical polypeptides compared to gels from polypeptides having random coil
secondary structure (Oelker et al. 2012). Gel crosslinking was demonstrated through anhydrous activation
of ethylene oxide grafted PPLG by a non-specific coupling agent, and, without purification, crosslinking
through 4-arm PEG thiol- 10k (Oelker et al. 2012). This work extends the published approach by
introducing a modular, biocompatible, synthetic strategy allowing peptide and even cell encapsulation.
The model system presented in this work enables foundational studies exploring controlled stepgrowth polypeptide crosslinked hydrogels. Specifically, a 4-arm PEG thiol is crosslinked with PPLG
macromonomers pre-grafted with a short PEG brush and two orthogonal crosslinking chemistries:
maleimides and norbornenes. Maleimides react extremely efficiently in a pH dependent addition reaction
with thiolates (Phelps et al. 2011; A. D. Baldwin and Kiick 2011; Garcia 2014) while norbornenes react
with UV generated thiolenes (Fairbanks, Schwartz, Halevi, et al. 2009; Alge, Donohue, and Anseth
2013). This design, which allows double crosslinking, builds on recent advances in fabricating cell culture
systems engineered with temporal and spatial modulation of both mechanical properties (Khetan, Katz,
and Burdick 2009; Guvendiren and Burdick 2012; Gramlich, Kim, and Burdick 2013) and bio-active
grafted groups (DeForest, Polizzotti, and Anseth 2009; Gramlich, Kim, and Burdick 2013). Fabricating
hydrogels from polypeptide macromers having orthogonal groups leverages the robust, efficient, and
synthetically tractable PPLG grafting to enable synthesis of a wide range of well-characterized, modular,
and highly versatile hydrogel systems. The strategies presented here for grafting norbornene and
maleimide functionality onto PPLG can be readily extended to introduce a variety of crosslinking
chemistries (Patterson, Nazarova, and Prescher 2014; Vanderhooft, Mann, and Prestwich 2007),
crosslinker lengths (Lee et al. 2000), and solubilizing brushes, creating a highly versatile system for tissue
engineering applications.
4.2
Materials and Methods
L-(+)-Glutamic acid 99% minimum was purchased from EMD Chemicals (Gibbstown, NJ). 3Maleimidopropionic acid N-hydroxysuccinimide ester was purchased from Alfa Aesar (Ward Hill, MA).
76
2-(2-azidoethoxy)ethanol (E02) was synthesized as previously reported (Chopko et al. 2012). 8-arm PEG
maleimide-10k and 8 arm 40k PEG maleimide were purchased from JenKemTechnologyUSA (Plano,
TX). 4-arm PEG thiol-IOk thiol was purchased from Laysan Bio (Arab, Alabama). Stock solutions of
multiarm PEG were dissolved at 10 wt% in ultrapure water, pH 5, and stored at -80 'C until use. Gels
were UV crosslinked with a PK50 Omnicure series 2000 lamp with a 365 nm filter at reported intensities
as measured by Dymax Corp Accu-Cal-50 Smart UV intensity meter from Dymax Corperation
(Torrington, Connecticut) which measures UV-A (320-390nm) intensity. Adhesive peptides, GCRERGDSP-NH 2(RGD), and negative control peptide, GCRE-RGESP-NH 2(RGE), were synthesized by
Boston Open Labs (Fall River, MA). Human telomerase reverse transcriptase (hTERT) immortalized
human mesenchymal stem cells (hTMSC) were a gift from Dr. Junya Toguchida (Kyoto University,
Kyoto, Japan) and were used passage 10 (Okamoto et al. 2002). A p-Plate Angiogenesis 96 well was
purchased from Ibidi, LLC (Verona, WI). All other chemicals were purchased from Sigma-Aldrich (St.
Louis, MO) and used as received.
lIx stock solutions diluted to lx in the crosslinking gels were made at pH 7.4, pH 6.0, pH 5.3 and
pH 5.2. All stock buffer solutions contained 10x phosphate buffered saline (PBS). Additional 200 mM
MES was added to stock solutions below pH 7.4 and buffers were titrated with HCl or NaOH to the
designated pH.
Synthesis and Characterization.
Poly(y-propargyl-L-glutamate)(PPLG). PPLG was synthesized as previously reported (Engler, Lee, and
Hammond 2009; Chopko et al. 2012). The polymer had a degree of polymerization (DOP) of 160 by
proton NMR (H NMR), as determined by comparing heptylamine initiator proteins to those on the PPLG
backbone, and a poly dispersity index (PDI) of 1.18 as determined by gel permeation chromatography
(GPC) against poly(methyl methacrylate) standards (Engler, Lee, and Hammond 2009).
5-Norbornene-2-carboxylicacidN-hydroxysuccinimideester. During a typical procedure 5-norbomene-2carboxylic acid (I g, 7.23 mmol), N-hydroxysuccinimide (lg, 8.7 mmol), and dicyclohexylcarbodiimide
(1.79g, 8.7 mmol) were dissolved in 1 OmL anhydrous tetrahydrofuran (THF) and stirred under N 2 for 3
hrs. The reaction was concentrated and purified on a silica column (CH 2 Cl 2 ) to yield a white solid (95%
yield).
In-situ azide crosslinker conjugation. During a typical procedure O-(2-aminoethyl)-O'-(2azidoethyl)nonaethylene glycol or N3-PEGo-NH 2 (0.032 g, 0.061 mmol), N,N,N',N',N"pentamethyldiethylenetriamine (PMDETA) (12.7 pL, 0.061 mmol) and N-(3-maleimidopropionyloxy)
succinimide (0.018g, 0.067 mmol) were all dissolved in dry dimethylformamide (DMF) (1.95 ml, amine
77
at 0.03 mmol/ml). After 20 minutes, consumption of the amine was verified by thin layer
chromotagrophy. Specifically, lanes were spotted with a 0.75 pL reaction mixture or a standard curve of
N 3-PEGiO-NH 2 and PMDETA diluted in DMF from 1/8 to 1/128 of the reaction amine starting
concentration and run in a solvent system of methanol:dichloromethane at a volume ratio of 20:1. The
plate was stained with ninhydrin (0.04 g ninhydrin, 40 ml acetone, and 200 pL acetic acid) and resolved
with heat. The reaction was assumed complete when the reaction spot for O-(2-aminoethyl)-O'-(2azidoethyl)nonaethylene glycol, (Rf=0.05) was less intensely pink than the spot of the 1/128 dilution,
representing greater than 99.2% amine conversion. Synthesis of N 3-PEGIO-nobomene is as above,
substituting 5-norbomene-2-carboxylic acid N-hydroxysuccinimide ester for the activated maleimide.
This procedure was adopted from the protocols recommended by the vendor at
ClickChemistryTools.com.
Azide-functionalized PPLG. A typical procedure includes grafting onto the PPLG backbone azide
terminated side chains in a two-stage reaction at an overall target molar feed ratio of
alkyne/azide/CuBr/PMDETA equal to 1/1.2/0.1/0.1. To target 5% grafting of maleimide functionality,
PPLG (0.05 g, 0.30 mmol alkyne repeat units), crude N 3-PEGIO-maleimide (0.5 ml at 0.03 mmol/mL with
0.0063 pL/mL PMDETA, 0.015 mmol azide) and neat PMDETA (3.1 pL, 0.015 mmol) were all dissolved
in DMF (1.25 mL). The copper bromide catalyst (0.002 g, 0.015 mmol) was added to the degassed
solution, and the reaction solution was stirred at room temperature. After 1 hour, 2-(2azidoethoxy)ethanol (0.047 g, 0.359 mmol) was added quickly under a blanket of argon. After another 1
hour, the reaction solution was precipitated in 40 mL cold diethylether, dissolved in 10 mL distilled
water, and incubated for 30 min with 5 mg Dowex@ M4195 sulfate copper chelating resin. The beads
were removed by filtration and the polymer solution was dialyzed against water acidified by HCl (pH <4)
for 24 hours and against distilled water for 12 hours. The polymer structure and degree of substitution
were verified by 'H-NMR in [d6] DMSO. Typical yield was 75%. Stock solutions of grafted-PPLG
macromers were dissolved at 10 wt% in ultrapure water, pH 5, and stored at -80'C until use.
For PPLG macromers we use the following terminology to describe the various grafting substitutions:
(i) mono-functionality macromers are designated by PPLG-g-(maleimideEO2z) or PPLG-g(norbomeneEO2), where the number of maleimide or norbornene functional groups (x) and inert, water
solubilizing 2-(2-azidoethoxy)ethanol chains (z) grafted per PPLG molecule were systematically varied to
cover a range of functional groups/PPLG while ensuring PPLG chains were fully grafted such that x + z=
DP (ii) di-functional macromers grafted with both norbornene and maleimide functional groups, along
with inert solubilizing E02 chains, are designated by PPLG-g-(maleimidesnorbomeneyEO2z) such that x
78
+ y + z = DP. A complete list of the specific compositions for the panel of macromers used in this work is
provided in Table 4.1.
Gelformation. Four general types of crosslinked gels were produced: PPLG-g-(maleimideEO2z) reacting
with 4-arm PEG-thiol (10K) in a spontaneous reaction; [2] PPLG-g-(norbomene,,EO2,), reacting with 4arm PEG-thiol, catalyzed by 365 nm UV light in the presence of Irgacure 2959 photoinitiator; [3] PPLGg-(maleimidexnorbomeneyEO2,) reacting with 4-arm PEG-thiol (10K) in a sequence of spontaneous
reaction (maleimide with thiol) then exposure to UV light in the presence of Irgacure 2959. [4] PEG
macromer gels prepared by reacting 8-arm PEG-maleimide 10k with 4-arm PEG-thiol-10k in a
spontaneous reaction, with addition of low molecular weight single-functionality thiol (2mercaptoethanol) in defined proportions as indicated to drive non-stoichiometric crosslinking. All gels
created with PPLG macromers contained 4 wt% grafted PPLG polymer and 1.3 wt% 4-arm PEG thiol10k, with stoichiometries of reactive groups as described in Results. The detailed reaction conditions are
given below.
PPLG-g-(maleimideEO2) - PEG thiol gels. To form a PPLG maleimide grafted macromer gel
(50 uL), 4-arm PEG thiol-IOk (6.5 pL, 5wt% in water) was added to a well-mixed solution of grafted
PPLG stock solution (20 pL, 10 wt% in water), lOx buffer (5 ptL), and water (12 ptL). The malemide
content per PPLG was varied from 0.9 to 9.8 systematically. The buffer pH was optimized for each
crosslinking polymer, as reported in Table 4.1, to tune gelation time for all polymer systems to between 1
and 15 minutes. The precursor solution was vigorously vortexed for 10 seconds and immediately pipetted
between two hydrophobic glass slides which had been treated with Rain - X9 Original Glass Treatment
and rinsed with water. After 20 minutes, crosslinked gels were transferred to PBS.
PPLG-g-(norbomeneEO2) - PEG thiol gels Precursor solutions for gels crosslinked through
PPLG-g-norbomene in final buffering conditions of lx PBS pH 7.4, were similar to those for PPLG-gmaleimide gels, but included addition of 1% 2-Hydroxy-4'-(2-hydroxyethoxy)-2-methylpropiophenone
(Irgacure 2959) in 10% dimethyl sulfoxide in water at a final concentration of 0.05 wt% in gel.
Crosslinking was initiated by UV exposure at 365 nm of either 250 mwatts/cm 2 for 3.5 seconds or 10
mwatts/cm 2 for 5 minutes, as calibrated by a Dymax Corp Accu-Cal-50 Smart UV intensity meter.
PPLG-g-(maleimidesnorbomene-EO2,) - PEG thiol gels. Precursor solutions were made as
described above. For two-stage crosslinking, precursor solutions were allowed to gel between glass slides
for 45 minutes in full humidity at room temperature before being exposed to UV, 10 mwatts/cm 2 for 5
minutes.
PEG-maleimide - PEG-thiol gels. Precursor solutions for control PEG-only gels were crosslinked
at pH 5.2. Three gel systems having equal molarity thiol and maleimide, C.IOa (1.9 wt% IOk-8arm PEGmaleimide, 3.4 wt% 4-arm PEG thiol- 10k), C.40a (3.4 wt% 40k-8arm PEG-maleimide, 1.9 wt% 4-arm
79
PEG thiol-IOk) and C.40b (2.3 wt% 40k-8arm PEG-maleimide, 1.3 wt% 4-arm PEG thiol-IOk), were
selected to match the conditions of the PPLG-g-(maleimideEO2z)-PEG-thiol gels. C. 1 Oa and C.40a
matched PPLG-based gels in having 5.4 wt% polymer in the precursor solutions while C.40b matched
PPLG-based gels in having 1.3 wt% 4-arm PEG thiol- 10k. The effect of non-stoichiometric crosslinking
was explored by adding 2-mercaptoethanol to the precursor solution at molar equivalents equal to 0 to
60% of the maleimides. Ellman's Assay was used to calibrate free thiol concentrations in the stock 2mercaptoethanol solution against fresh cysteine standards. The mixtures were incubated for 30 minutes to
ensure complete conjugation as measured by Ellman's Assay.
Gel characterization.All gels (i.e., PPLG-based and control PEG-only gels) were swollen in excess PBS
overnight at 37'C.
Swelling and Polymer Incorporation. Gels made from 30 ptL precursor solution (n=3) were
equilibrated to room temperature and the swollen mass of each gel was measured. Gels were then placed
on tarred glass slides and dehydrated under vacuum for 24 hours. The mass swelling ratio is reported as
the ratio of swollen mass to dry mass, each mass less the assumed mass of salt from PBS (10.7 mg/ml in
liquid fraction).
AFM Mechanical Measurements. The elastic modulus of the hydrogels was measured using a
commercial scanning probe microscope (Molecular Force Probe 3D, Asylum Research, Santa Barbara,
CA). Force spectroscopy measurements were taken with two different types of functionalized silicon
cantilevers. The first was modified with a 4.5 ptm diameter polystyrene particle (Novascan Technologies,
Ames, IA) and the second was modified with a 5 gm diameter silicon oxide particle (Novascan
Technologies, Ames, IA). The nominal spring constants of the PS and SiO 2 cantilevers were 0.35 N m'
and 0.18 N m- respectively. All hydrogels (20-30 pL precursor solution) were imaged in PBS at 25 0 C.
Five force-displacement curve measurements at ten random and disperse locations were acquired for each
gel. The quantitative elastic modulus (E) was extracted from the force-displacement data using the Hertz
model for a spherical tip.
2
3(1 - v )F
4
VR3/2
Where F = applied force, R = radius of particle, 5 = indentation depth, and v = Poisson's ratio, which was
assumed to be 0.5 for all hydrogels. IGOR Pro data analysis software (WaveMetrics, Lake Oswego, OR)
was used to analyze all force curves.
80
Cell culture and characterization
hTMSCs were maintained in a standard medium formulation containing Dulbecco's modified Eagle's
medium, 10% fetal bovine serum (FBS), 1 mM pyruvate, 1 mM L-glutamine, 1p M nonessential amino
acids, and 100 units/ml penicillin-streptomycin (Life Technologies, Grand Island, NY), at 37'C in a
humidified atmosphere of 5% C0 2/ 95% air.
2D cell adhesion. PPLG-g-(malemidexEO2z) precursor solutions were mixed with PEG-thiol as outlined
above with addition of RGD or RGE peptides (250 pM final nominal concentration before swelling). Gel
precursor solution (10 gL) was added to wells of an Ibidi 96 well angiogenesis plate, (specialized wells
that limited meniscus formation), swelled overnight in PBS, then incubated for 2 hours in full serum
media before being seeded with hTMSCs in 70 pL media (1,750 cells per well). After 6 hours nonadherent cells were washed with PBS, before adherent cells were fixed and stained with DAPI, staining
cell nuclei, and with rhodamine-phalloidin, staining fibular actin (Life Technologies, Grand Island, NY).
Average adherent cell number and cell area were observed by fluorescence microscopy.
3D cell viability. PPLG-based solutions contained PPLG-g-(norbomenegEO21 5 2) (4 wt%), 4-arm PEG
thiol- 10k (1.3 wt%), 0.05% Irgacure 2959, RGD solution (250pM final concentration, pre-swelling) and
hTMSCs (15,000 cells) in PBS. Control PEG precursor solutions contained 4-arm PEG norbornene- 10k
(1.2wt%), 4-arm PEG thiol-IOk (1.3 wt%), 0.05% Irgacure 2959, and hTMSCs (15,000 cells) in PBS. For
each gel, hydrogel precursor solution (25 pL) was added to a 1 mL tuberculin syringe with the tip
removed. The cell suspensions were crosslinked with 365 nm UV light (10 mwatt/cm 2, 5minutes) and
ejected into 2 ml medium by advancing the syringe plunger. Each crosslinked gel was submerged in 1 mL
medium in a well of a 24 well plate. After 30 minutes and 1, 3, 5 and 7 days, viability was assessed using
the LIVE/DEAD assay (Life Technologies, Grand Island, NY). At designated times, one gel from each
condition was incubated with calcein AM (1:2000 dilution) and ethidium homodimer (1:800 dilution) in
phenol free, serum free DMEM media for 1.5 hours and washed. Reported viability is the total live cells
per total cells counted in three randomly chosen gel regions from reconstructed 5x fluorescence z-stacks,
and processed using ImageJ.
4.3
Results and Discussion
4.3.1
PPLG Macromer Synthesis
PPLG was synthesized as reported (Engler, Lee, and Hammond 2009) having a degree of
polymerization of 160 as calculated by 'H NMR and a poly dispersity index (PDI) of 1.18. To explore
how systematically varying polypeptide macromer grafting affects hydrogel bulk properties, a library of
fully substituted PPLG backbones was synthesized through two-stage grafting: first with maleimide,
norbornene, or both crosslinkers and then with an excess water-solubilizing short ethylene oxide brush
81
...............
(Table 4.1). The azide-grafted crosslinkers were synthesized immediately prior to PPLG-grafting by
mixing N 3-PEG,= 10-NH2 (1 equivalent) and NHS-activated maleimide or norbornene (1.2 equivalents) in
dry DMF with base. Coupling was confirmed by amine consumption as monitored by ninhydrin thin layer
chromotography. The crude product was immediately grafted onto the PPLG backbone through copper
catalyzed 1,3 cycloaddition. Grafted ratios of N 3-PEG= 10-maleimide and N3-PEG,= 10-norbomene,
calculated through 1H NMR closely match that of the feed ratios (Figure 4.1). This robust, near-stochastic
functionalization makes grafted-PPLG an especially attractive and versatile hydrogel macromer. Further,
this strategy of azide crosslinker synthesis eliminates challenges in maintaining crosslinker stability
during purification, limits consumption of expensive discreet PEG linker, and allows high modularity in
both the length of the PEG linker and the choice of the crosslinking group.
A
O
Nor
+ OR
10
>y
N
NaON
O
OR
~,o10
0
o
N
OR
H
Noy-Z
1
B
H2
A" AwA
R
0
R
H2
O
N3
0
o
O
OH
R=(CH 2)cH 3
N
R'
Figure 4.1: A. In situ synthesis of N 3-PEG= 10-maleimide and N 3-PEG 10-norbomene. Product was used
without purification. B. Grafting on of azide-functionalize crosslinker and E02. Schematic shows
polypeptide backbone as blue coil.
80 90 95 97.5 98.8...
>99.2%
Ninhydrin std curve amine conv. rxn
Figure 4.2: Representative ninhydrin-stained thin layer chromatography plate with N 3-PEGn= 10-amine
standard curve and rxn (reaction product) on right demonstrating greater than 99.2% amine consumption
through activated carboxylic acid conjugation.
82
y =0.93x - 0.11
* Maleimides
z7
*
8
0 Norbornenes
S6
w4.2x- 046
4
2
C
o13
0
0
1
3
2
4
5
7
6
8
9
% Grafting Crosslinker Feed
Figure 4.3: Correlation of norbornene and maleimide grafting feed ratio and PPLG-grafted crosslinker as
quantified by 'H NMR PPLG backbone and grafting group peak integrals.
0
0
R
o
R'=
0
0
OR
2Hp
RN= ''d
Np
OR
R'= '
-'O
Hn
Hn
H
II
2Hp
2Hn
---
--
--
-
ii
-
2Hm
-~
I
I
8
7
6
4
6
3
2
1
ppm
I q!
.
Figure 4.4: Representative 'H NMR of PPLG-g-(maleimide 2gnorbomene8. 4E02 149), a PPLG polymer
having a degree of polymerization 160 and fully grafted with N 3-PEGn= 10-maleimide (average 2.8 per
PPLG), N3 -PEGn= 10-norbornene (average 8.4 per PPLG) and E02
4.3.2
Properties of Hydrogels containing PPLG-g-(maleimideEO22)
Crosslinking aqueous hydrogels through neutral polypeptides was first demonstrated for a series
of PPLG-g-(maleimide,EO2,) macromers. Specifically, precursor solutions of grafted PPLG (4 wt%, 0.6
to 6.1 average maleimide substitution, 1.6-7.4 mM maleimide) and 4-arm PEG thiol-IOk (1.3 wt%, 4.4
mM thiol) were crosslinked in aqueous buffers having stoichiometric ratios of PPLG-grafted maleimide
83
to PEG-conjugated thiol ranging from 0.18 to 1.68 (Figure 4.5). Precursor solution buffers were adjusted
between pH 5.2 and pH 7.4 to allow sufficient time for complete mixing before gelation. More acidic
conditions decrease the concentration of the thiolate reactive species and the overall crosslinking rate.
Crosslinking solutions containing PPLG polymers with the most grafted maleimides, having an average
of 9.8 maleimide crosslinking groups per PPLG backbone, (i.e., PPLG-g-(maleimideg.8 EO2 15o)), became
solid 1 minute after mixing with thiol crosslinker, even at pH 5.2. Precursor solution with PPLG-g(maleimidel 9 E02 158 ), remained fluid for 10 minutes at pH 6.0, while precursor solutions containing
PPLG-g-(maleimideo 9 EO2 159), did not gel even 2 hours after mixing with thiol crosslinker, at a pH of 7.4.
All PPLG-containing hydrogels and solutions were optically clear.
PPLG-g-(maleimideo9 E02 159 )
0.18
-
Crosslinking
buffer pH
pH 7.4
PPLG-g-(maleimidel9 E02 158 )
0.36
-
pH 6.0
PPLG-g-(maleimide 2.2EO2 15 g)
0.40
-
pH 6.0
PPLG-g-(maleimide 3.BE02156 )
0.70
-
pH 6.0
PPLG-g-(maleimide 5 .6 EO2 15 4)
1.04
-
pH 5.3
PPLG-g-(maleimide.,3EO2 150 )
1.68
-
pH 5.2
PPLG-g-(norbornene3 .EO2 15 7)
0.56
-
pH 7.4
PPLG-g-(norbornene6 .OE02 15 4)
1.07
-
pH 7.4
r (lx)* r (2x)**
PPLG-g-(norbornene8 .OE02 15 2)
1.47
-
pH 7.4
PPLG-g-(norbornene 12 .2 EO2148 )
2.20
-
pH 7.4
PPLG-g-(maleimide 1.3norbornene 0 OE02 15 1)
0.24
2.00
pH 6.0
PPLG-g-(maleimide 2.8norbornene8 .4EO214 9)
0.53
2.60
pH 6.0
PPLG-g-(maieiide 4 1norbornene 4 6EO 2
0.77
2.40
pH 6.0
1.67
-
pH 7.4
15 1)
PPLG-g-(norbornene1 O.4EO3 10.4)
Table 4.1: Inventory of grafted PPLG. * r(I x) refers to the molar ratio of non-thiol to thiol of gels single
crosslinked through either maleimide or norbomene, ** r(2x) refers to the molar ratio of non-thiol to thiol
of gels double crosslinked through both maleimide and norbornene.
84
A
B
0
0
[
0+
0 0
01,
Figure 4.5: A. Schematic of PPLG (4 wt%, 1.6-7.4 mM maleimide, 0.6 to 6.1 maleimide per PPLG)
crosslinking with 4-arm PEG thiol-IOk (1.3 wt%, 4.4 mM thiol) having stoichiometric ratios of
maleimide to thiol ranging from 0.18 to 1.68. B. Schematic of 8-arm PEG maleimide- 10k or -40k
to thiol
crosslinking with 4-arm PEG thiol-IOk. Stoichiometric ratios were varied from 0.5 to 1 maleimide
PEG-thiol.
with
crosslinking
to
prior
by capping maleimides with 2-mercaptoethanol
Gel point of hydrogels crosslinked through PPLG-g-(maleimideEO2)
The observed gelation can be described by the predicted critical extent of reactions for gels
crosslinked through step-growth addition reactions. Polymerization progresses from a prepolymer having
4.3.3
of
conversion defined as 0 to complete crosslinking with conversion of 1. At a certain critical extent
the
reaction, the crosslinking solution sharply transitions to a gel. Flory-Stockmayer theory approximates
critical extent of reaction by modeling reactions between reacting groups A and B as exclusively
intermolecular addition reactions between independent, equally reactive crosslinking groups (Stockmayer
and B is
1943; Flory 1941). The critical extent of reaction at gel point (H,) between the reacting groups A
modeled as the extent of conversion of A, the limiting reactant, at which the weight-average molecular
at
weight becomes infinitely large. Depending on the functionalities, or the f, of the macromers, where
1 (for gels
least one macromer must have f>2, the value of H, ranges in general from a high approaching
requiring nearly complete reactions to form a gel) down to values of~0.2 for the high functionality
14
macromers considered in this thesis. The formal representation of the Flory-Stockmayer theory models
as a function of r, the molar ratio of the reacting groups (r=nA/nB), and
fw,A
and
fw,B,
the weight-average
functionalities of groups A and B as follows:
1
H = {r(fWA - 1))(fwB
-
1))11/2
PEG
Applying this theory to the series of PPLG-g-(maleimideEO2) macromers crosslinked with 4-arm
thiol- 10k defines the theoretical critical extent of reaction for each crosslinking precursor solution.
Considering precursor solutions where maleimides are stoichiometrically-limiting,
85
nA
and
nB refer
to the
moles of maleimide and thiol reactive groups respectively, thus conforming to the standard convention for
calculating r in the Flory-Stockmayer theory. PEG thiol macromer functionality,
fwB,
is well-estimated as
4. As a first approximation, PPLG-g-(maleimideEO2z) functionality can be modeled as a homogeneous
population in which each PPLG macromer is grafted with the average maleimide functionality as
calculated from 'H NMR. The nearly stochastic nature of discreet PPLG grafting suggests extending
established theory by modeling maleimide substitution with the binomial probability density function.
Both approaches are discussed below.
Figure 4.6A shows the critical extent of reaction for gels crosslinked with PPLG having different
average maleimide grafting, modeled at each resulting stoichiometric ratio for PPLG having average
monodisperse or bionomially distributed maleimide substitution. The limiting molar ratio of the reacting
groups assuming average maleimide grafting,
grafting,
r*a
r*a,
cooresponds with an limiting average maleimide
of 1.9 by Flory-Stockmayer theory. Thisf*a approaches the intuitive limiting average
crosslinking value of 2 for a monodisperse system, as PPLG macromers grafted with 0 or 1 maleimide
groups cannot form structural crosslinks. Supporting the predictedf*a, gel precursor solutions made with
PPLG-g-(maleimideo 9E02 159), having a stoichiometric ratio of 0.18, remained a liquid even after 2 hours.
Generally, Flory-Stockmayer theoretical calculations are considered a lower bound on critical
stoichiometric ratios, as Flory's principles neglect intramolecular reactions and variable crosslinking
reactivities caused by the substitution effects (Platte et al. 2011). However, precursor solutions of PPLGg-(maleimidel 9 E02158 ), having r at the r *a, gelled rapidly, suggesting limitations to modeling crosslinker
grafted PPLG as an averaged, monodisperse population.
86
r*b
1
A
r*a
0.8
0.36,j *a=1.9r*b= 0.2 4 , f *b=. 3
r* 0 =
0.7
0.6
0.5
C
0
0.4
0.3
CC
0.2
-
0
0.1
-
x
Wl
Average Maleimide Grafting
Binomial Maleimide Grafting
-
0
0
1
0.75
0.5
0.25
Maleimide: SH Ratio (r)
B
r=0.181
PPLG-g-(maleimide1j.EO215s) r=0.36
PPLG-g-(maleimide2 2 EO2 5a) r=0.40
PPLG-g-(maleimide 3.8 EO2156 ) r=0.70
PPLG-g-(maleimide 5.6 E021 54) r=1.04
PPLG-g-(maleimide 9.8 E02j50 ) r=1.68
*- - ------- - --- - --0%
20%
40%
60%
- - -
PPLG-g-(maleimideo 9 E02,s,)
80%
0
E l:
100%
Figure 4.6: A. Extent of reaction at gel point as modeled with the Flory-Stockmayer theory for gels with
PPLG-g-(maleimideEO2z) (4 wt%), crosslinked with 4-arm PEG thiol-IOk (1.3 wt%). Curves show
extent of reaction of limiting reagent at gel point assuming monodisperse PPLG-g-(maleimideEO2z),
each PPLG having average maleimide substitution (with critical stoichiometric ratio r*a and critical
average maleimide substitutionf*a), or binomially distributed PPLG-g-(maleimideEO22 ) for which the
extent of reaction at gel point is calculated by weight-average functionalities (with critical stoichiometric
ratio r*a and critical average maleimide substitutionf*a). B. Molar distribution of functional grafting
expected for each PPLG-g-(maleimide.EO2y), as modeled by the binomial probability density function.
Previous solution phase studies strongly support stochastic grafting onto PPLG (Chopko et al.
2012, also refer to discussion in Section 3.2.1.1) motivating a second approach modeling the critical
extent of reaction assuming that maleimide substitution follows a binomial distribution. At each average
substitution, grafted PPLG can be modeled as a heterogeneous population having integer values of
grafting groups from 0 to infinity. Figure 4.6B shows the molar distribution of functional grafting
expected for each PPLG-g-(maleimideEO2,), as modeled by the binomial probability density function.
Considering the specific example PPLG-g-(maleimidel. 9 E02158 ), PPLG grafted with 0, 1, 2, or >3
represent 15%, 28%, 27% and 30% percent of the total population, on a molar basis. Applying FloryStockmayer theory, accounting for the weight average functionalities of binomially distributed maleimide
87
grafting, suggests a limiting stoichiometric ratio, r*a, of 0.24. This stoichiometric ratio corresponds to only
an average 1.3 maleimides per grafted PPLG, again aligning with experimentally observed conditions.
Both average and binomial models additionally trends with the observed time to crosslink the
different precursor solutions, where gels having stoichiometric ratios less than 0.4 require near complete
reaction of residual maleimides at gel point while those having stoichiometric ratios greater than 0.9
require less than 30% consumption of the limiting crosslinker (Figure 4.6A).
4.3.4
Swelling of hydrogels crosslinked through PPLG-g-(maleimideEO2)
The effect of the precursor crosslinking groups' stoichiometric ratio on the swelling behavior of
PPLG-g-(maleimideEO22) hydrogels was investigated and compared to three control PEG hydrogel
systems.
As introduced above, precursor solutions of grafted PPLG (4 wt%, 1.6-7.4mM maleimide) and 4arm PEG thiol-10k (1.3 wt%, 4.4 mM thiol) were crosslinked in aqueous buffers. After swelling around
16 hours in PBS, the gels were lyophilized and the swelling ratio,
Q, was
calculated for gel system,
(Q=swollen gel weight per dry polymer weight). For reference, the theoretical
Q of the precursor solution,
or that of a gel that does not change its hydration with swelling, is 19. Gels having stoichiometric ratios
from 0.36 to 1.68 exhibited reasonably consistent swelling ratios ranging from 13 to 16, while gels made
from PPLG-g-(maleimidel 9E02158) at r=0.36, showed only a slightly higher swelling ratio of 22. As such,
gels containing PPLG grafted with an average of 1.9 or more maleimide groups consistently synerise to
between 13% and 33% of the precursor solution volume upon swelling in PBS, attributed to greater
zconfinement of crosslinked polymers compared to polymers in the solution phase (Figure 4.7).
60
+ PPLG-g-Mal
-o
A
AC.10a
A C.40a
A
A C.40b
50
E
5 40
A A
S30
20
10
0
0.00
1.00
0.50
Maleimide: SH Ratio
1.50
Figure 4.7: Swelling ratios of gels crosslinked through PPLG-g-(maleimide.EO2z) (blue diamonds) and
control PEG gels crosslinked through 8-arm PEG maleimide (triangles), preincubated with 2mercaptoethanol to give non stoichiometric crosslinking ratios. All maleimides crosslinked with 4-arm
PEG thiol-IOk.
88
Three series of control gels having precursor solutions of 4-arm PEG thiol- 10k and 8-arm PEG
maleimide-1 Ok or -40k were also crosslinked. Gel precursors having equimolar thiol and maleimide
functional groups were selected to match PPLG-g-(maleimideEO2z) gels in either total wt% polymer or
molarity of thiol. Specifically, gels C.IOa (1.9 wt% 8-arm PEG-mal-IOk and 3.4 wt% 4-arm PEG thiol10k) and C.40a (3.4 wt% 8-arm PEG-mal-40k and 1.9 wt% 4-arm PEG thiol-10k) both contained 5.3
wt% polymer. Gel C.40b (2.3 wt% 8-arm PEG mal-40k and 1.3 wt% 4-arm PEG thiol-10k) matched
grafted-PPLG gel crosslinker molarity but was only 3.7 wt% polymer. From each base polymer system, a
series of control gels having stoichiometric ratios less than 1 were generated by reacting 8-arm PEGmaleimide macromers with varied ratios of 2-mercaptoethanol prior to crosslinking, effectively removing
crosslinking maleimides while preserving constant precursor polymer wt%.
Unlike the relatively constant swelling observed for grafted-PPLG hydrogels having
stoichiometric ratios greater than 0.4, PEG hydrogel volume increased with swelling for all conditions,
with decreased crosslinking resulting in a further increase in swelling. Without maleimide capping, all
three control PEG gel systems, C.IOa, C.40a, and C.40b, showed limited swelling ratios of 20, 22 and 24,
respectively, corresponding to 3%, 4% and 33% volume-increases experienced during swelling. However,
as predicted by elastic polymer crosslinking theory and demonstrated in published PEG systems (Lutolf
and Hubbell 2003), control PEG gels swelled 2 to 3 times their original volume with stoichiometric ratios
decreased to 0.6 by maleimide consumption prior to gel crosslinking (Figure 4.7).
Increased swelling ratios observed for gels crosslinked only with PEG compared to those of gels
crosslinked through grafted PPLG is attributed to grafted PPLG's increased mechanical resistance to
deformation as compared to that of PEG macromers. In the classic hydrogel swelling model developed by
Flory and Rehner (Flory and Rehner 1943a; Flory and Rehner 1943b), and extended by Bray, Peppas and
Merrill (Bray and Merrill 1973; Peppas and Merrill 1976), equilibrium polymer swelling of solutioncrosslinked gels can be systematically modeled as the balance between 1) enthalpic and entropic forces
driving the distribution of polymer chains into the solvent and 2) the crosslinked polymers' mechanical
resistance to extension. In more traditional terms, the free energy of the gel is assumed to be the sum of
the free energy associated with mixing the solvent and the polymer and the free energy associated with
stretching the polymer network.
It is proposed that for the specific PPLG-crosslinked gels examined here, grafted PPLG
polypeptides maintain sufficiently stabilized a-helical secondary structure to resist the nanoscale
extension required for macroscopic swelling. From another prospective, entropic effects are quite
different for the PPLG gels than for gels crosslinking with PEG random coils. More extended chain
confirmations lead to less entropic driving force for further expansion. The rod-like behavior of grafted ahelical polypeptides is especially significant in this particular gel system as PPLG-g-(maleimidexEO2,)
89
accounts for 75% of the total polymer and 21% of the total crosslinking volume of an unswollen grafted
PPLG gel (modeled in detail in Section 1.3.2). PPLG gels that are crosslinked through longer PEG
crosslinkers, have less stabilized secondary structure, or have lower DOP are expected to demonstrate
swelling more similar to that observed in gels crosslinked only through PEG. [Section 1.2.1 more fully
explores justifies PPLG's assumed a-helical structure in a crosslinked gel and Section 1.2.2 quantitatively
explores differences in the single molecule response of PEG and a-helical polypeptides to force].
4.3.5
Mechanical properties of hydrogels crosslinked through PPLG-g-(maleimideEO2z)
Elastic moduli measurements characterizing the series of PPLG-g-(maleimidexEO2z) gels show
gel stiffnesses increasing from 3 kPa to 17 kPa for gels crosslinked with stoichiometric ratios of 0.36 to
1.68, corresponding to 1.9 to 9.8 grafted maleimides per PPLG (Figure 4.8). PEG-only control gels had
moduli ranging from 9 kPa to 16 kPa.
20
* PPLG-g-MaI
18
A C. 10a
a- 16
U,
A C.40a
A C.40b
14
.5
-o 12
0
10
8
w 6
U
U,
U-
4
2
n
0.00
0.50
1.00
1.50
Maleimide: SH Ratio
Figure 4.8: Elastic moduli quantified by AFM indentation of gels crosslinked through PPLG-g(maleimideEO2,) (blue diamonds) and control PEG gels crosslinked through 8-arm PEG maleimide
(triangles). All maleimides are crosslinked with 4-arm PEG thiol- 10k.
Gels crosslinked with 1.3 through PPLG-g-(maleimide,EO2z), having x=1.9, 2.2, and 3.8, all had
insufficient maleimides to fully react with the 4-arm PEG-thiol-1 Ok, defined as having a ratio
maleimide:thiol less than 1. These gels demonstrate expected increasing stiffness from 2.6 kPa to 7.9 kPa
with increasing maleimide substitution (Figure 4.8). However, gels crosslinked with PPLG-g(maleimide5 .6 EO2 154) (r=- 1.01) and PPLG-g-(maleimide9 8. .EO2 15 o) (r= 1.68) both have bulk maleimide
concentrations sufficient to allow complete PEG thiol consumption and maximum gel crosslinking,
suggesting that these gels should have similar bulk properties. As such, the dramatic increase in stiffness
from 8.7 kPa to 17.3 kPa for gels crosslinked with PPLG-g-(maleimide. 8.EO21 50) compared to gels
crosslinked with PPLG-g-(maleimides. 6 E02 154) deserves particular attention.
90
Generally step-growth gels with higher molarity of crosslinker and more ordered gel structure
have stronger mechanical properties (Sakai et al. 2008). This section explores one explanation of why
PPLG grafted with an average 5.6 and 9.8 maleimides might crosslink to form gels with different levels
of crosslinking and overall structure. As an extension of Figure 4.6B, substitution of both PPLGs can be
modeled with the binomial probability density function (Figure 4.9). Gels crosslinked with PPLG-g(maleimide5 .6 EO2 154) have sufficient maleimides to fully react with thiols. However, for these gels, only
49% of the PPLG backbones, or 66% of PPLG grafted-maleimide groups, are modeled as being tethered
with 6 or more additional maleimides on a single tethered backbone, where 6 is the average integer
number of crosslinks per PPLG required for complete thiol consumption. Mismatch between such
heterogeneous distribution of maleimide crosslinkers and homogeneous distribution of thiol crosslinkers
is expected to both limit overall crosslinking and drive a crosslinked gel with pronounced heterogeneties
mirroring heterogeneities in the crosslinking PPLG. By stochastic maleimide grafting, PPLG-g(maleimideg.8 E02 150 ) is modeled as having 93% of PPLG polymers with at least the 6 maleimides per
backbone required for complete crosslinking. Therefore, 4-arm PEG-thiols-10k crosslinkers are expected
to completely crosslink with PPLG grafted maleimides into a more heterogeneous, stiffer, gels in which
PEG thiols are consumed locally and are not required to "reach" to react with clusters of distant PPLG
grafted maleimides.
The spatial model of the crosslinking polymers developed in Section 1.3.2 and summarized in
Figure 1.6 is a useful tool for gaining a more intuitive understanding of the implications of
heterogenerous PPLG grafting and its effect on gel structure.
0.2
0.18
4
E_
0.14
Z 0.12
0.1
r0
-
0.1
0.08
0.06
UPPLG-g-mal(9.8
0
0---
ON
-
----
-
0.04
per)
0
--------
0
0.02
-
-
0.16
+
-
w
o PPLG-g-mal(5.6 per)
0
N
0
n
0
10
5
15
20
# maleimides per PPLG
Figure 4.9: Distribution of grafting groups per polymer of polymers PPLG-g-(maleimide AE02 154 ) and
PPLG-g-(maleimide9 .8 E02 150 ), as modeled with the binomial probability density function.
91
Further exploration of the impact of PPLG macromer substitution on gel stiffness would benefit
from testing additional polymers, as well as quantifying free thiols remaining in the PPLG crosslinked
gels. Absolute quantification, especially for gels having a stoichiometric ratio greater than 1, is
complicated by addition of the Ellman's Reagent colorimetric product, 2-nitro-5-thiobenzoic acid, to
excess maleimides. However at first approximation, solutions of Ellman's Reagent incubated with the two
gel systems show slight signal above background only for gels crosslinked through PPLG-g(maleimide5 6. EO21 54). Future studies might consider more quantitative detection through NMR (A.
Baldwin and Kiick 2013), or alternative reagent-based thiol detections (X. Chen et al. 2010).
4.3.6
Introducing hydrogels crosslinked through PPLG-g-(maleimidenorborneneyEO2,)
Highly controlled PPLG-grafting with a single functional group and E0 2 brush can be extended to
consider the grafting on of two grafted groups having orthogonal functionalities. This work demonstrates
additional PPLG grafting with the norbornene functional group resulting in increased control of both bulk
gel properties and grafted bio-functionality through two orthogonal chemistries. UV-mediated addition of
thiol radicals to norbornenes is a well-established technique for both crosslinking cells into PEG
hydrogels (Fairbanks, Schwartz, Halevi, et al. 2009; Alge and Anseth 2013) and grafting biofunctionality
to already crosslinked gels with temporal and spatial control (Gramlich, Kim, and Burdick 2013). In a
second reaction, norbornenes have been shown to react with electron-deficient tetrazines through cellcompatible inverse-electron-demand Diels-Alder cycloaddition. Within the field of biomaterials,
conjugation between tetrazine and a norbornene derivative has been demonstrated to crosslink bulk PEG
hydrogels (Cok, Zhou, and Johnson 2013; Alge et al. 2013) and enable specific bioconjugation (Han et al.
2010). Thiolene and tetrazine conjugations are generally orthogonal to each other and to maleimide
thiolate addition, making PPLG grafted with both maleimides and norbornenes a useful system to explore
multiplexed PPLG grafting for tissue culture application. The following section first explores the
crosslinking PPLG-g-(norborneneyEO2z) as demonstrated by the bulk properties of polypeptide
crosslinked hydrogels before extending these studies to gel systems containing PPLG grafted with both
maleimides and norbornenes.
4.3.6
Swelling of hydrogels crosslinked through PPLG-g-(norborneneyEO2,)
A series of PPLG-g-(norborneneyEO2z) backbones were synthesized by a two-stage click reaction
having an average grafting of 3.0 to 12.2 norbornes per backbone and fully grafted with E0 2 . Precursor
solutions containing PPLG-g-(norborneneyEO2z) (4%) with 4 arm-PEG thiol-10k (1.3 wt%) and 0.05%
2
Irgacure in Ix PBS at pH 7.4 were crosslinked through UV exposure, 10 mwatts/cm for 5 minutes or 250
92
-,- -
-
-
-
-
-
-
-
-
-.- " ,
" ......
.........
"
:Z-L
mwatts/cm 2 for 3.5 seconds. Solutions of grafted polypeptides having stoichiometric ratios of norbornene
to thiol ranging from 0.56 to 2.20 formed gels when exposed to UV as predicted by Flory-Stockmayer
2
UV for 5 minutes (slow) or 250 mwatts/cm UV for 3.5 seconds (fast).
Unlike the optically clear gels formed from PPLG-g-(maleimidexEO2,) polypeptides, swollen
gels crosslinked over five minutes at lower UV intensity through PPLG-g-(norbomeneyEO2,), where
y=6.0, 8.0 or 12.2 norbornene groups per PPLG, appeared translucent and showed regular circular
heterogeneities when viewed with brightfield microscopy (Figure 4.10 B. 1, C. 1, and D. 1). Moreover, the
density of circular heterogeneities increased and the diameter decreased with increasing norbornene
substitution. Transparent gels were recovered when the same precursor solutions were crosslinked with
higher UV intensities for only 3.5 seconds and the frequency of circularly heterogeneities drastically
reduced (Figure 4.10 B.2, C.2, and D.2). Substituting E0 2 sidechains with the more hydrophilic 2-(2-(2azidoethoxy)ethoxy)ethanol (E0 3 ) sidechains of PPLG grafted with on average 10.4 norborene groups per
backbone greatly reduced high norbornene substitution (on average 10.4 per PPLG) both low, (Figure
4.10 E.1) ad and E.2).
3.0 per
r=0.56
6.0 per
r=1.07
8.0 per
r=1.47
12.2 per
r=2.20
10.4 per
r=1.67
Figure 4.10: Phase contrast (20x) images of PPLG-g-(norbomeneyEO2z) gels having PPLG grafted with
3.0 to 12.2 norbornene groups per PPLG backbone and fully substituted with either E0 2 (A-D)2 or E0 3(E).
2
Gels were crosslinked with either 10 mwatts/cm UV for 5 minutes (A-E). 1 or 250 mwatts/cm UV for
3.5 seconds (A-E).
Recent small angle neutron scattering studies of aqueous crosslinked thiolene norbornene PEG
hydrogels noted similar findings, reporting a crosslinker length dependent network structure transitioning
from a homogeneous network with long PEG crosslinkers to a two phase structure when crosslinking
through PEG having arms of 4k or around 90 repeat units. The authors of this study accredit phase
93
separation to segregation of the norbornene hydrophobic pendent and crosslinked side chains both during
gelation and especially during aqueous swelling. The norbornene functionality is generally recognized as
being more hydrophobic than maleimides, supporting that PPLG-g-(maleimideEO2z) crosslinked gels do
not phase separate (Shih and Lin 2012). These findings suggest that phase separation in future gels
crosslinked only with norbornene could be limited by increasing the bulk solubility of the crosslinking
grafted PPLG by either 1) increasing the norbornene tether length (Dudowicz, Freed, and Douglas 2011;
Shih and Lin 2012), or 2) grafting a more hydrophilic solubilizing brush, either of longer ethylene glycol
chains or charged groups. An interesting alternative approach might be to graft norbornenes directly to the
PPLG polypeptide with a minimal linker, relying on steric
Swelling measurements of gels crosslinked through PPLG-g-(norborneneyEO2z) showed limited
dependence on the gel stoichiometric crosslinker ratio, as observed for gels crosslinked through PPLG-g(maleimideEO2z). Swelling was also not greatly influenced by rate of crosslinking as controlled by UV
intensity (Figure 4.11).
60
+ PPLG-g-Mal
50
C
-
40
30
A C.10a
G
0
-
PPLG-g-Norslow
PPLG-g-Norfast
AC.40a
A
-----------------------------------
20 ------------
140%
-----------
100%
0
0.00
0.50
1.50
1.00
Mal or Nor :SH Ratio
2.00
Figure 4.11: Swelling ratios of gels crosslinked through PPLG-g-(norborneneyEO2z) (red circles - 10
mwatts/cm 2 UV for 5 minutes, or yellow open circles- 250 mwatts/cm 2 UV for 3.5 seconds), PPLG-g(maleimideEO2z) (blue diamonds) and control PEG gels crosslinked through 8-arm PEG maleimide
(triangles). All gels were crosslinked with 4-arm PEG thiol- 10k at a total 5.3 wt% polymer in gel
precursor. Percentages indicate swollen gel volume as a fraction of the crosslinking volume.
4.3.7
Mechanical properties of hydrogels crosslinked through PPLG-g-(norborneneyEO2,)
AFM elastic moduli measurements for PPLG-g-(norborneneyEO2z) gels crosslinked at low UV
intensities showed gel elastic moduli significantly lower than those observed for gels formed through
maleimide crosslinking at the same stoichiometric ratio (Figure 4.12). However, gels crosslinked at a
stoichiometric ratio of 1.07 and 1.47 had elastic moduli 4 times greater than did gels having a stochastic
94
ratio of 0.56, demonstrating moduli dependence on norbornene substitution. PPLG-gin stiffness
(norbornene6 EO21 54) (r=1.07) gels crosslinked with higher UV intensity showed no increase
(Figure 4.12) compared to the same polymer crosslinked at low UV intensities with significant gel
heterogeneity (Figure 4.10 B.1 and B.2).
-
6.00
M 5.00
.24.00
o3.00
.Y 2.00
1.00
0.00
r= 0.56
r= 1.07
r= 1.47
r= 1.07
slow
slow
slow
fast
Figure 4.12: Elastic moduli quantified by AFM indentation of gels crosslinked through PPLG-g10 mwatts/cm
(norborneneyEO2,) having stoichiometric ratios (r) as reported and crosslinked with UV at
2
Decreased mechanical properties of gels crosslinked through PPLG-g-(norborneneyEO2,)
rate
compared to those through PPLG-g-(maleimide,EO2,) support visual observation of a crosslinking
and solubility dependent phase separation of the swollen PPLG-g-(norborneneyEO2,) gels.
Swelling and mechanical properties of hydrogels crosslinked through PPLG-g4.3.8
(maleimidenorborneneyEO2.)
A third set of grafted PPLG macromers demonstrates the potential to form gels with PPLG
of
backbones having both maleimide and norbornene crosslinkers, and supports the stochastic nature
only
grafting with multiplexed groups. Without UV exposure, PPLG macromers crosslink with thiols
thiols
through maleimides, forming a single crosslinked gel. When exposed to a UV source, remaining
can crosslink through grafted norbornenes, forming a double crosslinked gel. In an alternative strategy,
residual norbornenes may conjugate with soluble tetrazine-conjugated functional groups.
PPLG polymers grafted with both norbornenes and maleimides were synthesized in a two-stage
click reaction similar to the approaches to single crosslinker grafting outlined above, but with grafting
both norbornene- and maleimide-conjugated azides onto the PPLG polypeptides before grafting with
1.3 wt% 4excess E2 (Table 4.1). Solutions containing 4 wt% PPLG-g-maleimide and norbomene with
arm PEG thiol-IOk and 0.05wt % Irgacure at pH 5.2 were incubated for 45 minutes, forming single
at10
crosslinked gels. Double crosslinked gels resulted from exposing single crosslinked gels to UV
mwatts/cm 2 for 5 minutes. All gels were swollen overnight in PBS.
Single and double crosslinked gels show swelling ratios (Figure 4.13A) and fraction of the
precursor polymer crosslinked after swelling (Figure 4.14B) closely matching those of gels crosslinked
95
through PPLG grafted with only a single functional group. Gels both single and double crosslinked
through PPLG-g-(maleimide 28. norbomene. 4EO2 149) (Ix r=0.56, 2x r-2.60) and PPLG-g(maleimide 4. norbomene 46 EO2 151) (Ix r=0.77, 2x r-2.40), demonstrated almost complete polymer
retention after swelling, consistent swelling ratios, and swollen gel volume averaging 70% of the
crosslinking volume (Figure 4.13A and B).
E 0 25
- --
20
~
--
--------
30
A-
-
-
A
- -----
5
E
10
tw
5
0 .-
C
0
--
1.2-
C
1.0
Z
E
0
-
12
- .0 4
.-
..--
j
nr(..er4o(46pr
norC. p0)
2.
28
~
7
PPGg-a212pr)
P8g-a(. pr,
PL-g-a(.
e)
Figure 4.13: Single and double crosslinking of three starting polymers, PPLG-g(maleimidexnorborneneyEO2z) fully substituted with EO2. A. Swelling ratios of gels, B. Fraction of
polymer crosslinked into the gel calculated as mass of dry polymer/mass polymer in gel precursor, and C.
AFM elastic moduli of gels after swelling in PBS.
96
PPLG-g-(maleimide1. 2norbomene8 EO2 151) (lx r=0.24, 2x r=2.00) when single crosslinked formed
a gel, but only retained 70% of the crosslinking polymer after swelling (Figure 4.13B). Double
crosslinking allowed nearly complete polymer retention (Figure 4.14B). Polymer mass loss for the single
crosslinked gel was attributed to both non-crosslinking of the significant fraction of polymer expected to
have no grafting groups (Figure) and incomplete crosslinking of gels with only 1 or 2 grafted maleimides
(Figure 4.14).
0.5 -
1
(D
W
0
0.4
- 0.9
00
0
0
-0
0.
0.8 3
0
0.2
00
.
00
00
t
0
0.
0,
02
0.1
0-0
0.00
0
00
%g
.
.6
f.
0.6
0.5
1
1.5
2
%graft.
0.80
0.60
0.40
0.20
2.5
3
r
Figure 4.14: Left axis (squares) shows theoretical estimates of the stoichiometric or molar fraction of
PPLG DOP 160 crosslinked with no crosslinker (f=0) for a given "% graft." (average % substitution
crosslinking per PPLG). For gels crosslinked with 4 wt% grafted PPLG and 1.3 wt% 4-arm PEG thiol10k, each % graft corresponds to gels having a given r (molar ratio maleimide:thiol) as shown on the
lower x axis. Right axis (circles) shows the maximum fraction polymer incorporated, calculated as mass
of dry polymer/mass polymer in gel precursor, for crosslinked through PPLG with a given % graft. and r,
assuming no incorporation of PPLG polypeptides grafted with no crosslinkers
Significant literature has demonstrated the utility of temporal and spatial control of gel
mechanical properties (Khetan, Katz, and Burdick 2009; Yang et al. 2014). In grafted-PPLG gels,
mechanical properties can be tuned by introducing crosslinkers with orthogonal temporal control. Elastic
moduli measurements show single and double crosslinked gels as promising platforms to modulate gel
mechanical properties in a physiologically relevant range. As seen with gels having only grafted
maleimides, single crosslinked gels have stiffness increasing with PPLG maleimide substitution. For each
grafted PPLG, double crosslinked gels were consistently stiffer by AFM than gels with only single
crosslinking, presumably due to norbornene conjugation to pendent thiol groups (Figure 4.14C). The
extent of stiffness increase with double crosslinking would likely be increased with an alternative UV
reactive group with greater solubility and less tendency towards phase separation. constraints to limit
norbornene hydrophobic pendent aggregation.
97
4.3.9
Cell adhesion to 2D hydrogels crosslinked through PPLG-g-(maleimide.EO2.)
The appropriateness of PPLG hydrogels as modular 2D synthetic extracellular matrices was
explored by characterizing cell response to soft and hard grafted-PPLG hydrogels functionalized with the
integrin engaging peptide, RGD, and a negative control peptide, RGE. Mesenchymal stem cells are well
known to attach and spread only on 2D PEG gels that are both sufficiently stiff and tethered with an
adhesive peptide, here RGD (MacQueen, Sun, and Simmons 2013). Without sufficient mechanical and
adhesive cues, these cells fail to adhere to a gel or remain as balls. Grafted PPLG gels with RGE did not
support cell adhesion and spreading, even of cells seeded in full serum (Figure 4.14). The limited nonspecific adhesion of cells to grafted PPLG hydrogels is attributed to the anti-fouling properties of the E0
2
brush (Prime and Whitesides 1993). Significantly, cells also failed to adhere to grafted PPLG gels with
stoichiometric ratios greater than one, suggesting limited non-specific conjugation of serum proteins even
in gels having excess reactive maleimides. As expected, the most robust adhesion and cell spreading was
observed for cells seeded on the stiffest PPLG gels with adhesive peptide, supporting AFM based
characterization of gel elastic moduli.
A.
O.S
-
-
B.
+
__n
10.34
0.2
C.1
0
kPa
No peptide
C.
2
9
17
PPLG-g-maleimide
12
16
9
PEG maleimide Sarm
Figure 4.15: Adhesion of hTMSCs seeded on PPLG and PEG hydrogels after 6 hours culture A. Number
of cells on each gel type after 6 hours culture as a fraction of adherent cells on tissue culture poly(styrene)
(TCPS). Reported gel stiffness corresponds to those reported in Figure 4.8, without peptide incorporation.
B. Image of fixed and stained rounded cells seeded on soft PPLG hydrogels. B. Image of fixed and
stained well-spread cells seeded on hard PPLG hydrogel.
4.3.10 Cell viability in 3D hydrogels crosslinked through PPLG-g-(norborneneyEO2.)
The biocompatibility and stability of PPLG-g-(norbomeneEO2154 ) hydrogels were further
explored via the 3D encapsulation of hTMSCs. Cells were UV-crosslinked into PPLG and PEG-only gels
and cultured in full serum media. At 0, 1, 3, 5, and 7 days, the fraction of live cells was quantified using
Invitrogen Live/Dead stain. Cells encapsulated in well-established PEG norbornene gels showed only
98
slightly higher viability than those crosslinked through grafted PPLG (Figure 4.13). This preliminary
screen suggests that E02-grafted PPLG is non-toxic to cultured cells and, with optimization, PPLG based
hydrogel could be excellent engineered extracellular matrices for in vitro and in vivo applications.
PEG Norbornene
PPLG Norbornene
-_--
-
90
80
70
-
60
-
50
40
30
.
-
20
10
0
1
3
Day
5
7
Figure 4.16: Overall viability as quantified by Live/Dead staining of hTMSC encapsulated in hydrogels
(25 piL, 15k cells) crosslinked through 4-arm PEG norbomene- 10k or PPLG-g-(norbornene6 EO2 154) with
4-arm PEG thiol- 10k through UV irradiation 10 mwatts/cm 2 for 5 minutes.
Cell viability within both PPLG and control gels is expected to be improved through increasing
the concentration of gel-incorporated adhesive ligand. Previous studies encapsulating human
mesenchymal stem cells in unmodified PEG diacrylate gels showed only 15% viability after 1 week
compared to 75% in gels incorporating 2.8 mM RGD, suggesting the potential for increasing viability by
increasing the RGD concentration of 250 pM used in this study (Nuttelman, Tripodi, and Anseth 2005).
In a related strategy, viability might be improved by grafting the more potent cyclic RGD adhesive group
154
)
(Kilian and Mrksich 2012). Characterizing cells encapsulated in PPLG-g-(norborneneEO2
crosslinked gels would also benefit from increased optical clarity by crosslinking more quickly, either
through short bursts of higher UV intensities as outlined above or by using a more efficient photoinitiator
such as lithium phenyl-2,4,6- trimethylbenzoylphosphinate (Fairbanks, Schwartz, Bowman, et al. 2009).
Finally, while both PEG and PPLG gels systems remained intact for the 7 day culture period, PPLG gels
noticeably degraded by day 7, as PPLG is well-known to degrade through ester hydrolysis of side chain
propargyl groups (Engler, Bonner, et al. 2011). While gel degradation is a feature for some in vivo
applications, long term cell studies may benefit from crosslinking through more hydrolytically stable
polypeptides (Barz, Duro-Castano, and Vicent 2013; Huang et al. 2010).
4.4
Conclusions
Step-growth hydrogels crosslinked through grafted PPLG is a promising platform for extending
the potential of PEG hydrogels, especially for applications in tissue engineering. Foundational
characterization presented in this work demonstrates a modular, well-controlled synthetic platform for
99
synthesizing crosslinker-grafted PPLG, easily extended to a wide variety of covalent crosslinking
chemistries. Characterization of gels crosslinked through grafted PPLG supports robust control of bulk
hydrogel properties through systematically modulating the grafted polypeptide's nanostructure. Further,
swelling ratios, polymer incorporation, and bulk gel stiffness measurements strongly support both
stochastic substitution of PPLG grafting groups, and the a-helical secondary structure of grafted-PPLG
even when crosslinked into a gel. Preliminary studies identify grafted PPLG as supporting both 2D and
3D culture of hTMSCS. PPLG-g-(maleimideEO2z) hydrogels were shown to support specific-2D cell
adhesion through short, thiol terminated peptides, where cell adhesion could also be modified by
adjusting PPLG functionalization. 3D studies suggest minimal toxicity of grafted PPLG.
Together, these findings recommend grafted PPLG hydrogels as a promising platform for
investigating and controlling cellular response. Future studies look to leverage robust chemical control of
grafted functionality to offer unparalleled spatial and temporal control of bulk mechanical properties and
biofunctionalilty. While this study focuses on the specific polypeptide, PPLG, with a short PEG brush and
maleimide and norbome crosslinkers, these findings readily expanded to direct design of other grafted
polypeptide hydrogels with other grafting strategies, especially of polypeptides having defined a-helical
secondary structure (Deng et al. 2014; Feng et al. 2011; Rhodes and Deming 2013; Lu et al. 2014). As
such, this thesis presents a broadly applicable practical platform and theoretical framework for
crosslinked polypeptide hydrogels.
4.5
References
Alge, Daniel L., and Kristi S. Anseth. 2013. Thiol-X Chemistries in Polymer and MaterialsScience.
Edited by Andrew Lowe and Christopher Bowman. RSC Polymer Chemistry Series. Cambridge:
Royal Society of Chemistry. doi:10.1039/9781849736961.
Alge, Daniel L., Malar Azagarsamy, Dillon F. Donohue, and Kristi S. Anseth. 2013. "Synthetically
Tractable Click Hydrogels for Three-Dimensional Cell Culture Formed Using Tetrazine-Norbornene
Chemistry." Biomacromolecules 14 (4): 949-53. doi:10.1021/bm4000508.
Alge, Daniel L., Dillon F. Donohue, and Kristi S. Anseth. 2013. "Facile and Efficient Lewis Acid
Catalyzed Synthesis of an Asymmetric Tetrazine Useful for Bio-Orthogonal Click Chemistry
Applications." TetrahedronLetters 54 (41). Elsevier Ltd: 5639-41. doi: 10.101 6/j.tetlet.2013.08.010.
Baldwin, Aaron D., and Kristi L. Kiick. 2011. "Tunable Degradation of Maleimide-Thiol Adducts in
Reducing Environments." Bioconjugate Chemistry 22 (10): 1946-53. doi:10.1021/bc200148v.
Baldwin, Aaron D., and Kristi L. Kiick. 2013. "Reversible Maleimide-thiol Adducts Yield GlutathioneSensitive Poly (ethylene Glycol)-heparin Hydrogels." Polymer Chemistry, 133-43.
doi: 10.1039/c2py20576a.
Barz, Matthias, Aroa Duro-Castano, and Maria J. Vicent. 2013. "A Versatile Post-Polymerization
Modification Method for Polyglutamic Acid: Synthesis of Orthogonal Reactive Polyglutamates and
Their Use in 'click Chemistry."' Polymer Chemistry 4 (10): 2989. doi:10.1039/c3pyOOI89j.
Bray, James C., and Edward W. Merrill. 1973. "Poly(vinyl Alcohol) Hydrogels. Formation by Electron
Beam Irradiation of Aqueous Solutions and Subsequent Crystallization." Journal ofApplied
Polymer Science 17 (12): 3779-94. doi: 10.1002/app. 1973.070171219.
100
.
Chakrabarti, Buddhapriya, and Alex Levine. 2005. "Nonlinear Elasticity of an A-Helical Polypeptide."
Physical Review E 71 (3): 031905. doi:10.1 103/PhysRevE.71.031905.
Chen, Ping, Chuang Li, Dongsheng Liu, and Zhibo Li. 2012. "DNA-Grafted Polypeptide Molecular
Bottlebrush Prepared via Ring-Opening Polymerization and Click Chemistry." Macromolecules 45
(24): 9579-84. doi:10.1021/ma302233m.
Chen, Xiaoqiang, Ying Zhou, Xiaojun Peng, and Juyoung Yoon. 2010. "Fluorescent and Colorimetric
Probes for Detection of Thiols." Chemical Society Reviews 39 (6): 2120-35. doi:10.1039/b925092a.
Cheng, Yilong, Chaoliang He, Chunsheng Xiao, Jianxun Ding, Haitao Cui, Xiuli Zhuang, and Xuesi
Chen. 2013. "Versatile Biofunctionalization of Polypeptide-Based Thermosensitive Hydrogels via
Click Chemistry." Biomacromolecules 14 (2): 468-75. doi:10.1021/bm3017059.
Chopko, Caroline M., Erika L. Lowden, Amanda C. Engler, Linda G. Griffith, and Paula T. Hammond.
2012. "Dual Responsiveness of a Tunable Thermosensitive Polypeptide." ACS Macro Letters 1:
727-31. doi:10.1021/mz300088w.
Cima, Linda Griffith, and Stephanie T. Lopina. 1995. "Network Structures of Radiation-Crosslinked Star
Polymer Gels." Macromolecules 28 (20): 6787-94. doi:10. 1021/ma00124a013.
Cok, Alexandra M., Huaxing Zhou, and Jeremiah Johnson. 2013. "Synthesis of Model Network
Hydrogels via Tetrazine-Olefin Inverse Electron Demand Diels-Alder Cycloaddition."
MacromolecularSymposia 329 (1): 108-12. doi: 10.1002/masy.201300008.
DeForest, Cole, Brian D. Polizzotti, and Kristi S. Anseth. 2009. "Sequential Click Reactions for
Synthesizing and Patterning Three-Dimensional Cell Microenvironments." Nature Materials 8 (8).
Nature Publishing Group: 659-64. doi:10.1038/nmat2473.
Deng, Chao, Jintian Wu, Ru Cheng, Fenghua Meng, Harm Anton Klok, and Zhiyuan Zhong. 2014.
"Functional Polypeptide and Hybrid Materials: Precision Synthesis via A-Amino Acid NCarboxyanhydride Polymerization and Emerging Biomedical Applications." Progressin Polymer
Science 39 (2). Elsevier Ltd: 330-64. doi: 10.101 6/j.progpolymsci.2013.10.008.
Ding, Jianxun, Xiuli Zhuang, Chunsheng Xiao, Yilong Cheng, Li Zhao, Chaoliang He, Zhaohui
Tanga and Xuesi Chen. 2011. "Preparation of Photo-Cross-Linked pH-Responsive Polypeptide
Nanogels as Potential Carriers for Controlled Drug Delivery." JournalofMaterials Chemistry 21:
11383-91. doi:10.1039/cljmlO391a.
Dudowicz, Jacek, Karl F. Freed, and Jack F. Douglas. 2011. "Can the Miscibility of Telechelic Polymer
Solutions Increase with Polymer Chain Length?" ACS Macro Letters 1: 88-91.
doi:10.1021/mz200101p
Elbert, Donald L., Alison B. Pratt, Matthias P. Lutolf, Sven Halstenberg, and Jeffrey Hubbell. 2001.
"Protein Delivery from Materials Formed by Self-Selective Conjugate Addition Reactions." Journal
of ControlledRelease 76 (1-2): 11-25. doi: 10.1016/S0168-3659(01)00398-4.
Engler, Amanda C., Hyung-il Lee, and Paula T. Hammond. 2009. "Highly Efficient 'Grafting onto' a
Polypeptide Backbone Using Click Chemistry." Angewandte Chemie (InternationalEd. in English)
48 (49): 9334-38. doi:10.1002/anie.200904070.
Engler, Amanda C., Anita Shukla, Sravanthi Puranam, Hilda G. Buss, Nina Jreige, and Paula T.
Hammond. 2011. "Effects of Side Group Functionality and Molecular Weight on the Activity of
Synthetic Antimicrobial Polypeptides." Biomacromolecules 12 (5): 1666-74.
doi:10.1021/bm2000583.
Engler, Amanda C., Daniel K. Bonner, Hilda G. Buss, Eva Y. Cheung, and Paula T. Hammond. 2011.
"The Synthetic Tuning of Clickable pH Responsive Cationic Polypeptides and Block
Copolypeptides." Soft Matter 7 (12): 5627. doi:10.1039/c lsm05O64h.
Fairbanks, Benjamin D., Michael P. Schwartz, Christopher N. Bowman, and Kristi S. Anseth. 2009.
"Photoinitiated Polymerization of PEG-Diacrylate with Lithium Phenyl-2,4,6Trimethylbenzoylphosphinate: Polymerization Rate and Cytocompatibility." Biomaterials 30 (35).
Elsevier Ltd: 6702-7. doi: 10.101 6/j.biomaterials.2009.08.055.
Fairbanks, Benjamin D., Michael P. Schwartz, Alexandra E. Halevi, Charles R. Nuttelman, Christopher
N. Bowman, and Kristi S. Anseth. 2009. "A Versatile Synthetic Extracellular Matrix Mimic via
101
Thiol-Norbomene Photopolymerization." Advanced Materials 21 (48): 5005-10.
doi:10. 1002/adma.200901808.
Feng, Chun, Yongjun Li, Dong Yang, Jianhua Hu, Xiaohuan Zhang, and Xiaoyu Huang. 2011. "WellDefined Graft Copolymers: From Controlled Synthesis to Multipurpose Applications." Chemical
Society Reviews 40 (3): 1282-95. doi:10.1039/b921358a.
Flory, Paul J. 1941. "Molecular Size Distribution in Three Dimensional Polymers. 1. Gelation 1." Journal
ofthe American Chemical Society 63 (11): 3083-90. doi: 10.1021/jaOl 856a061.
. 1950. "Statistical Mechanics of Swelling of Network Structures." The Journalof Chemical
Physics 18 (1): 108. doi:10.1063/1.1747424.
Flory, Paul J., and John Rehner. 1943a. "Statistical Mechanics of Cross-Linked Polymer Networks II.
Swelling." The Journalof Chemical Physics 11 (11): 521. doi:10.1063/1.1723792.
. 1943b. "Statistical Mechanics of Cross-Linked Polymer Networks I. Rubberlike Elasticity." The
Journalof Chemical Physics 11 (11): 512. doi:10.1063/1.1723791.
Garcia, Andres J. 2014. "PEG-Maleimide Hydrogels for Protein and Cell Delivery in Regenerative
Medicine." Annals of Biomedical Engineering42 (2): 312-22. doi:10.1007/si0439-013-0870-y.
Gramlich, William M., Iris L. Kim, and Jason Burdick. 2013. "Synthesis and Orthogonal Photopatteming
of Hyaluronic Acid Hydrogels with Thiol-Norbornene Chemistry." Biomaterials34 (38). Elsevier
Ltd: 9803-11. doi: 10.101 6/j.biomaterials.2013.08.089.
Guvendiren, Murat, and Jason Burdick. 2012. "Stiffening Hydrogels to Probe Short- and Long-Term
Cellular Responses to Dynamic Mechanics." Nature Communications3 (January). Nature
Publishing Group: 792. doi:10.1038/ncommsl792.
Gyenes, Tamis, Vikt6ria Torma, Benjimin Gyarmati, and Mikl6s Zrinyi. 2008. "Synthesis and Swelling
Properties of Novel pH-Sensitive Poly(Aspartic Acid) Gels." Acta Biomaterialia4: 733-44.
doi: 10.1016/j.actbio.2007.12.004.
Han, Hee-Sun, Neal K. Devaraj, Jungmin Lee, Scott Hilderbrand, Ralph Weissleder, and Moungi G.
Bawendi. 2010. "Development of a Bioorthogonal and Highly Efficient Conjugation Method for
Quantum Dots Using Tetrazine-Norbornene Cycloaddition." Journalof the American Chemical
Society 132 (23): 7838-39. doi:10.1021/ja101677r.
He, Chaoliang, Xiuli Zhuang, Zhaohui Tang, Huayu Tian, and Xuesi Chen. 2012. "Stimuli-Sensitive
Synthetic Polypeptide-Based Materials for Drug and Gene Delivery." Advanced Healthcare
Materials 1: 48-78. doi:10.1002/adhm.201100008.
Head, David A., Alex J. Levine, and F. C. Mackintosh. 2003. "Deformation of Cross-Linked Semiflexible
Polymer Networks." Physical Review Letters 91 (September): 2-5.
doi:10.1 103/PhysRevLett.91.108102.
Huang, Jin, Gijs Habraken, Fabrice Audouin, and Andreas Heise. 2010. "Hydrolytically Stable Bioactive
Synthetic Glycopeptide Homo-and Copolymers by Combination of Nca Polymerization and Click
Reaction." Macromolecules 43: 6050-57. doi:10.1021/mal01096h.
Kharkar, Prathamesh M., Kristi L. Kiick, and April M. Kloxin. 2013. "Designing Degradable Hydrogels
for Orthogonal Control of Cell Microenvironments." ChemicalSociety Reviews 42: 7335-72.
doi: 10. 1039/c3cs6004Oh.
Khetan, Sudhir, Joshua S. Katz, and Jason Burdick. 2009. "Sequential Crosslinking to Control Cellular
Spreading in 3-Dimensional Hydrogels." Soft Matter 5 (8): 1601. doi:10.1039/b820385g.
Kilian, Kristopher, and Milan Mrksich. 2012. "Directing Stem Cell Fate by Controlling the Affinity and
Density of Ligand-Receptor Interactions at the Biomaterials Interface." Angewandte Chemie
(InternationalEd. in English) 51 (20): 4891-95. doi:10.1002/anie.201108746.
Lee, Kuen Yong, Jon A. Rowley, Petra Eiselt, Erick M. Moy, Kamal H. Bouhadir, and David J. Mooney.
2000. "Controlling Mechanical and Swelling Properties of Alginate Hydrogels Independently by
Cross-Linker Type and Cross-Linking Density." Macromolecules 33 (11): 4291-94.
doi: 10.1021/ma9921347.
Lu, Hua, Jing Wang, Ziyuan Song, Lichen Yin, Yanfeng Zhang, Haoyu Tang, Chunlai Tu, Yao Lin, and
Jianjun Cheng. 2014. "Recent Advances in Amino Acid N-Carboxyanhydrides and Synthetic
102
-
Polypeptides: Chemistry, Self-Assembly and Biological Applications." Chemical Communications
(Cambridge, England) 50 (2): 139-55. doi:10.1039/c3cc46317f.
Lutolf, Matthias P., and Jeffery A. Hubbell. 2003. "Synthesis and Physicochemical Characterization of
End-Linked Poly(ethylene Glycol)-Co-Peptide Hydrogels Formed by Michael-Type Addition."
Biomacromolecules 4 (3): 713-22. doi:10.1021/bm025744e.
MacQueen, Luke, Yu Sun, and Craig A. Simmons. 2013. "Mesenchymal Stem Cell Mechanobiology and
Emerging Experimental Platforms." Journal of the Royal Society, Interface / the Royal Society 10
(84): 20130179. doi:10.1098/rsif.2013.0179.
Markland, Peter, Yuehua Zhang, G L Amidon, and V C Yang. 1999. "A pH- and Ionic StrengthResponsive Polypeptide Hydrogel: Synthesis, Characterization, and Preliminary Protein Release
Studies." JournalofBiomedical Materials Research 47 (4): 595-602. doi:10.1002/(SICI)10974636(19991215)47:4<595::AID-JBMI 7>3.0.CO;2-I.
Nguyen, Kaytai Truong, and Jenn West. 2002. "Photopolymerizable Hydrogels for Tissue Engineering
Applications." Biomaterials23: 4307-14. doi: 10.1007/978-1-4419-5919-5 11.
Nuttelman, Charles R, Margaret C Tripodi, and Kristi S Anseth. 2005. "Synthetic Hydrogel Niches That
Promote hMSC Viability." Matrix Biology: Journalof the InternationalSociety for Matrix Biology
24 (3): 208-18. doi: 10.101 6/j.matbio.2005.03.004.
Oelker, Abigail M., Shannon M. Morey, G Griffith, Paula T. Hammond, and Linda G. Griffith. 2012.
"Helix versus Coil Polypeptide Macromers: Gel Networks with Decoupled Stiffness and
Permeability." Soft Matter 8 (42): 10887. doi:10.1039/c2sm26487k.
Oesterhelt, F, M Rief, and H E Gaub. 1999. "Single Molecule Force Spectroscopy by AFM Indicates
Helical Structure of Poly(ethylene-Glycol) in Water." New Journal ofPhysics 1 (January): 6-6.
doi: 10.1088/1367-2630/1/1/006.
Okamoto, Takeshi, Tomoki Aoyama, Tomitaka Nakayama, Takeharu Nakamata, Taisuke Hosaka, Koichi
Nishijo, Takashi Nakamura, Tohru Kiyono, and Junya Toguchida. 2002. "Clonal Heterogeneity in
Differentiation Potential of Immortalized Human Mesenchymal Stem Cells." Biochemical and
Biophysical Research Communications 295 (2): 354-61. doi: 10.101 6/S0006-291X(02)00661-7.
Oliveira, Eder D., Samuel G. Hirsch, Richard J. Spontak, and Stevin H. Gehrke. 2003. "Influence of
Polymer Conformation on the Shear Modulus and Morphology of Polyallylamine and Poly(a-L
Lysine) Hydrogels." Macromolecules 36 (16): 6189-6201. doi:10.1021/ma021468k.
Patterson, David M., Lidia Nazarova, and Jennifer Prescher. 2014. "Finding the Right (bioorthogonal)
Chemistry." ACS Chemical Biology 9: 592-605. doi:10.1021/cb400828a.
Peppas, Nikolaos A., and Edward W. Merrill. 1976. "Poly(vinyl Alcohol) Hydrogels: Reinforcement of
Radiation-Crosslinked Networks by Crystallization." JournalofPolymer Science: Polymer
Chemistry Edition 14 (2): 441-57. doi:10.1002/pol.1976.170140215.
Phelps, Edward A., Nduka 0. Enemchukwu, Vincent F. Fiore, Jay C. Sy, Niren Murthy, Todd Sulchek,
Thomas H. Barker, and Andr6s J. Garcia. 2011. "Maleimide Cross-Linked Bioactive PEG Hydrogel
Exhibits Improved Reaction Kinetics and Cross-Linking for Cell Encapsulation and in Situ
Delivery." Advanced Materials (DeerfieldBeach, Fla.) 24 (1): 64-70, 2.
doi:10.1002/adma.201103574.
Platte, Daniela, Uta Helbig, Ruth Houbertz, and Gerhard Sextl. 2011. "Localization of Critical Molar
Ratio Intervals for Highly Branched Step-Growth Michael Addition Networks." Macromolecules
44: 5123-26. doi:10.1021/ma20081Oe.
Prime, Kevin L., and George M. Whitesides. 1993. "Adsorption of Proteins onto Surfaces Containing
End-Attached Oligo(ethylene Oxide): A Model System Using Self-Assembled Monolayers."
Journal of the American Chemical Society 115 (23): 10714-21. doi: 10.102 1/ja00076a032.
Quadir, Mohiuddin, Mackenzie Martin, and Paula T. Hammond. 2014. "Clickable Synthetic Polypeptides
- Routes to New Highly Adaptive Biomaterials." Chemistry ofMaterials 26: 461-76.
doi: 10.102 1/cm4023997.
103
Rhodes, Allison J., and Timothy J. Deming. 2013. "Soluble, Clickable Polypeptides from AzideContaining N -Carboxyanhydride Monomers." A CS Macro Letters 2 (5): 351-54.
doi: 10.102 1/mz4001089.
Sakai, Takamasa, Takuro Matsunaga, Yuji Yamamoto, Chika Ito, Ryo Yoshida, Shigeki Suzuki, Nobuo
Sasaki, Mitsuhiro Shibayama, and Ung-il Chung. 2008. "Design and Fabrication of a High-Strength
Hydrogel with Ideally Homogeneous Network Structure from Tetrahedron-like Macromonomers."
Macromolecules 41 (14): 5379-84. doi: 10.102 1/ma800476x.
Shih, Han, and Chien-Chi Lin. 2012. "Cross-Linking and Degradation of Step-Growth Hydrogels Formed
by Thiol-Ene Photoclick Chemistry." Biomacromolecules 13 (7): 2003-12. doi:10.1021/bm300752j.
Shih, Ke-ying, Tai-shen Hsiao, Shiang-lin Deng, and Jin-long Hong. 2014. "Water-Soluble Poly(7Propargyl- L -Glutamate) Containing Pendant Sulfonate Ions and Terminal Fluorophore:
Aggregation-Enhanced Emission and Secondary Structure." Macromolecules 47 (12): 4037-47.
doi: 10.1021/ma500886v.
Stockmayer, Walter H. 1943. "Theory of Molecular Size Distribution and Gel Formation in BranchedChain Polymers." The Journalof Chemical Physics 11 (2): 45. doi:10.1063/1.1723803.
Tang, Haoyu, and Donghui Zhang. 2010. "General Route toward Side-Chain-Functionalized AlphaHelical Polypeptides." Biomacromolecules 11 (6): 1585-92. doi:10.1021/bmi1002174.
Vanderhooft, Janssen L., Brenda K. Mann, and Glenn D. Prestwich. 2007. "Synthesis and
Characterization of Novel Thiol-Reactive Poly(ethylene Glycol) Cross-Linkers for ExtracellularMatrix-Mimetic Biomaterials."Biomacromolecules 8 (9): 2883-89. doi:10.1021/bm0703564.
Wang, Jing, Hua Lu, Yuan Ren, Yanfeng Zhang, Martha Morton, Jianjun Cheng, and Yao Lin. 2011.
"Interrupted Helical Structure of Grafted Polypeptides in Brush-like Macromolecules."
Macromolecules 44: 8699-8708. doi: 10. 1021/ma201833b.
Xiao, Chunsheng, Changwen Zhao, Pan He, Zhaohui Tang, Xuesi Chen, and Xiabin Jing. 2010. "Facile
Synthesis of Glycopolypeptides by Combination of Ring-Opening Polymerization of an AlkyneSubstituted N-Carboxyanhydride and Click 'Glycosylation'." MacromolecularRapid
Communications 31 (11): 991-97. doi:10.1002/marc.200900821.
Yan, Congqi, and Darrin J. Pochan. 2010. "Rheological Properties of Peptide-Based Hydrogels for
Biomedical and Other Applications." Chemical Society Reviews 39 (9): 3528-40.
doi: 10. 1039/b9l9449p.
Yang, Chun, MW Tibbitt, Lena Basta, and Kristi S. Anseth. 2014. "Mechanical Memory and Dosing
influence Stem Cell Fate." Nature Materials, no. March: 1-8. doi: 10. 1038/NMAT3889.
Zhang, Zhe, Li Chen, Mingxiao Deng, Yunyan Bai, Xuesi Chen, and Xiabin Jing. 2011. "Biodegradable
Thermo- and pH-Responsive Hydrogels for Oral Drug Delivery." JournalofPolymer Science Part
A: Polymer Chemistry 49 (13): 2941-5 1. doi:10.1002/pola.24730.
Zhu, Junmin. 2010. "Bioactive Modification of Poly(ethylene Glycol) Hydrogels for Tissue Engineering."
Biomaterials31 (17). Elsevier Ltd: 4639-56. doi:10.1016/j.biomaterials.2010.02.044.
104
5
Thesis summary and future works
5.1
Summary of thesis
This thesis explores both the utility and practicality of hydrogels formed through step-growth
crosslinking between multiarm PEG macromers and PPLG grafted with complementary crosslinking
groups. As such, it develops a platform to integrate the significant nanoscale control of NCA-polymerized
grafted polypeptides with well-established techniques of step-growth hydrogels to enable design and
fabrication of well-controlled and highly useful synthetic extracellular matrices.
Chapter 1 introduces a new theoretical framework for visualizing the nano-scale structure of
grafted polypeptide hydrogels. Results presented in Chapter 2 identify grafting PPLG with 2-(2azidoethoxy) ethanol as sufficient to solubilize PPLG in aqueous buffers and strongly support the
stochastic nature of grafting azide functionalized small molecules to PPLG pendent alkynes through
copper catalyzed 1,3-cycloaddition. Chapter 3 introduces the first step-growth hydrogels crosslinked
through a grafted polypeptide. It also demonstrates grafting on of the biofunctional adhesive peptide,
RGD, thereby demonstrating the potential for PPLG crosslinkers to additionally present biofunctional
groups. Chapter 4 introduces gels crosslinked through the Michael-type addition between maleimides and
thiols, as well as the radical mediated addition of a thiolene to PPLG-grafted norbornene. Systematic
screens use these crosslinking systems to demonstrate experimentally: 1) distinct structural differences
between hydrogels crosslinked through grafted polypeptides having defined secondary structure and those
crosslinked through PEG and 2) the utility of PPLG in presenting complementary crosslinking
chemistries.
Signficicantly, this thesis presents a broadly applicable practical platform and theoretical
framework for crosslinked polypeptide hydrogels. Recent reviews highlight the ever growing strategies
for crosslinking PEG only or PEG-polysaccharide gels, many of which could be readily transferred to
grafted polypeptide gels (Patterson, Nazarova, and Prescher 2014; Tang and Becker 2014). Optimal
crosslinking chemistry should be selected depending on gel application. Similarly, while this thesis
introduces fabricating crosslinked neutral polypeptide hydrogels grafted with 2-(2-azidoethoxy) ethanol,
the robust character of the azide alkyne grafting offers significant synthetic flexibility. For example the
ability to introduce a wide variety of charged, short and long side chains allows gels to be highly
engineered for particular applications. Further, while this thesis focuses on the single polypeptide, PPLG,
these findings could readily be expanded to direct the design of other grafted polypeptide hydrogels,
particularly those with defined a-helical secondary structure (Deng et al. 2014; Feng et al. 2011; Rhodes
and Deming 2013; Lu et al. 2014). As such,
Together, novel gels crosslinked through PPLG and other grafted peptides are expected to enable
extended existing hydrogel platforms to enable improved cellular control.
105
5.2
Future work
The techniques and framework presented in this thesis establishes grafted-PPLG as a powerful
system for enabling synthetic control of engineered extracellular matrices, but leaves these application
primarily to future work. Future directions can be divided into two categories: 1) improving synthetic
approaches to expand the usefulness of grafted PPLG-hydrogels, and 2) applying PPLG to clinically
relevant cell systems.
Regarding synthetic improvements, a significant constraint remaining in crosslinking grafted
polypeptide hydrogels is maintaining macromer stability and solubility. Alternative crosslinking strategies
appropriate for a wide variety of cell applications build on foundational characterizations presented in this
thesis of gels crosslinked through nearly ideal crosslinking between maleimides and thiols. Specifically,
sections below outline three specific crosslinking strategies: a protection strategy for improving
maleimide crosslinking (Section 5.2.1.1), enzymatic crosslinking through the sortase enzyme (Section
5.2.1.2), and non-covalent affinity crosslinking through large proteoglycans (Section 5.2.1.3).
Regarding applications, the non-covalent gels presented in Section 5.2.1.3 are expected to be
highly clinically relevant especially for applications such as synthetic cartilage. Preliminary data
generated but not included in this thesis also strongly suggest increased 2D cell adhesion to adhesive
peptides grafted into hydrogels to PPLG rather than through long PEG chains, even at similar bulk gel
properties. These preliminary findings recommend using PPLG as a platform for understanding and
optimizing local control of cell response to hydrogel incorporated biofunctionality (Section 5.2.2).
Finally, a significant utility of hydrogels incorporating grafted polypeptides is for these polypeptides to
serve as a synthetically simple and modular handle to allow hydrogel functionalization with a broad range
of biofunctional groups, particularly proteins. Section 5.2.1 outlines specific strategies for using grafted
PPLG as a modular handle for introducing into hydrogel a wide variety of functional groups. Specific
future work could focus on conjugating EFG and EGF protein binders through aqueous grafting onto geltethered PPLG-g-norbornene, where the PPLG grafted protein is presented as isolated clusters or as
clusters of proteins tethered to a single backbone.
5.2.1
Alternative crosslinking strategies for grafted PPLG hydrogels
This thesis introduces gels crosslinked through three specific crosslinking chemistries: acrylates,
maleimides, and norbornenes. PPLG-grafted step-growth hydrogels crosslinked through these groups
serve as a platform to begin characterizing the fundamental theory and properties governing polypeptide
gel systems. However, future work would benefit from extending crosslinking to chemistries optimized
for a targeted application. Three alternative crosslinking strategies are suggested which build on those
proposed in this thesis: protected chemical crosslinking, enzymatic crosslinking, and affinity crosslinking.
106
5.2.1.1 Chemical Crosslinking
Hydrogels crosslinked through acrylates and maleimides are popular for cell culture application
because both groups react quickly with thiols at neutral pH. However, this high reactivity also introduces
significant synthetic challenges in maintaining stable macromers prior to crosslinking. For example,
amines also conjugate with both acrylates and maleimide groups at neutral pH, though at lower rates than
thiols. Therefore, these crosslinking chemistries are inappropriate for crosslinking PPLG macromers also
grafted with crosslinking small molecule amines or peptides having amine functionality. Experimental
results have demonstrated challenges preventing premature crosslinking and autopolymeriation of even
PPLG substituted with greater than 10% of either acrylate or maleimide crosslinking groups, without
additional grafted amines. In contrast, norbornenes are relatively stable and inert until exposed to UV
light during the crosslinking reaction. The group's hydrophobic nature was observed to promote phase
separation in the crosslinking polymer (Chapter 4).
A proposed improvement to the established technique of grafting unprotected maleimides would
be to protect the crosslinking group during synthesis. A promising approach would be to graft on a furan
protected maleimide through the organize click reaction. The furan protecting group is expected to
stabilize the maleimide during synthesis but can be readily removed by heating in dilute, aqueous
conditions immediately prior to gel crosslinking (Sanchez, Pedroso, and Grandas 2011; Elduque et al.
2013; Koehler et al. 2013).
5.2.1.2 Enzymatic crosslinking
A promising alternative to small-molecule chemical crosslinking is enzymatically catalyzed
crosslinking through PPLG-grafted peptides. In enzymatic crosslinking, proteins catalyze amide bond
formation between specific peptide sequences (Teixeira et al. 2012). Compared to small-molecule
chemical crosslinking, enzymatic crosslinking allows increased crosslinking orthogonally with common
chemistries and robust macromer stability. Crosslinking peptides can be grafted onto PPLG either through
the organic grafting of azide-grafted peptides (Chapter 3), or through the aqueous grafting via
preincubation of thiol-terminated crosslinking peptides with PPLG-grafted maleimides. For many
applications, organic grafting is expected to be favorable as it allows crosslinking through sortase
peptides without requiring that all PPLG grafting groups be non-reactive with grafted maleimides.
Unpublished studies explored PPLG crosslinking through Factor XIIIa, a fibrin crosslinking
transglutaminase routinely used to crosslink peptide-grafted PEG gels (Ehrbar et al. 2007). PPLG was
grafted through an organic click reaction with E02, Factor XIIIa crosslinking peptides, and fluorescein,
and incorporated at less than 10% of the total polymer into bulk PEG gels crosslinked through Factor
XIIIa. Fluorescence studies showed quantitative PPLG-g-fluorescein incorporation in the swollen gel.
107
However, the gels failed to form when crosslinked through peptide-grafted PPLG (through PEG= 10
linker), where PPLG was grafted with one peptide and PEG was grafted with the complementary peptide.
Limited crosslinking was attributed to insufficient accessibility of the large crosslinking Factor XIIIa
protein (130 kDa) to the PPLG-grafted peptides.
These preliminary findings highlight inherent steric constraints in crosslinking through a protein
rather than through small molecules. Future work would benefit from extending enzymatic crosslinking to
crosslinking through the sortase enzyme (Chen, Dorr, and Liu 2011; Sijbrandij et al. 2013). For example,
the small enzyme, which rapidly conjugates R1-LPXTG with NH 2-GGG-R 2, both short peptides. Sortase
(25 kDa) is expected to more easily access PPLG-grafted peptides than Factor XIIIa (130 kDa), enabling
more efficient crosslinking without requiring the synthetic challenge of increasing peptide PEG tether
length.
5.2.1.3 Non-covalent affinity crosslinking
A proposed alternative to PPLG gels crosslinked only through chemical crosslinking is
crosslinking through high affinity non-covalent interactions, in which crosslinking is either exclusively
non-covalent, or a combination of both covalent and non-covalent affinity crosslinking. A particularly
well-characterized and clinically useful platform has been developed over the last 10 to 15 years for
localizing proteoglycans including heparin and hyaluronan. For example, the Panitch group has
established polysaccharide-PEG gels crosslinked by affinity interactions between long proteoglycans and
multi-arm PEG, where each arm is substituted by a terminal peptides optimized to bind the specific
proteoglycan. Examples of two specific systems include gels crosslinked through binding of heparin and
the heparinbinding peptide, GKAFAKLAARLYRKAGC-NH2(Mummert 2000; Seal and Panitch 2006a;
Seal and Panitch 2006b; Seal and Panitch 2003) and more recently of hyaluronan and the hyaluronan
binding peptide, GAHWQFNALTVRGGGC-NH 2 (Sharma et al. 2013). Incorporating heparin in
biomaterials has well-established clinical utility due to its growth factor binding, anticoagulant activity,
and anti-angiogenic and apoptotic effects (Liang and Kiick 2013). PPLG grafted with heparin binding
peptides might be designed to form gels that incorporate small amounts of heparin as a vehicle for drug
delivery or that bind heparin as a structural gel crosslinker. Further, mechanical properties could be
compared between gels having PPLG grafted with binding peptides already incorporated into PEGproteoglycan platforms and those bound with peptides routinely used by the Griffith lab to engage cell
surface proteoglycans. Comparisons of gels crosslinked through peptide grafted PPLG with those
crosslinked through peptide-grafted PEG would allow more fundamental studies exploring how structural
differences observed for polypeptide and PEG crosslinked gels systems extend to non-covalent
interactions. More importantly, mechanical and biofunctional properties of grafted polypeptide hydrogels
108
gels crosslinked with proteoglycans are expected readily optimized to maximize this gel system's in vivo
potential.
5.2.2
The influence of local presentation and stiffness on cellular adhesion to PPLG-grafted
peptides
In unpublished studies, hTMSCs seeded on grafted-PPLG gels were consistently observed to be
dramatically more adherent and more spread than cells seeded on PEG hydrogel gels having similar or
higher bulk elastic moduli and the same molar adhesive peptide concentration (i.e. RGD).The improved
cell attachment to grafted-PPLG gels suggests that cells may be more responsive to adhesive peptides
presented from PPLG than from star PEG molecules, possibly because of 1) the improved steric
accessibility of cell integrins to peptides grafted on the EO 2functionalized PPLG brush, or 2) the
increased local stiffness of peptides tethered through short linkers to rigid a-helical polypeptides
compared to that of peptides conjugated through long PEG linkers.
The relative contributions to greater cell adhesion of both increased peptide accessibility and
increased local mechanical stiffness might be explored by comparing the cell responses to PPLG-g-RGD
and E0 2 grafted into bulk PEG gels through two different grafting strategies. In the first strategy, PPLG
could be crosslinked to the bulk gel through many side-chain grafted crosslinkers as presented in this
thesis while in the second, PPLG could be conjugated through a single terminal grafting crosslinker
conjugated through a protected initiator (unpublished data). Both PPLG presentations are expected to
reduce steric hindrance through the E0 2 brush, but only PPLG-g-RGD crosslinked through multiple side
chains is expected to increase local mechanical stiffness of the grafted peptide due to its locally increased
rigidity. Incorporation of fluoresce in tagged RGD can be used to quantify adhesive peptide concentration
across gel systems. A greater understanding of the mechanisms driving robust cell adhesion to PPLG-gRGD is expected to be useful for two reasons: 1) it would enable improved design of synthetic cell culture
environments that promote cell adhesion and 2) it could be used as a highly controlled synthetic system to
begin exploring factors governing cell response to native ECM.
A more careful consideration of local mechanical properties suggests that cell adhesion might be
reasonably improved through polypeptide-mediated increases in local stiffness. Fibroblast-like cell
adhesion on 2D gels long has been recognized as being generally enhanced with increased elastic
modulus of the bulk gel (Ahearne 2014; Trappmann and Chen 2013) with the recently increased focus on
the effects of nanoscale rigidity on cell response (Vogel and Sheetz 2006; Trappmann et al. 2012). At first
approximation, length scale and force estimates suggest that PPLG grafted with a PEG,= 10 crosslinker
conjugated to RGD may positively influence cell response compared to RGD grafted to 8-arm PEG
maleimide- 10k and -40k having arm lengths of PEGn= 2 8 or PEG,,=, 1 4 , respectively. At length scales
dependent on the specific crosslinking polymers, gel mechanical properties are expected to transition
109
from affine to non-affine. Using current polymerization techniques, PPLG can be synthesized having a
maximum degree of polymerization of around 300 corresponding to a helical length of 45 nm. PPLG
readily typically synthesized and most used in this thesis has a length of only 25nm. Also, gels presented
in this thesis are crosslinked with 4-arm PEG -2.5k (n=56 for each arm), making the gels most
represented as rigid polypeptide rods connected by flexible PEG linkers. Taken together, these
assumptions suggest the PPLG-crosslinked hydrogels presented in this thesis are expected to demonstrate
non-affine network behavior around the length scales of tens of nanometers.
In the context of cellular engagement, cells experience gel stiffness through different length
scales including the protein (1-4 nm), assembled proteins such as actin filaments (<10 nm), actin
networks or collagen fibers (<1 im), the whole cell (<10 pm) and multiple cells (<lmm).With non-affine
properties expected at length scales of tens of nanometers cellular mechanical sensing to grafted PPLG
may respond to local polypeptide stiffness at the length scale of individual integrin bonds or nascent
adhesion formations (as small as 30-40 nm) (Shroff et al. 2008; Shroff and Galbraith 2007), but likely
does not contribute to the mechanics of growing adhesions (> 100 nm) or traction mediated migration as
does the non-affine nature of the native extracellular matrix (Wen and Janmey 2013). Further, estimates
of 2-40 pN for the force generated by a single integrin (Morimatsu et al. 2013), is a similar order of
magnitude as that required for a-helical polypeptide extension and of that required for bending (Section
1.5.2). One study of poly(glutamic acid) DOP 40 in water reports forces of 0.04 N m' required to extend
the helical polypeptide at intermediate stiffness, which suggests around 80 pN is required to extend it an
additional 3 nm (Zegarra et al. 2009), well within the mechanical strength expected from a single integrin.
Together, these published results support that both the length scale and modulus of crosslinking
polypeptide could enable nanoscale control of cell engagement inherent in the native extra cellular matrix
but not demonstrated in affine, PEG-only gel systems.
5.2.3
Expanding PPLG grafting groups and introducing grafted group clustering
A significant feature of PPLG crosslinked gel systems is the ability to systematically introduce a
wide range of gel functionality through different grafted groups, including both small molecules and
biologically active peptides and proteins.
5.2.3.1 Grafting on small molecules
The E02 grafting group used in this thesis to solubilized PPLG could be supplemented or
replaced with small molecule cues, such as conjugated charge molecules, including amines (Engler et al.
2011) and phosphate groups, or longer PEG chains (Engler, Lee, and Hammond 2009). Grafting PPLG
with alternative side chains could systematically extend existing literature findings showing how changes
in bulk (Benoit et al. 2008) and local physicochemical gel properties, including hydrophobicity (Ayala et
110
al. 2011) and charge could result in a modified cell response. The figh densities of these small molecules
could also have therapeutic utility in areas of responsive drug delivery and sensing (Hendrickson and
Andrew Lyon 2009).
5.2.3.2 Grafting on proteins and peptides - bulk and nano- organization
Compared to the many published step-growth PEG hydrogels incorporating grafted peptides,
relatively few gels incorporate grafted proteins, primarily because of significant synthetic challenges
inherent to conjugating proteins while maintaining their bioactivity (Zhu 2010; Cabanas-Dands, Huskens,
and Jonkheijm 2014). Recent innovations in synthesizing proteins with well-defined handles for gel
conjugation (e.g. sortase ligation and unnatural amino acids) offer opportunities to expand protein
grafting. PPLG provides a robust, modular platform that covalently binds proteins to hydrogels, both by
streamlining the synthesis and characterization of protein-grafted gels, and by enabling facile nano-scale
organization of gel-grafted proteins across a variety of covalently crosslinked gel platforms.
For applications as a general protein-conjugating platform, PPLG might be grafted with 1) groups
enabling PPLG-crosslinking throughout a target bulk hydrogel, 2) one of a wide variety of conjugating
chemistries appropriate for reacting with the specific target protein, and 3) a label such as a fluorescent
dye. Proteins engineered with a specific handle can be tethered to the PPLG molecule via the appropriate
aqueous click reaction. Protein conjugation can be verified by gel or column size exclusion
chromatography, specifically through monitoring the disappearance of signal from the unconjugated
protein in the grafted product. Protein-grafted PPLG can then be mixed with the bulk gel precursor
solution, where PPLG incorporation into the crosslinked gel can be quantified by monitoring the
incorporation of the PPLG-grafted label. This approach to protein grafting would leverage PPLG's highly
defined stochastic substitution, organized secondary structure, and remarkable synthetic control to link
protein and hydrogel chemistries, which will enable facile conjugation of proteins to synthetic hydrogels.
Further, the nanoscale distribution of grafted biofunctionlity can be modulated by tuning the
molar ratio of PPLG polymer to grafting protein in the aqueous click reaction. A significant
stoichiometric excess of the PPLG polymer would result in proteins being distributed homogeneously
throughout the bulk gel, whereas excess peptide would result in clustered proteins as modeled by
stochastic grafting presented in this thesis. The size of protein clusters can be further controlled by PPLG
DOP and the length of the PEG protein linker. Several recent reviews (Ekerdt, Segalman, and Schaffer
2013; Kiessling, Gestwicki, and Strong 2006; Shekaran and Garcia 2011; Wheeldon et al. 2011) and
research articles (Huang et al. 2009; Lam and Segura 2013; Comisar et al. 2006; Englund et al. 2012;
Toepke et al. 2012) highlight the utility of controlled nanoscale organization and biofunctional clustering
in synthetic ECMs, particularly through this approach of pre-incubation.
111
As a specific example, tetrazine-functionalized epidermal growth factor (EGF) (Krueger et al.
2014) might be conjugated to norbornenes grafted with maleimides onto PPLG, as in the PPLG-gnorbomene and maleimide cases presented in this thesis (though cell accessibility and dimerization of
tethered EGF is expected to be enhanced by grafting the norboronene groups through a longer PEG tether,
such as PEG,= 2 7 ). The polymer-protein complex could be crosslinked into a Michael-type, step-growth
hydrogel through PPLG-grafted maleimides. Homogeneously distributed tethered-EGF has been shown to
dramatically influence cell phenotype (Kuhl and Griffith-Cima 1996; Platt et al. 2009; Mehta et al. 2010;
Gobin and West 2003). It is hypothesized, based on solution-base studies of divalent EGF (Krueger et al.
2014; Jay et al. 2011) that EGF bioactivity may be increased through optimized clustered PPLG-grafting,
though future studies should carefully consider the most appropriate geometries and the entropic effects
governing cell response to tethered multivalent ligands (Martinez-Veracoechea and Leunissen 2013). In
addition to directly grafting EGF, an interesting complementary approach could be to graft tight binders
to EGF, locally concentrating soluble growth factors. Finally, EGF signaling through EGFR has been
demonstrated to signal cooperatively with integrins and other cell surface proteins, suggesting the
interesting potential of using nanoclustering through PPLG tethering to explore the effects of colocalizing tethered EGF and integrin binding domains on cellular response.
References
5.3
Ahearne, Mark. 2014. "Introduction to Cell-Hydrogel Mechanosensing." Interface Focus 4 (2):
20130038. doi:10.1098/rsfs.2013.0038.
Ayala, Ramses, Chao Zhang, Darren Yang, and Yongsung Hwang. 2011. "Engineering the Cell-material
Interface for Controlling Stem Cell Adhesion, Migration, and Differentiation." Biomaterials32 (15).
Elsevier Ltd: 3700-3711. doi: 10.101 6/j.biomaterials.2011.02.004.
Benoit, Danielle S. W., Michael P. Schwartz, Andrew R. Durney, and Kristi S. Anseth. 2008. "Small
Functional Groups for Controlled Differentiation of Hydrogel-Encapsulated Human Mesenchymal
Stem Cells." Nature Materials7 (10): 816-23. doi:10.1038/nmat2269.
Cabanas-Dands, Jordi, Jurriaan Huskens, and Pascal Jonkheijm. 2014. "Chemical Strategies for the
Presentation and Delivery of Growth Factors." JournalofMaterials Chemistry B.
doi: 10. 1039/c3tb20853b.
Chen, Irwin, Brent M. Dorr, and David R. Liu. 2011. "A General Strategy for the Evolution of BondForming Enzymes Using Yeast Display." Proceedings ofthe NationalAcademy of Sciences of the
United States ofAmerica 108 (28): 11399-404. doi:10.1073/pnas.1101046108.
Comisar, Wendy, Susan X. Hsiong, Hyun-Joon Kong, David J. Mooney, and Jennifer J. Linderman.
2006. "Multi-Scale Modeling to Predict Ligand Presentation within RGD Nanopatterned
Hydrogels." Biomaterials27 (10): 2322-29. doi: 10. 1016/j.biomaterials.2005.10.037.
Deng, Chao, Jintian Wu, Ru Cheng, Fenghua Meng, Harm Anton Klok, and Zhiyuan Zhong. 2014.
"Functional Polypeptide and Hybrid Materials: Precision Synthesis via A-Amino Acid NCarboxyanhydride Polymerization and Emerging Biomedical Applications." Progressin Polymer
Science 39 (2). Elsevier Ltd: 330-64. doi: 10.10 16/j.progpolymsci.2013.10.008.
Ehrbar, Martin, Simone C. Rizzi, Ruslan Hlushchuk, Valentin Djonov, Andreas H. Zisch, Jeffrey
Hubbell, Franz E. Weber, and Matthias P. Lutolf. 2007. "Enzymatic Formation of Modular Cell-
112
Instructive Fibrin Analogs for Tissue Engineering." Biomaterials 28 (26): 3856-66.
doi: 10.101 6/j.biomaterials.2007.03.027.
Ekerdt, Barbara L., Rachel Segalman, and David V. Schaffer. 2013. "Spatial Organization of CellAdhesive Ligands for Advanced Cell Culture." Biotechnology Journal 8 (12): 1411-23.
doi: 10.1 002/biot.201300302.
Elduque, Xavier, Albert Sanchez, Kapil Sharma, Enrique Pedroso, and Anna Grandas. 2013. "Protected
Maleimide Building Blocks for the Decoration of Peptides, Peptoids, and Peptide Nucleic Acids."
Bioconjugate Chemistry 24 (5): 832-39. doi:10.1021/bc4000614.
Engler, Amanda C., Hyung-il Lee, and Paula T. Hammond. 2009. "Highly Efficient 'Grafting onto' a
Polypeptide Backbone Using Click Chemistry." Angewandte Chemie (InternationalEd. in English)
48 (49): 9334-38. doi:10.1002/anie.200904070.
Engler, Amanda C., Daniel K. Bonner, Hilda G. Buss, Eva Y. Cheung, and Paula T. Hammond. 2011.
"The Synthetic Tuning of Clickable pH Responsive Cationic Polypeptides and Block
Copolypeptides." Soft Matter 7 (12): 5627. doi: 10.1039/cl sm05064h.
Englund, Ethan, Deyun Wang, Hidetsugu Fujigaki, Hiroyasu Sakai, Christopher M Micklitsch, Rodolfo
Ghirlando, Gema Martin-Manso, et al. 2012. "Programmable Multivalent Display of Receptor
Ligands Using Peptide Nucleic Acid Nanoscaffolds." Nature Communications3 (January). Nature
Publishing Group: 614. doi: 10.1038/ncomms1629.
Feng, Chun, Yongjun Li, Dong Yang, Jianhua Hu, Xiaohuan Zhang, and Xiaoyu Huang. 2011. "WellDefined Graft Copolymers: From Controlled Synthesis to Multipurpose Applications." Chemical
Society Reviews 40 (3): 1282-95. doi: 10.1039/b921 358a.
Gobin, Andrea S., and Jennifer L. West. 2003. "Effects of Epidermal Growth Factor on Fibroblast
Migration through Biomimetic Hydrogels." Cell, 1781-85.
Hendrickson, Grant R., and L. Andrew Lyon. 2009. "Bioresponsive Hydrogels for Sensing Applications."
Soft Matter 5 (1): 29. doi:10.1039/b811620b.
Huang, Jinghuan, Stefan V. Grater, Francesca Corbellini, Sabine Rinck, Eva Bock, Ralf Kemkemer,
Horst Kessler, Jiandong Ding, and Joachim P. Spatz. "Impact of order and disorder in RGD
nanopatterns on cell adhesion." Nano letters 9, no. 3 (2009): 1111-1116. doi:10.1021/nl803548b.
Jay, Steven M., Elma Kurtagic, Luis M. Alvarez, Seymour de Picciotto, Edgar Sanchez, Jessica F.
Hawkins, Robin N. Prince, et al. 2011. "Engineered Bivalent Ligands to Bias ErbB ReceptorMediated Signaling and Phenotypes." The JournalofBiological Chemistry 286 (31): 27729-40.
doi:10.1074/jbc.M 111.221093.
Kiessling, Laura L., Jason E. Gestwicki, and Laura E. Strong. 2006. "Synthetic Multivalent Ligands as
Probes of Signal Transduction." Angewandte Chemie (InternationalEd. in English) 45 (15): 234868. doi: 10.1002/anie.200502794.
Koehler, Kenneth C., Daniel L. Alge, Kristi S. Anseth, and Christopher N. Bowman. 2013. "A DielsAlder Modulated Approach to Control and Sustain the Release of Dexamethasone and Induce
Osteogenic Differentiation of Human Mesenchymal Stem Cells." Biomaterials 34 (16). Elsevier
Ltd: 4150-58. doi:10.1016/j.biomaterials.2013.02.020.
Krueger, Andrew T., Carsten Kroll, Edgar Sanchez, Linda G. Griffith, and Barbara Imperiali. 2014.
"Tailoring Chimeric Ligands for Studying and Biasing ErbB Receptor Family Interactions."
Angewandte Chemie (InternationalEd. in English) 53 (10): 2662-66. doi:10.1002/anie.201307869.
Kuhl, Philip R., and Linda G. Griffith-Cima. 1996. "Tethered Epidermal Growth Factor as a Paradigm for
Growth Factor-induced Stimulation from the Solid Phase." Nature Medicine 2: 1022-27.
doi: 10. 103 8/nm0996-1022.
Lam, Jonathan, and Tatiana Segura. 2013. "The Modulation of MSC Integrin Expression by RGD
Presentation." Biomaterials34 (16). Elsevier Ltd: 3938-47. doi:10.1016/j.biomaterials.2013.01.091.
Liang, Yingkai, and Kristi L. Kiick. 2013. "Heparin-Functionalized Polymeric Biomaterials in Tissue
Engineering and Drug Delivery Applications." Acta Biomaterialia,August. Acta Materialia Inc.
doi:10.1016/j.actbio.2013.07.03 1.
113
Lu, Hua, Jing Wang, Ziyuan Song, Lichen Yin, Yanfeng Zhang, Haoyu Tang, Chunlai Tu, Yao Lin, and
Jianjun Cheng. 2014. "Recent Advances in Amino Acid N-Carboxyanhydrides and Synthetic
Polypeptides: Chemistry, Self-Assembly and Biological Applications." Chemical Communications
(Cambridge, England) 50 (2): 139-55. doi:10.1039/c3cc46317f.
Martinez-Veracoechea, Francisco J., and Mirjam E. Leunissen. 2013. "The Entropic Impact of Tethering,
Multivalency and Dynamic Recruitment in Systems with Specific Binding Groups." Soft Matter 9
(12): 3213. doi:10.1039/c3sm27766f.
Mehta, Geeta, Courtney M. Williams, Luis Alvarez, Martha Lesniewski, Roger D. Kamm, and Linda G.
Griffith. 2010. "Synergistic Effects of Tethered Growth Factors and Adhesion Ligands on DNA
Synthesis and Function of Primary Hepatocytes Cultured on Soft Synthetic Hydrogels."
Biomaterials31 (17). Elsevier Ltd: 4657-71. doi: 10. 1016/j.biomaterials.2010.01.138.
Morimatsu, Masatoshi, Armen H. Mekhdjian, Arjun S. Adhikari, and Alexander R. Dunn. 2013.
"Molecular Tension Sensors Report Forces Generated by Single Integrin Molecules in Living
Cells." Nano Letters 13 (9): 3985-89. doi:10.1021/n4005145.
Mummert, M. E. 2000. "Development of a Peptide Inhibitor of Hyaluronan-Mediated Leukocyte
Trafficking." Journalof ExperimentalMedicine 192 (6): 769-80. doi: 10. 1084/jem. 192.6.769.
Patterson, David M., Lidia Nazarova, and Jennifer Prescher. 2014. "Finding the Right (bioorthogonal)
Chemistry." ACS Chemical Biology 9: 592-605. doi:10.1021/cb400828a.
Platt, Manu 0., Arian J. Roman, Alan Wells, Douglas Lauffenburger, and Linda G. Griffith. 2009.
"Sustained Epidermal Growth Factor Receptor Levels and Activation by Tethered Ligand Binding
Enhances Osteogenic Differentiation of Multi-Potent Marrow Stromal Cells." Journal of Cellular
Physiology 221 (2): 306-17. doi: 10. 1002/jcp.21854.
Rhodes, Allison J., and Timothy J. Deming. 2013. "Soluble, Clickable Polypeptides from AzideContaining N -Carboxyanhydride Monomers." A CS Macro Letters 2 (5): 351-54.
doi: 10.1021/mz4001089.
Sdnchez, Albert, Enrique Pedroso, and Anna Grandas. 2011. "Maleimide-Dimethylfuran Exo Adducts:
Effective Maleimide Protection in the Synthesis of Oligonucleotide Conjugates." OrganicLetters 13
(16): 4364-67. doi:10.1021/o1201690b.
Seal, Brandon L., and Alyssa Panitch. 2003. "Physical Polymer Matrices Based on Affinity Interactions
between Peptides and Polysaccharides." Biomacromolecules 4 (6): 1572-82.
doi:10.1021/bm0342032.
2006a. "Physical Matrices Stabilized by Enzymatically Sensitive Covalent Crosslinks." Acta
Biomaterialia2 (3): 241-51. doi: 10.1016/j.actbio.2005.12.008.
. 2006b. "Viscoelastic Behavior of Environmentally Sensitive Biomimetic Polymer Matrices."
Macromolecules 39 (6): 2268-74. doi: 10.102 1/ma0524528.
Sharma, Shaili, Aeju Lee, Kuiwon Choi, Kwangmeyung Kim, Inchan Youn, Stephen B Trippel, and
Alyssa Panitch. 2013. "Biomimetic Aggrecan Reduces Cartilage Extracellular Matrix from
Degradation and Lowers Catabolic Activity in Ex Vivo and in Vivo Models." Macromolecular
Bioscience 13 (9): 1228-37. doi:10.1002/mabi.201300112.
Shekaran, Asha, and Andrds J. Garcia. 2011. "Extracellular Matrix-Mimetic Adhesive Biomaterials for
Bone Repair." Journalof Biomedical MaterialsResearch. PartA 96 (1): 261-72.
doi: 10.1002/jbm.a.32979.
Shroff, Hari, Catherine G. Galbraith, James Galbraith, and Eric Betzig. 2008. "Live-Cell Photoactivated
Localization Microscopy of Nanoscale Adhesion Dynamics." Nature Methods 5 (5): 417-23.
doi: 10.1038/nmeth. 1202.
Shroff, Hari, Catherine G. Galbraith, James A. Galbraith, Helen White, Jennifer Gillette, Scott Olenych,
Michael W. Davidson, and Eric Betzig. "Dual-color superresolution imaging of genetically
expressed probes within individual adhesion complexes." Proceedingsofthe NationalAcademy of
Sciences 104, no. 51 (2007): 20308-20313. doi: 10.1073/pnas.0710517105.
114
Sijbrandij, Tjitske, Nivedita Cukkemane, Kamran Nazmi, Enno C. I. Veerman, and Floris J. Bikker. 2013.
"Sortase A as a Tool to Functionalize Surfaces." Bioconjugate Chemistry 24 (5): 828-31.
doi: 10. 1021/bc4000146.
Tang, Wen, and Matthew L. Becker. 2014. "'Click' Reactions: A Versatile Toolbox for the Synthesis of
Peptide-Conjugates."Chemical Society Reviews. doi:10.1039/c4csOO139g.
Teixeira, Liliana S. Moreira, Jan Feijen, Clemens van Blitterswijk, Pieter J. Dijkstra, and Marcel
Karperien. 2012. "Enzyme-Catalyzed Crosslinkable Hydrogels: Emerging Strategies for Tissue
Engineering." Biomaterials33 (5). Elsevier Ltd: 1281-90. doi: 10.101 6/j.biomaterials.2011.10.067.
Toepke, Michael W., Nicholas Impellitteri, Sheeny K. Lan Levengood, Derek S. Boeldt, Ian M. Bird, and
William L. Murphy. 2012. "Regulating Specific Growth Factor Signaling Using Immobilized
Branched Ligands." Advanced HealthcareMaterials 1 (4): 457-60. doi: 10.1002/adhm.201200077.
Trappmann, Britta, and Christopher S. Chen. 2013. "How Cells Sense Extracellular Matrix Stiffness: A
Material's Perspective." Current Opinion in Biotechnology 24 (5). Elsevier Ltd: 948-53.
doi: 10. 1016/j.copbio.2013.03.020.
Trappmann, Britta, Julien E. Gautrot, John T. Connelly, Daniel G. T. Strange, Yuan Li, Michelle L.
Oyen, Martien Cohen Stuart, et al. 2012. "Extracellular-Matrix Tethering Regulates Stem-Cell
Fate." Nature Materials 11 (7). Nature Publishing Group: 642-49. doi:10.1038/nmat3339.
Vogel, Viola, and Michael Sheetz. 2006. "Local Force and Geometry Sensing Regulate Cell Functions."
Nature Reviews. Molecular Cell Biology 7 (4): 265-75. doi: 10.1038/nrm 1890.
Wen, Qi, and Paul Janmey. 2013. "Effects of Non-Linearity on Cell-ECM Interactions." Experimental
Cell Research 319 (16). Elsevier: 2481-89. doi: 10.1016/j.yexcr.2013.05.017.
Wheeldon, Ian, Arash Farhadi, Alexander G. Bick, Esmaiel Jabbari, and Ali Khademhosseini. 2011.
"Nanoscale Tissue Engineering: Spatial Control over Cell-Materials Interactions." Nanotechnology
22 (21): 212001. doi:10.1088/0957-4484/22/21/212001.
Zegarra, Fabio C, Gian N. Peralta, Alberto M. Coronado, and Yi Qin Gao. 2009. "Free Energies and
Forces in Helix-Coil Transition of Homopolypeptides under Stretching." Physical Chemistry
Chemical Physics: PCCP 11: 4019-24. doi:10.1039/b820021a.
Zhu, Junmin. 2010. "Bioactive Modification of Poly(ethylene Glycol) Hydrogels for Tissue Engineering."
Biomaterials31 (17). Elsevier Ltd: 4639-56. doi: 10.101 6/j.biomaterials.2010.02.044.
115