Effect of Molecular Weight and Functionality on Acrylated Poly

Jul 25, 2018 - File failed to load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js .... micelles for intracellular delivery of dox...
0 downloads 0 Views 1MB Size
Subscriber access provided by Kaohsiung Medical University

Article

Effect of Molecular Weight and Functionality on Acrylated Poly(caprolactone) for Stereolithography and Biomedical Applications Brian J. Green, Kristan S. Worthington, Jessica R. Thompson, Spencer J. Bunn, Mary Rethwisch, Emily E. Kaalberg, Chunhua Jiao, Luke A. Wiley, Robert F. Mullins, Edwin M. Stone, Elliott H. Sohn, Budd A. Tucker, and C. Allan Guymon Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.8b00784 • Publication Date (Web): 25 Jul 2018 Downloaded from http://pubs.acs.org on July 27, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Effect of Molecular Weight and Functionality on Acrylated Poly(caprolactone) for Stereolithography and Biomedical Applications Brian J. Green,1 Kristan S. Worthington,2,3 Jessica R. Thompson,2,3 Spencer J. Bunn,1 Mary Rethwisch,1 Emily E. Kaalberg,2 Chunhua Jiao,2 Luke A. Wiley,2 Robert F. Mullins,2 Edwin M. Stone,2 Elliott H. Sohn ,2 Budd A. Tucker,2 C. Allan Guymon*1 1

Department of Chemical and Biochemical Engineering, The University of Iowa, 4133 Seamans Center, Iowa City, IA 52242, USA 2

Institute of Vision Research, Department of Ophthalmology and Visual Science, Carver College of Medicine, The University of Iowa, 4111 Medical Education and Research Facility, Iowa City, IA 52242, USA 3

Department of Biomedical Engineering, The University of Iowa, 5602 Seamans Center, Iowa City, IA 52242, USA *Corresponding Author

1 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

Abstract: Degradable polymers are integral components in many biomedical polymer applications. The ability of these materials to decompose in situ has become a critical component for tissue engineering, allowing scaffolds to guide cell and tissue growth while facilitating gradual regeneration of native tissue. The objective of this work is to understand the role of prepolymer molecular weight and functionality of photocurable poly(caprolactone) (PCL) in determining reaction kinetics, mechanical properties, polymer degradation, biocompatibility, and suitability for stereolithography. PCL, a degradable polymer used in a number

of

biomedical applications,

was

functionalized

with acrylate

groups

to

enable

photopolymerization and 3D printing via stereolithography. PCL prepolymers with different molecular weight and functionality were studied to understand the role of molecular structure on reaction kinetics, mechanical properties and degradation rates. The mechanical properties of photocured PCL were dependent on cross-link density and directly related to the molecular weight and functionality of the prepolymers. High molecular weight, low functionality PCLDA prepolymers exhibited lower modulus and higher strain at break while low molecular weight, high functionality PCLTA prepolymers exhibited lower strain at break and higher modulus. Additionally, degradation profiles of cross-linked PCL followed a similar trend, with low cross-link density leading to degradation times up to 2.5 times shorter than more highly cross-linked polymers. Furthermore, photopolymerized PCL showed biocompatibility both in vitro and in vivo, causing no observed detrimental effects on seeded murine induced pluripotent stem cells or when implanted into pig retinas. Finally, the ability to create 3-dimensional PCL structures is shown by fabrication of simple structures using digital light projection stereolithography. Low molecular weight, high functionality PCLTA prepolymers printed objects with feature sizes near the hardware resolution limit of 50 μm. This work lays the foundation for future work in fabricating micro-scale PCL structures for a wide range of tissue regeneration applications.

2 ACS Paragon Plus Environment

Page 3 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Keywords: Photopolymerization, Stereolithography, Poly(caprolactone), Tissue Scaffolds, Stem Cells, Degradable Polymers, Biocompatability

3 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction: Degradable polymers have become increasingly important in biomedical applications, enabled by their ability to degrade through biological means. The market for degradable polymer products is growing at a fast pace, fueled by demand for consumer goods such as packaging, cosmetics, and disposable utensils.1,2 Indeed, degradable polymers have allowed significant progress to occur in biomedical applications such as sutures, tissue scaffolds, drug delivery, and other biomedical devices.3–5 Inherent degradability is particularly important in tissue scaffolds, which are designed to be temporary structures that provide support and guidance to growing cells and tissues.6,7 Resorbable polymers typically degrade either by traditional chemical reactions, such as hydrolysis or photodegradation, or biological processes, such as enzymatic degradation. The method and time-scale of degradation rely on both polymer chemistry as well as the microenvironment of the polymer.4 Thus, a number of polymer characteristics influence degradation including the location and type of functional groups that comprise the polymer structure, polymer chain length, and cross-link density. Meanwhile, environmental factors including pH, temperature, water content, and presence of enzymes and microorganisms also play significant roles.1,8–11 Regardless of the decomposition mechanism, the degradation rate of these polymers is an important consideration across broad applications. For example, many degradable polymers currently used in the biomedical field degrade via hydrolysis of polyester bonds, resulting in the release of carboxylic acid containing byproducts. When such an implanted polymer degrades quickly (hours to weeks), the local tissue may not be able to effectively remove or neutralize the degradation products before they cause damage to sensitive biological processes. Such rapid degradation may cause a sharp decrease in the local pH that can adversely affect the surrounding tissue, limiting the efficacy of the tissue scaffold.12

4 ACS Paragon Plus Environment

Page 4 of 38

Page 5 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Poly(caprolactone) (PCL) is a degradable polyester with a slower degradation rate than many other degradable polymers used in bioengineering applications, such as poly(lactic acid) (PLA), poly(glycolic acid), and poly(lactic-co-glycolic acid) (PLGA). This rate leads to favorable tolerance in sensitive human tissues, such as the retina, due to lower local decreases in pH than would otherwise occur if using PLGA, for example.13 PCL is formed by a ring opening polymerization of -caprolactone in the presence of a catalyst such as stannous octoate and has been used in biomedical applications such as sutures, drug delivery devices, and tissue scaffolds.14 It is commonly used as a linear, high molecular weight polymer due to its favorable rheological and viscoelastic properties that allow it to be easily shaped and molded into materials with a broad range of properties.15 However, many biomedical applications would benefit from the ability to fabricate three-dimensional poly(caprolactone) structures rapidly at ambient conditions enabled by photopolymerization induced cross-linking. To allow photopolymerization, PCL has been modified with functional groups such as acrylates,16–19 methacrylates,20 or fumarates.21 For example, PCL with alcohol end groups was functionalized with acrylate groups using acryloyl chloride and either triethylamine16,17 or potassium carbonate18 as a proton scavenger. These studies produced PCL polymers of varying surface chemistry, stiffness, and crystallinity. Variation in crystallinity played a significant role in regulating cell growth of osteoblasts and Schwann cells18 and in observed degradation rate.16 Other work has synthesized vinyl-functionalized PCL by ringopening polymerization in the presence of 4-hydroxybutyl vinyl ether.22 Additionally, PCL-based macrophotoinitiators have been synthesized using an enzyme catalyst and used in acrylate polymerization.23 PCL has been used for biomedical applications as a degradable subunit of PEG-based photopolymerized block-copolymer hydrogels with tunable degradation properties.24,25 Futhermore, acrylated PCL and PCL macroinitiators have been copolymerized with hydroxyethyl methacrylate to create hydrogels for degradable tissue scaffolds.19 This study produced biocompatible hydrogels with appropriate

mechanical

properties

and

tunable

degradation

5 ACS Paragon Plus Environment

rates.19

Recent

progress

in

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 38

photopolymerization and stereolithography could enable the use of 3D photocured systems to achieve the spatial control necessary to utilize cross-linked PCL in custom tissue scaffolds or degradable implants. 3D printed scaffolds have been used for drug delivery devices,26 wound healing,27 and bone tissue regeneration,28 but the majority of these scaffolds are fabricated using extrusion based 3D printing techniques. Conversely, stereolithography uses photopolymerization to cure liquid resins to build objects layer by layer. The inherent spatial and temporal control of photopolymerization is ideal for 3D printing and creating custom structures. Features small enough to influence the cell and tissue response of a biomaterial may be achieved through stereolithography which would be much more difficult to create through extrusion-based techniques.3,29 However, relatively few degradable prepolymers have been examined as potential stereolithographic resins and little is known about their relevant final properties. Increased understanding of the influence of cross-linked networks on polymer degradation behavior may allow widespread use of stereolithography for creating degradable tissue scaffolds, drug delivery devices or consumer goods. The ability to create biocompatible, 3D scaffolds capable of controlled degradation is necessary to build on recent progress and advance the field of tissue engineering. In this work, we explore the impact of prepolymer molecular weight and functionality on the properties of cross-linked poly(caprolactone) and on the ability to fabricate biocompatible 3D structures. PCL with varying molecular weights and functionalities are functionalized with acrylate groups based on a previously

developed

method.16,18

These

acrylate

groups

allow

cross-linking

through

photopolymerization, enabling spatial and temporal control of the cross-linking reaction. Double bond conversion and reaction kinetics are measured to understand the polymerization reaction and to assess compatibility with stereolithography. The mechanical properties of cross-linked PCL are determined, and the degradation profile is established. Additionally, the biocompatibility of photopolymerized PCL is determined in vitro using murine induced pluripotent stem cells and in vivo in pig retinas. Finally, the

6 ACS Paragon Plus Environment

Page 7 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

capability to create high-resolution structured PCL is shown by fabricating objects with micro-scale structure using digital light projection stereolithography.

Experimental PCL Functionalization Polycaprolactone diol (average Mn of 530 and 1250 g/mol) and polycaprolactone triol (average Mn of 300 and 900 g/mol) (Sigma-Aldrich, St. Louis, MO) were dried at 110 °C for 12 hours and cooled under vacuum. Potassium carbonate (Fisher Scientific, Hampton, NH) was dried at 110 °C for 24 hours. Acryloyl chloride and dichloromethane (Sigma-Aldrich) were both used as received. Polycaprolactone diol and triol of different molecular weights using a modified procedure based on a method by Cai and Wang.18 First, approximately 10 grams of PCL were dissolved in 100 mL dichloromethane. A slurry was formed when potassium carbonate was added to the solution at a 1.5 molar ratio with respect to the PCL alcohol end groups. Acryloyl chloride was dissolved in dichloromethane (1:5 v:v) in a separate vessel, at an equimolar ratio with respect to the potassium carbonate. The reaction flask containing the slurry was purged with argon gas for 10 minutes before dropwise addition of the acryloyl chloride solution while stirring. The reaction was then allowed to proceed at room temperature for 24 hours under argon atmosphere. The mixture was then filtered to remove solids formed during the reaction. The resulting solution was rotary-evaporated, leaving the functionalized PCL as a waxy solid or viscous liquid, depending on the molecular weight. PCL acrylation was confirmed via 1H NMR. After vacuum purification, each PCL prepolymer was dissolved in CDCl3 (Sigma-Aldrich) and the solutions were analyzed using a Bruker Avance-300 probe with a field strength of 300 MHz. The appearance of NMR peaks at δ 5.8 (1H), 6.1 (1H), and 6.4 (1H) ppm, corresponding to the hydrogens on the acrylate group, were used to confirm the presence of acrylate

7 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

groups on the prepolymers. The degree of acrylation was determined by comparing the integrated area under the NMR peaks at δ 5.8 through 6.4 (acrylate hydrogens) to δ 3.7 (diethylene glycol linker) for diacrylates or δ 0.9 (methyl group on the 4th arm) for triacrylates (Figure S1). Formulation Acrylated PCL prepolymers were dissolved in dioxane (Sigma-Aldrich) to form 1:1 (wt:wt) PCL:dioxane solutions. A photoinitiator, 2-Benzyl-2-dimethylamino-1-(4-morpholinophenyl)-butanone-1 (commonly known as Irgacure 369 or I-369, BASF, Ludwigshafen, Germany) was added to the solution at 3 wt% to allow cross-linking upon exposure to UV light. The mixtures were then sonicated for 15 minutes until the prepolymer and photoinitiator were fully dissolved. These solutions were stored in the dark, away from any light sources, to prevent unwanted polymerization. The viscosity of all solutions was in the range of 20-30 cP at room temperature as measured by a Brookfield Model DV-I + Viscometer equipped with an SC4-18 spindle operating at a shear rate of 36 s-1. UV Polymerization Kinetics Double bond conversion was determined using a Thermo Electron Nexus 670 real-time Fourier transform infrared (RT-FTIR) spectrometer equipped with a horizontal transmission accessory. A small droplet of each sample was placed between two NaCl salt plates with 15 µm silica spacers to ensure consistent path length. The sample was illuminated with 2.0 mW/cm2 light from an EFOS Acticure 4000 UV lamp equipped with a medium pressure mercury arc lamp. Reduction of the acrylate peak height at 810 cm-1 was monitored to determine double bond conversion and compared to the carbonyl reference peak height at 1730 cm-1. Photopolymerization kinetics were examined using a Perkin Elmer Diamond differential scanning calorimeter modified with a medium pressure mercury arc lamp (photo-DSC). Formulations were prepared with the same composition as for RT-FTIR except a high boiling point solvent, ethylene glycol 8 ACS Paragon Plus Environment

Page 8 of 38

Page 9 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

diacetate (EGDA), was to prevent heat flow variations caused by dioxane evaporation during polymerization. Between 3 and 3.5 mg of the total formulation were placed into a DSC pan and polymerized with 2.0 mW/cm2 UV light from a medium pressure mercury arc lamp. Polymerization rates were determined using the evolved heat of polymerization per acrylate group according to Equation 1.

𝑅𝑝 =

𝑄 ∆𝐻 𝑛 𝑚 𝑀𝑊

(1)

Where Rp is the normalized rate of polymerization, Q is the heat flow as measured by the DSC, ΔH is the enthalpy of reaction for an acrylate functional group (86190.4 J/mol),30 n is the number of reactive groups, m is the mass of monomer in the sample, and MW is the molecular weight of the monomer. The resulting Rp is normalized per reactive group. Mechanical Behavior Dynamic mechanical analysis (DMA; Q800 DMA TA Instruments) was conducted to investigate the effect of monomer molecular weight and functionality on the resulting polymer mechanical properties. Samples were prepared by placing monomer mixtures between two glass slides with 200 µm spacers. The samples were then illuminated with UV light at 10 mW/cm2 for 10 minutes to induce polymerization. The polymerized films were then cut to create rectangular samples with dimensions approximately 5 mm x 20 mm x 0.2 mm. No steps were taken to remove the solvent from the polymerized samples. To measure the tensile properties, the films were held at 30 °C and a force rate of 1.5 N/min was applied until break. Degradation Cross-linked PCL films were prepared by placing prepolymer mixtures between two glass slides with 1 mm spacers. The monomers were then illuminated with 10 mW/cm2 UV light for 10 minutes. An 8 mm biopsy punch was used to create discs from the resulting polymer films. Initial masses were measured after

9 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

vacuum drying for 24 hours to remove solvent. Degradation rates were determined gravimetrically by placing these cross-linked PCL discs in either 1x phosphate buffered saline (PBS) or 1 M NaOH in an incubator held at 37 °C. NaOH was used to accelerate the hydrolysis of the ester bonds present in PCL. This method has previously been used to determine accelerated degradation profiles of linear PCL, the relative profiles of which were similar to those performed in a pH neutral system.31 At set intervals that covered a total time of roughly six weeks (PBS) or three weeks (NaOH), PCL discs were removed from the solvent, dried under vacuum for 24 hours, and the dry weight was recorded. These measurements were performed in triplicate. Dried discs were discarded after weighing. Biocompatibility In Vitro To test the biocompatibility of each formulation, murine induced pluripotent stem cells (MiPSCs) were cultured on 5 mm diameter UV-polymerized discs, as described in a previous study.32 Briefly, PCL discs were submerged in 100% EtOH overnight and rinsed with three PBS washes the following day. The discs were then placed in 40 m nylon mesh inserts (three identical samples per insert, Falcon cell strainer, Corning) and placed in a 6-well plate. The cross-linked PCL samples were coated in Matrigel (3 mL per well) to enable cell attachment. Cells were seeded onto each of the samples at a concentration of 2 × 104 cells in 10 μL of pluripotency media and incubated at 37 °C for 30 min. After initial cell attachment, 4mL of fresh media was gently added to each well. For the rest of the week the pluripotency media was changed daily. Cell proliferation and viability were analyzed after 1 week of culture using immunocytochemistry. Samples were fixed in 4% paraformaldehyde, rinsed three times with PBS, and incubated at 4 °C overnight in primary antibody solution consisting of rabbit anti-phospho-Histone H3 (Cat#: 9701; 1:500; Cell Signaling Danvers, MA) and AlexaFluor 488-conjugated Phalloidin (#A12379; 1:1000; Life Technologies) in blocking 10 ACS Paragon Plus Environment

Page 10 of 38

Page 11 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

buffer (5% normal goat serum, 0.5% Triton-X, and 3% bovine serum albumin (BSA) in PBS). The next day, samples were rinsed three times with PBS and incubated for 2 h at room temperature in secondary antibody solution containing fluorescently conjugated Alexa Fluor secondary antibodies (1:500; Thermo Fisher Scientific; Waltham, MA) in blocking buffer. Samples were rinsed three times with PBS again, counterstained with 4’,6-diamidino-2-phenylindole (DAPI) to fluorescently label cells, and mounted with a coverslip. Immunofluorescent images were collected using a Leica DM 2500 SPE confocal microscope (Leica Microsystems, Wetzlar, Germany). Seven or more fields were captured for each sample, and the mitotic index (MI) was calculated as follows:

𝑀𝐼 =

𝑃𝐻3+ 𝐷𝐴𝑃𝐼+

× 100

(2)

where PH3+ and DAPI+ are the number of cells undergoing mitosis and the total number of cells, respectively. Counting was performed using thresholding, watershed and the Analyze Particles function in ImageJ. For each independent image of mouse iPSCS on cross-linked PCL (in vitro data), the number of cells per field and mitotic index were determined using ImageJ. Neither dataset passed the d’Agostino and Pearson normality test, so non-parametric analysis was performed thereafter. In particular, a Kruskal-Wallis test was performed for each dataset, followed by Dunn’s multiple comparisons tests. Both were performed at a confidence interval of 95%. In Vivo Prior to transplantation, PCL samples were sterilized by submersion in 100% ethanol for 24 hours, followed by three 30-minute rinses with sterile 1xHBSS. All animal procedures were performed with permission of the University of Iowa IACUC and complied with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Polymer implantation was performed on 5 four-month-old

11 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Yucatan mini pigs, each of which received a sub-retinal 1.25 x 4 x 0.1 mm (W x L x T) PCLDA 1250 film. Surgery was performed on both eyes in each pig, but the polymer film was placed in only one of the two eyes, allowing the second eye to act as a surgical control. Specifically, a 23-gauge three-port pars plana vitrectomy was used for removal of the central vitreous and detachment of the posterior hyaloid. A subretinal bleb near the optic nerve and visual streak was raised with balanced salt solution on a 41-gauge needle as previously described.33 An approximately 1.5 mm retinotomy was made within the bleb using endocautery and vertical scissors, and the scaffold was placed through a fourth sclerotomy into the subretinal space with custom-built polymer forceps. Fluid-air exchange was performed. The cannulas were removed and sclerotomies sutured before 5% povidone-iodine was used to rinse the eye. The pigs received no systemic post-operative immune suppressive agents. Animals were sacrificed by barbiturate overdose at one-month post-surgery, at which time indirect ophthalmoscopy and spectral-domain OCT (SD-OCT) were performed (Bioptigen Envisu R2200, Bioptogen, Inc., Morrisville, NC). Immediately postophthalmoscopic evaluation, eyes were enucleated and fixed in 4% paraformaldehyde overnight, then processed for morphologic analysis. Retinal sections were assessed using paraffin sections stained by haemotoxylin and eosin (H&E) and immunohistochemistry then imaged using light and confocal microscopy, respectively. For immunohistochemistry, tissues were treated with anti-RPE65 and antiCD31 antibodies to detect retinal pigment epithelium and blood vessels, respectively.

Digital Light Projection Stereolithography 3D printed objects were created using an Autodesk Ember 3D printer equipped with a DLP 0.45" WXGA digital micromirror device and a 405 nm LED light source that outputs approximately 25 mW/cm2 at the resin interface. STL files were loaded into Autodesk Netfabb software and sliced into 50 µm thick layers. Formulations were prepared as described earlier except the I-369 photoinitiator was replaced by

12 ACS Paragon Plus Environment

Page 12 of 38

Page 13 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide) (TPO, Ciba Specialty Chemicals, Basel, Switzerland) due to TPO more readily absorbing 405 nm light. During polymerization, the first layer was illuminated for 8 seconds to ensure adhesion to the build head and subsequent layers were illuminated for 2.2 seconds. To quantify resolution, a 3D object was designed using square boxes that were 200 µm tall and widths of 500 µm, 400 µm, 300 µm, 200 µm, 150 µm, 125 µm, 100 µm, 75 µm, and 50 µm. Scanning electron micrographs were obtained using a Hitachi S-3400N (Hitachi, Krefeld, Germany) scanning electron microscope with a 5 kV accelerating voltage. Prior to being placed in the sample chamber, test objects were vacuum dried for 12 hours to remove the solvent and any other volatile components. ImageJ software was used to measure height and width of printed features. The measured dimensions were then divided by the designed dimensions to determine height and width ratios. Circularity was calculated according to Equation 3: 𝐶𝑖𝑟𝑐𝑢𝑙𝑎𝑟𝑖𝑡𝑦 = 4𝜋

𝐴 𝑃2

(3)

where A is the measured area and P is the measured perimeter. Based on equation 3, the circularity of a perfect circle is 1 and a square is 0.785.

Results and Discussion In this study, we aim to functionalize poly(caprolactone) to enable photopolymerization and determine the effect of monomer molecular weight and prepolymer functionality on the resulting polymer properties, stem cell biocompatibility, and stereolithography. We report successful acrylation of PCL diols and triols, describe the kinetics of their photopolymerization, and characterize the properties of the resulting polymers. The double bond conversion and reaction kinetics are measured to screen suitability for 3D printing, and the resulting mechanical properties of polymerized films are characterized. Degradation profiles of cross-linked PCL acrylates are determined in PBS and under accelerated conditions 13 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

in 1M NaOH. Additionally, biocompatibility of cross-linked PCL films is determined both in vitro using murine induced pluripotent stem cells and in vivo through implantation in pig retinas. Finally, structures are prepared using digital light projection stereolithography to illustrate the capabilities of using PCL prepolymers in 3D printing as degradable tissue scaffolds. PCL Functionalization Poly(caprolactone) is commonly used for tissue engineering applications because it is degradable, biocompatible, and has robust mechanical and viscoelastic properties. We successfully functionalized PCL diols and triols of several molecular weights with acrylate groups (referred to hereafter as PCL diacrylate and triacrylate, or PCLDA and PCLTA, respectively) using previous methods.18 These acrylate groups enable photopolymerization which provides spatial and temporal control of the radical cross-linking reactions. This synthesis was accomplished by reacting PCL molecules with acryloyl chloride in the presence of potassium carbonate under anhydrous and inert conditions (Figure 1A). Successful acrylation is confirmed by the appearance of NMR peaks at δ 5.8 (1H), 6.1 (1H), and 6.4 (1H) ppm, corresponding to the hydrogens on the acrylate group. An example spectrum is shown in Figure 1B with full spectra with appropriate integration shown in Figure S2. The acrylate functionality of all prepolymers ranged from approximately 75% to 85% (Table S1).

14 ACS Paragon Plus Environment

Page 14 of 38

Page 15 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Figure 1. A) Reaction schematic for the functionalization of PCL Diol with acrylate functional groups using acryloyl chloride and potassium carbonate. B) Representative NMR spectrum indicating the successful incorporation of acrylate functional groups.

Polymerization Kinetics Many factors such as prepolymer functionality, final acrylate conversion, polymerization kinetics, and prepolymer molecular weight can affect the ability to use stereolithography (SL) as a method of printing 3-dimensional structures. For example, prepolymers with higher functionalities are expected to form a more highly cross-linked network upon polymerization, leading to better structure fidelity and accuracy of printed parts. Furthermore, final acrylate conversion corresponds with cross-link density, where higher conversion indicates that more reactive groups are converted into functional cross-links. Polymerization kinetics, in turn, dictate how quickly these cross-links are formed, influencing required printing time. A fourth factor that affects the SL process is molecular weight between cross-links, which is directly tied to prepolymer molecular weight, functionality, and conversion. Prepolymers with higher molecular weight between cross-links form looser networks when fully cross-linked, resulting in lower resolution and poorer structure retention than their low molecular weight counterparts. Essentially, these factors indicate that lower molecular weight and higher functionality should result in high functional group density, which

15 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

would lead to high conversion, fast reaction rates, high cross-link density and likely more effective printing. Figure 2 shows the real-time acrylate conversion for each of the functionalized PCL monomers. Once illuminated with UV light in the presence of a photoinitiator, acrylated PCL prepolymers react quickly to form cross-linked networks. Both prepolymer molecular weight and the number of functional groups per molecule play a large role in the reaction and resulting conversion profiles. All PCL prepolymers reach high conversion: nearly 100% for PCLTA 300 and around 90% for the remaining prepolymers. Both PCLTA prepolymers appear to have a similar initial reaction rate that is higher than the two PCLDA prepolymers. The effect of molecular weight becomes more evident in the pair of PCLDA prepolymers: the PCLDA 1250 prepolymer has a much lower polymerization rate than the PCLDA 530 prepolymer. In addition to having the lowest initial polymerization rate, PCLDA 1250 takes roughly twice as long to reach final conversion than the other prepolymers. Generally, decreasing molecular weight and increasing functionality increase the rate of reaction and, to a lesser degree, the final conversion. These results confirm the welldocumented role that molecular weight and functionality play in acrylate conversion and polymerization kinetics.

16 ACS Paragon Plus Environment

Page 16 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

F r a c t io n a l C o n v e r s io n

Page 17 of 38

1 .0

PCLTA PCLTA PCLDA PCLDA

0 .5

300 900 530 1250

0 .0 0 .0

0 .5

1 .0

1 .5

T im e (m in )

Figure 2. Prepolymer conversion profile for PCLTA 300 ( ), PCLTA 900 ( ), PCLDA 530 ( ), and PCLDA 1250 ( ) measured using RT-FTIR. The samples were prepared with a 1:1 (w:w) PCL prepolymer:dioxane ratio and illuminated with 2.0 mW/cm2 full-spectrum UV light. The disappearance of the acrylate peak at 810 cm-1 was monitored in real-time during the reaction to determine double bond conversion.

The effect of molecular weight on the kinetic trends observed in Figure 2 can be further explored using photo-DSC, which tracks the heat of reaction and allows calculation of the resulting reaction rates. This experiment (Figure 3) confirms the previous results, showing that the lower molecular weight PCL prepolymers reach a maximum polymerization rate that is higher than the corresponding higher molecular weight PCL prepolymer. Decreasing the molecular weight of acrylated PCL increases the maximum polymerization rate by approximately 50% in the case of PCLDA 1250 compared to PCLDA 530 and by roughly 33% for PCLTA 900 versus PCLTA 300. In these systems, the major factors that influence the rate of polymerization are the monomer molecular weight and the monomer functionality. The higher functionality monomers show faster polymerization rates than those with lower functionalities. When comparing monomers of the same functionality, higher polymerization rates are observed with lower molecular weights, presumably due to the proximity of functional groups during polymerization. This proximity leads to a greater local concentration of reactive groups and thereby allows for growing polymer

17 ACS Paragon Plus Environment

Biomacromolecules

chains to easily find and react with unreacted double bonds, limiting the need for monomer diffusion for reaction.

0 .0 5

PCLTA PCLTA PCLDA PCLDA

0 .0 5

N o r m a liz e d R p ( 1 / s )

N o r m a liz e d R p ( 1 / s )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 38

0 .0 4

0 .0 3

0 .0 4

0 .0 3

300 900 530 1250

0 .0 2

0 .0 1

0 .0 2 0 .0 0 0

5

10

0 .0 1

15

20

25

30

T im e (s )

0 .0 0 0

50

100

150

T im e (s )

Figure 3. Normalized prepolymer polymerization rate for PCLTA 300 ( ), PCLTA 900 ( ), PCLDA 530 ( ), and PCLDA 1250 ( ) measured using Photo-DSC. Samples were prepared with a 1:1 PCL prepolymer:EGDA ratio and illuminated with 2.0 mW/cm2 full-spectrum UV light. The subsequent heat of reaction was measured to determine the rate of polymerization. The inset shows the first 30 seconds of the reaction to further demonstrate the difference in the early reaction rates.

Based on the fast reaction kinetics observed, it is expected that all four prepolymers are suitable for 3D printing using DLP stereolithography. Each prepolymer reaches its maximum polymerization rate within five seconds at light intensities approximately one-tenth of those commonly used in stereolithography. Objects fabricated using PCLTA 300 prepolymer should allow printing of the smallest features with the highest resolution and structure fidelity. Due to its faster reaction kinetics and higher final conversion, the ability of radicals to diffuse and react outside of the illuminated area would be more limited compared to the other prepolymers. Similarly, objects fabricated using PCLDA 1250 prepolymer would likely exhibit the lowest resolution and fidelity. In this case, the relatively slow reaction rate is expected to allow more time for propagating chains to diffuse out of the illuminated area. Additionally, the PCLDA 1250 prepolymer is the slowest to reach its final conversion, requiring more illumination time to reach appropriate cross-link density which could lead to difficulty fabricating small structures. 18 ACS Paragon Plus Environment

Page 19 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Mechanical Behavior Tuning and understanding the mechanical properties of a polymer are important in tissue engineering in order to avoid damaging delicate local tissue and to enable surgical manipulation of the material. Additionally, the modulus of a material is capable of influencing the behavior of implanted cells, including differentiation, proliferation, and orientation.34–37 In this work, the tensile stress/strain behavior of crosslinked PCL is examined using dynamic mechanical analysis to more fully understand the effect of molecular weight and functionality on the mechanical properties for this system. As with reaction kinetics, molecular weight and functionality play large roles in the mechanical properties of cross-linked PCL. This trend can be seen in Figure 4 where low molecular weight and high functionality lead to relatively high modulus and low strain at break. When comparing polymers made from prepolymers with the same functionality, polymers formed from lower molecular weight prepolymers exhibit higher modulus and stress at break, accompanied by lower strain at break. For example, PCLTA 300 polymer exhibits a 75% higher modulus, 22% greater stress at break, and 30% lower strain at break compared to PCLTA 900 (Table 1). Similar trends can be observed for polymers created from the pair of lower functionality PCLDA prepolymers; PCLDA 530 polymer shows a higher modulus, greater stress at break, and lower strain at break compared to PCLDA 1250. Higher modulus and lower strain at break is observed for higher functionality prepolymers, as expected. Specifically, when comparing the lower molecular weight prepolymers with different functionalities, PCLTA 300 has a higher modulus and lower strain at break than PCLDA 530. The same general trend can be observed for the pair of higher molecular weight prepolymers, where PCLTA 900 has a larger modulus and lower strain at break than PCLDA 1250. As expected, network structure formed upon cross-linking plays an integral role in understanding the observed results. In general, when molecular weight between cross-links decreases, either by using lower molecular weight or higher functionality prepolymers, cross-link density increases. The close-knit

19 ACS Paragon Plus Environment

Biomacromolecules

networks limit the ability of the polymer chains to stretch and extend in response to the applied stress, resulting in lower strain at break. Additionally, higher cross-linked networks offer more reinforcement to the applied stress, effectively distributing the stress through a larger number of polymer chains resulting in a stiffer, higher modulus material. Based on these results, the PCLDA 1250 polymer is expected to be the most amenable to surgical manipulation as it has a significantly higher strain at break than the other polymers, allowing for much more flexibility during the implantation process. Additionally, the lower modulus of the PCLDA 1250 polymer may make it more suitable as a scaffold for soft tissue, such as the retina, than the other tested polymers. While the kinetic studies indicate that the PCLTA 300 polymer is likely the most suitable prepolymer for achieving high-resolution and high-fidelity 3D printing, which will be required to generate the intricate 3D scaffolds needed to increase cellular packing density and promote photoreceptor cell alignment post-transplantation.

PCLTA PCLTA PCLDA PCLDA

1 .0

0 .8

S tre s s (M P a )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 38

300 900 530 1250

0 .6

0 .4

0 .2

0 .0 0

10

20

30

40

S tr a in (% )

Figure 4- Representative stress-strain plots for cross-linked PCLTA 300 ( ), PCLTA 900 ( ), PCLDA 530 ( ), and PCLDA 1250 ( ) polymers. Rectangular samples with dimensions approximately 5 mm x 20 mm x 0.2 mm (W x L x T) were polymerized using 10 mW/cm2 full-spectrum UV light and tested at room temperature in tensile mode with a force rate of 1.5 N/min until break.

20 ACS Paragon Plus Environment

Page 21 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Table 1- Stress-strain results for tensile testing of cross-linked PCL polymers. Experiments were performed in triplicate and results reported as mean  standard deviation. Prepolymer Tensile Modulus (MPa) Strain at Break (%) Stress at Break (MPa) Toughness (kJ/m3) 13  1 0.70  0.07 53  5 PCLTA 300 6.9  0.7 19  1 0.58  0.05 61  6 PCLTA 900 4.0  0.5 19  2 0.78  0.09 90  10 PCLDA 530 4.9  0.6 39  4 0.58  0.04 150  20 PCLDA 1250 3.3  0.2

Degradation Polymer degradation rate is a critical performance parameter in tissue engineering materials. For example, in an ideal case, the polymer scaffold is present to guide initial delivery and proliferation of implanted cells, then subsequently degrades to allow integration to host tissue and avoid harmful effects of permanent fixtures.38 For cross-linked PCL systems, network variations and cross-link density are expected to play large roles in observed degradation rates, similar to their effect on mechanical properties. To determine the effect of molecular weight and functionality on PCL degradation, 8 mm diameter polymer discs were placed in both 1x PBS and 1 M NaOH at 37 °C. Since PCL can take several years to dissolve completely at neutral pH, NaOH is used to accelerate the hydrolysis reaction, providing short term results where the degradation profiles can be compared to those in phosphate buffered saline.31 Figure 5 shows the degradation rate as a function of time for cross-linked PCL prepolymers in PBS (A) and NaOH (B). It is expected that materials fabricated using prepolymers with higher molecular weights and lower functionality will degrade faster because the cross-link density is decreased compared to the lower molecular weight, higher functionality prepolymers. This trend is illustrated experimentally in Figure 5, where PCLTA 300 takes nearly three times longer to completely degrade than PCLDA 1250. When comparing prepolymers with the same functionality, lower molecular weight prepolymers take slightly longer to degrade than higher molecular weight counterparts. However, prepolymer functionality plays a much larger role in determining degradation time as the higher functionality PCLTA prepolymers degrade two to three times more slowly than the PCLDA prepolymers. This behavior indicates that higher cross21 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

link density attributed to increased prepolymer functionality plays a larger role in determining the degradation time than the increase in cross-link density that occurs due to decreasing prepolymer molecular weight. This trend of higher molecular weight prepolymers leading to decreased degradation time in cross-linked systems is the opposite of that observed in non-cross-linked systems, where the degradation time typically increases as polymer molecular weight increases. When cross-links are formed, acrylate groups react to form saturated, non-degradable chains that are quite hydrophobic. Any increase in cross-link density further prohibits water swelling and limits the ability of water to reach the degradable PCL sections and induce hydrolysis. Additionally, a large difference in degradation time is observed between difunctional and trifunctional prepolymers due to further changes in network connectivity. The increased functionality of the PCLTA prepolymers leads to a more highly cross-linked network with many connection points that subsequently require more degradation reactions for polymer chains to decouple from the polymer network. The combination of increased cross-link density and increased network connectivity leads to polymerized PCLTA 300 degrading nearly three times slower than polymerized PCLDA 1250 under accelerated conditions. Regardless of prepolymer selection, in vivo degradation od cross-linked PCL is expected to occur faster than in PBS due to enzymatic degradation.25

22 ACS Paragon Plus Environment

Page 22 of 38

Page 23 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Figure 5- Degradation profiles for cross-linked PCLTA 300 ( ), PCLTA 900 ( ), PCLDA 530 ( ), and PCLDA 1250 ( ) polymer discs in A) PBS and B) 1 M NaOH to accelerate hydrolytic degradation. Mass loss was determined gravimetrically using dried samples. Experiments were performed with three replicates for each polymer.

In Vitro Biocompatibility To be considered effective, biomaterials must be biocompatible and perform desired functions with an appropriate host response.39 Changes in prepolymer molecular weight and functionality are expected to influence biocompatibility and cell survival due to the range of mechanical properties that result from variations in cross-link density. In this study, cross-linked PCL coated with a layer of extracellular matrix (ECM) proteins (i.e., Matrigel) supports the attachment and growth of mouse induced pluripotent stem cells (MiPSCs), regardless of molecular weight or functionality (Figure 6). One week after being seeded, densely packed MiPSCs are observed on all cross-linked PCL samples (Figure 6A-D). Although cells on both samples are within the expected proliferative range,30 cells are significantly less abundant (p < 0.01) but more mitotically active (p < 0.05) on PCLTA 530 films than on PCLTA 300 films (Figure 6E-F). Since mitotic index is a relative measure of actively dividing cells, this result could indicate that cells seeded on PCLTA 300 proliferate more rapidly and have already reached confluence by one week, while cells on PCLDA 530 may take longer to reach a contact-inhibited state. Given the otherwise similar quantity and mitotic index 23 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

of cells attached to cross-linked PCL samples, distinct differences in biocompatibility due to prepolymer molecular weight or functionality cannot be identified. Thus, this experiment provides strong evidence that the cross-linked PCL described in this study is a suitable substrate for many cell types. Given that stem cells are notoriously sensitive to acidic or otherwise toxic environments, their attachment and growth on cross-linked PCL is a promising indicator of its suitability for 3D-printed regenerative medicine applications.

Figure 6- In vitro biocompatibility of mouse induced pluripotent stem cells on cross-linked PCLTA 300 (A), PCLTA 900 (B), PCLDA 530 (C), and PCLDA 1250 (D) polymers. Immunocytochemistry and confocal microscopy were used to determine number of cells on each sample (E) and the mitotic index (F). Seven fields were imaged for each sample.

24 ACS Paragon Plus Environment

Page 24 of 38

Page 25 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

In Vivo Biocompatibility Rapid resorption of degradable polyesters can cause severe damage to host tissues, particularly those that comprise sensitive neuronal cells.12,40 In this study, we aim to demonstrate the feasibility of using slow-degrading cross-linked poly(caprolactone) in regenerative medicine applications, including those involving sensitive neuronal tissue. We chose the eye as our model organ because it contains the most metabolically demanding neurons in the human body (photoreceptors), is readily accessible, and its status is relatively easy to assess compared to other neurological tissues. As this is a pilot study to merely assess the in vivo compatibility and ability to surgically manipulate appropriately-sized PCL materials, we selected only one prepolymer for in vivo evaluation: PCLDA 1250. This system was selected based on the mechanical results observed in Figure 4, because it exhibits significantly higher strain at break than the other polymers, allowing much more flexibility during the implantation process. Five animals received sub-retinal transplantation of cross-linked PCLDA 1250 in one eye and a sham surgery (control) in the contralateral eye. One month after surgery, ophthalmoscopy revealed that all 10 eyes had complete, spontaneous retinal reattachment (Table 2): a positive sign of the natural healing process after surgical manipulation of the retina (Table 2). Cross-linked PCL implants are detectable by ophthalmoscopy in all treated eyes (Table 2, Figure 7A), and their anatomic location in the subretinal space confirmed with spectral-domain optical coherence tomography (SD-OCT) (Figure 7B). On SD-OCT, the surface of the polymer is hyper-reflective while the body is hypo-reflective compared to the retina. Importantly, none of the eyes show evidence of intraocular inflammation or other vitreous opacities (Table 2), indicating a quiescent immune response to the procedure. Hematoxylin-eosin and immunohistochemically stained sections through a subset of injected eyes demonstrate subretinal location of the implanted polymer, and preservation of retinal layers adjacent to the implant (Figure 7C-D). These results demonstrate that cross-linked PCL can be successfully delivered to the subretinal space and is well-tolerated ophthalmically, even without immunosuppressants. The polymer selected for sub-retinal 25 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

transplantation (PCLDA 1250) is the fastest to degrade in this study. Thus, it is unlikely that the use of any of the other three prepolymers would result in excessive local pH changes. The very positive surgical outcomes described here, paired with the in vitro biocompatibility described above, suggest that any of the four prepolymers are likely suitable starting materials for retinal tissue scaffolds and for regenerative medicine applications in a wide range of other organs. This knowledge will be fundamental to the development of effective autologous cell-based sub-retinal transplantation grafts for retinal degenerative diseases.

Figure 7- Cross-linked PCLDA 1250 in the pig sub-retinal space 30 days after transplantation. Representative photograph (A) and spectral-domain optical coherence tomography image (B) of the post-mortem eye show placement of the transplanted polymer (marked with *). Representative histological (C) and immunocytochemical (D) images of retinal sections show the embedded PCL scaffolds (marked with * and also autofluorescent in green) and surrounding tissue. Inset in C is a comparative section of porcine retina that did not receive a polymer scaffold.

Table 2- Ophthalmoscopic evaluation of cross-linked PCLDA 1250 in the pig 30 days after sub-retinal transplantation. Control Treated 26 ACS Paragon Plus Environment

Page 26 of 38

Page 27 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Scaffold Detection Retinal Reattachment Intraocular Inflammation Vitreous Opacities

N/A 5/5 0/5 0/5

5/5 5/5 0/5 0/5

3D Printing Precise control of scaffold design is integral when creating scaffolds with features at an appropriate sizescale to direct cell growth and differentiation in vivo. The high degree of biocompatibility exhibited by cross-linked PCL enables all synthesized prepolymers to be considered for tissue engineering, making the printed resolution a particularly important factor. The ability of PCL prepolymers to form high-resolution features is expected to be heavily influenced by prepolymer reaction rate and ultimate cross-link density. The reaction rates measured in Figure 3 were determined using a different light source and photoinitiator than are used for DLP stereolithography. As irradiation conditions are different, the previously measured rates cannot be used to directly determine polymerization rates in DLP SL. Although the rate of initiation and, consequently, rate of polymerization will change based on irradiation conditions, the overall trend in reaction rate of the prepolymers as functionality and molecular weight changes should be relatively independent of initiation rate.41 Therefore, the kinetic trends observed between prepolymers may be used to facilitate understanding of DLP SL results PLCDA 1250 and PCLTA 300 prepolymers were selected for 3D printing via DLP stereolithography because these prepolymers represent the slowest and fastest reacting prepolymers in this study, respectively, due to their molecular weight and functionality. As discussed earlier, the PCLTA 300 prepolymer is expected to be the most suitable prepolymer for creating high-resolution features on the micro-scale while the PCLDA 1250 prepolymer will likely be the least suitable of the acrylated prepolymers based on the reaction kinetics. However, many of the other properties such as modulus, strain at break, and degradation rate 27 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

make PCLDA 1250 a more desirable prepolymer for tissue scaffold fabrication. Thus, it is important to understand the effect of prepolymer molecular weight and functionality on the resolution and ability to print small features. Figure 8A and 8B show that both PCLDA 1250 and PCLTA 300 prepolymers are capable of forming cmscale structures via stereolithography. At this size scale, it is difficult to discern differences in micro-scale capabilities, therefore suitable small features were fabricated. A test object was designed with feature sizes starting at 500 µm and progressively decreasing in size until 50 µm, the resolution limit of the Autodesk Ember 3D printer. The objects were then printed using both prepolymers and the resulting structures imaged using SEM. Selected SEM images can be seen in Figures 8C and 8D with full results quantified in Figures 8E and 8F. Objects printed with PCLDA 1250 are 10% to 20% smaller than designed whereas the printed features for the PCLTA 300 ranged from 25% to 100% larger than designed. The smaller PCLDA 1250 features are likely due to solvent evaporation and inherent shrinkage that occurs during polymerization. This decrease in size is more pronounced in the PCLDA 1250 polymer than the PCLTA 300 because the PCLTA 300 polymer has a higher cross-link density and is therefore able to resist the shrinkage to a greater degree. The increase in feature size for the PCLTA 300 polymer results from a combination of increased diffusion of radicals outside the illuminated area, enabled by smaller prepolymer molecular weight, and faster reaction kinetics that allow the bulk prepolymer to quickly reach gel point around the fringes of light exposure, despite not being directly illuminated. Further measurements of fidelity can be observed in Figure 8G, where the circularity is measured for both polymers. Structures printed using PCLDA 1250 exhibit low values of circularity, near the values expected for a square. Structures printed using PCLTA 300 increase in circularity as the feature size decreases, until plateauing in a region of high deviation for structures below 150 µm. The largest difference between the

28 ACS Paragon Plus Environment

Page 28 of 38

Page 29 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

prepolymers is observed in the smallest feature size of objects that can be printed. Under these conditions, the smallest features printed from the PCLDA 1250 prepolymer are around 300 µm, while the PCLTA 300 prepolymer is polymerized to form features that are as small as 100 µm. These results confirm the hypothesis that under the same illumination conditions, a prepolymer with a faster polymerization rate leads to the ability to fabricate small-scale structures near the resolution limit of the 3D printer. The ability to create small features with high fidelity is an important consideration in choosing a material for 3D printed tissue scaffolds. This study underscores the ability to easily achieve microscale features using acrylated polycaprolactone prepolymers that exhibit a high degree of in vitro and in vivo biocompatibility.

29 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 8- 3D Printed PCLDA 1250 (left column) and PCLTA 300 (right column) fabricated using DLP stereolithography. Centimeter-scale objects (A and B) were printed to demonstrate the ability of the prepolymers to form 3-dimensional structures. Objects with micrometer-scale features were printed and imaged using SEM (C and D). The resulting resolution and fidelity are quantified using both a height and a width ratio (E and F), calculated by dividing the printed feature size by the designed feature size. Circularity (G) was also calculated by dividing the area by the perimeter squared (equation 3). A circularity value of 1 indicates a perfect circle and a circularity value of 0.785 indicates a square. Conclusions In this work, we demonstrate the ability to create three-dimensional structures using degradable, biocompatible poly(caprolactone) acrylate prepolymers and evaluate the role of molecular weight and prepolymer functionality on the reaction kinetics, mechanical properties, and degradation of cross-linked PCL. Higher cross-link density that results from lower molecular weight and higher functionality leads to

30 ACS Paragon Plus Environment

Page 30 of 38

Page 31 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

faster reaction kinetics, larger strain at break, and higher modulus than lower cross-link density polymers. Similarly, the high cross-link density of low-molecular weight, high-functionality prepolymers leads to longer degradation times due to the decreased availability of degradation sites. Regardless of molecular weight or functionality, cross-linked PCL coated with a layer of ECM proteins supports the attachment and growth of mouse induced pluripotent stem cells in vitro. High-molecular weight, low-functionality PCLDA was successfully implanted into a pig retina and is shown to cause no damage to host retinal tissue in vivo. Suitability for controlled scaffold fabrication was examined using digital light projection stereolithography, where the higher cross-link density PCLTA 300 prepolymer is capable of printing feature sizes on the order of 100 µm in this system, three times smaller than is possible with lower cross-link density PCLDA 1250. Together, this work sets the stage for cross-linked PCL to be used in tissue engineering applications that require control of mechanical properties, degradation rates, and 3-dimensional structure.

Supporting Information General structures for PCL diacrylates and triacrylates highlighting acrylate and reference hydrogens used in NMR analysis are presented in Figure S1. Integrated NMR spectra for all PCL prepolymers are shown in Figure S2. The degree of acrylation for PCL prepolymers was calculated using NMR by comparing acrylate hydrogens to reference hydrogens with results quantified in Table S1.

Acknowledgements This work funded in part by the National Science Foundation (CBET-1438486), Research to Prevent Blindness, the International Retinal Research Foundation, the Howard F. Ruby Endowment for Human Retinal Engineering, and the Stephen A. Wynn Foundation. The authors would like to thank Autodesk for

31 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

donating the Autodesk Ember 3D printer used in this research. This work utilized the Hitachi S-3400N in the University of Iowa Central Microscopy Research Facilities that was purchased with funding from the NIH SIG grant 1 S10 RR022498-01. Nuclear magnetic resonance was performed at the University of Iowa Central NMR Facility, which is supported by the NIH (S10-RR025500), the NSF (CHE-0840371) and University of Iowa funds. The authors would like to thank the staff at both facilities for their assistance and support. TOC Graphic

32 ACS Paragon Plus Environment

Page 32 of 38

Page 33 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

References (1)

Ghanbarzadeh, B.; Almasi, H. Biodegradable Polymers. In Biodegradation - Life of Science; InTech, 2013.

(2)

Siracusa, V.; Rocculi, P.; Romani, S.; Rosa, M. D. Biodegradable Polymers for Food Packaging: A Review. Trends Food Sci. Technol. 2008, 19 (12), 634–643.

(3)

Do, A.-V.; Khorsand, B.; Geary, S. M.; Salem, A. K. 3D Printing of Scaffolds for Tissue Regeneration Applications. Adv. Healthc. Mater. 2015, 4 (12), 1742–1762.

(4)

Ulery, B. D.; Nair, L. S.; Laurencin, C. T. Biomedical Applications of Biodegradable Polymers. J. Polym. Sci. Part B Polym. Phys. 2011, 49 (12), 832–864.

(5)

Ifkovits, J. L.; Burdick, J. a. Review: Photopolymerizable and Degradable Biomaterials for Tissue Engineering Applications. Tissue Eng. 2007, 13 (10), 2369–2385.

(6)

Hynes, S. R.; Lavik, E. B. A Tissue-Engineered Approach towards Retinal Repair: Scaffolds for Cell Transplantation to the Subretinal Space. Graefes Arch. Clin. Exp. Ophthalmol. 2010, 248 (6), 763– 778.

(7)

Pallua, N.; Suschek, C. V. Tissue Engineering; Pallua, N., Suscheck, C. V., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg, 2011.

(8)

Tamada, J. a; Langer, R. Erosion Kinetics of Hydrolytically Degradable Polymers. Proc. Natl. Acad. Sci. U. S. A. 1993, 90 (2), 552–556.

(9)

Sawhney, A. S.; Pathak, C. P.; Hubbell, J. A. Bioerodible Hydrogels Based on Photopolymerized Poly(Ethylene Glycol)-Co-Poly(.Alpha.-Hydroxy Acid) Diacrylate Macromers. Macromolecules 1993, 26 (4), 581–587.

33 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(10)

Clapper, J. D.; Guymon, C. A. Nanostructured Biodegradable Polymer Composites Generated Using Lyotropic Liquid Crystalline Media. Macromolecules 2007, 40 (22), 7951–7959.

(11)

Hashimoto, K.; Hamano, T.; Okada, M. Degradation of Several Polyamides in Soils. J. Appl. Polym. Sci. 1994, 54 (10), 1579–1583.

(12)

Bettinger, C. J.; Weinberg, E. J.; Kulig, K. M.; Vacanti, J. P.; Wang, Y.; Borenstein, J. T.; Langer, R. Three-Dimensional Microfluidic Tissue Engineering Scaffolds Using a Flexible Biodegradable Polymer. Adv Mater. 2005, 18 (2), 165–169.

(13)

Yao, J.; Tao, S. L.; Young, M. J. Synthetic Polymer Scaffolds for Stem Cell Transplantation in Retinal Tissue Engineering. Polymers (Basel). 2011, 3 (4), 899–914.

(14)

Mondal, D.; Griffith, M.; Venkatraman, S. S. Polycaprolactone-Based Biomaterials for Tissue Engineering and Drug Delivery: Current Scenario and Challenges. Int. J. Polym. Mater. Polym. Biomater. 2016, 65 (5), 255–265.

(15)

Woodruff, M. A.; Hutmacher, D. W. The Return of a Forgotten Polymer - Polycaprolactone in the 21st Century. Prog. Polym. Sci. 2010, 35 (10), 1217–1256.

(16)

Kweon, H.; Yoo, M. K.; Park, I. K.; Kim, T. H.; Lee, H. C.; Lee, H.-S.; Oh, J.-S.; Akaike, T.; Cho, C. S. A Novel Degradable Polycaprolactone Networks for Tissue Engineering. Biomaterials 2003, 24 (5), 801–808.

(17)

Kim, H. M.; Kim, H. R.; Hou, C. T.; Kim, B. S. Biodegradable Photo-Crosslinked Thin Polymer Networks Based on Vegetable Oil Hydroxy Fatty Acids. JAOCS, J. Am. Oil Chem. Soc. 2010, 87 (12), 1451–1459.

(18)

Cai, L.; Wang, S. Poly(ε-Caprolactone) Acrylates Synthesized Using a Facile Method for Fabricating Networks to Achieve Controllable Physicochemical Properties and Tunable Cell Responses. 34 ACS Paragon Plus Environment

Page 34 of 38

Page 35 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Polymer (Guildf). 2010, 51 (1), 164–177. (19)

Atzet, S.; Curtin, S.; Trinh, P.; Bryant, S.; Ratner, B. Degradable Poly(2-Hydroxyethyl Methacrylate)- Co -Polycaprolactone Hydrogels for Tissue Engineering Scaffolds. Biomacromolecules 2008, 9 (12), 3370–3377.

(20)

Chung, I.; Xie, D.; Puckett, A. D.; Mays, J. W. Syntheses and Evaluation of Biodegradable Multifunctional Polymer Networks. Eur. Polym. J. 2003, 39 (9), 1817–1822.

(21)

Wang, S.; Lu, L.; Gruetzmacher, J. A.; Currier, B. L.; Yaszemski, M. J. Synthesis and Characterizations of Biodegradable and Crosslinkable Poly(ε-Caprolactone Fumarate), Poly(Ethylene Glycol Fumarate), and Their Amphiphilic Copolymer. Biomaterials 2006, 27 (6), 832–841.

(22)

Liow, S. S.; Widjaja, L. K.; Lipik, V. T.; Abadie, M. J. M. Synthesis, Characterization and Photopolymerization of Vinyl Functionalized Poly (ε-Caprolactone). Express Polym. Lett. 2009, 3 (3), 159–167.

(23)

Kaya, N. U.; Onen, A.; Guvenilir, Y. Photopolymerization of Acrylates by Enzymatically Synthesized PCL Based Macrophotoinitiator. Express Polym. Lett. 2017, 11 (6), 493–503.

(24)

Davis, K. A.; Burdick, J. A.; Anseth, K. S. Photoinitiated Crosslinked Degradable Copolymer Networks for Tissue Engineering Applications. Biomaterials 2003, 24 (14), 2485–2495.

(25)

Rice, M. A.; Sanchez-Adams, J.; Anseth, K. S. Exogenously Triggered, Enzymatic Degradation of Photopolymerized Hydrogels with Polycaprolactone Subunits: Experimental Observation and Modeling of Mass Loss Behavior. Biomacromolecules 2006, 7 (6), 1968–1975.

(26)

Do, A. V.; Akkouch, A.; Green, B.; Ozbolat, I.; Debabneh, A.; Geary, S.; Salem, A. K. Controlled and Sequential Delivery of Fluorophores from 3D Printed Alginate-PLGA Tubes. Ann. Biomed. Eng. 35 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2017, 45 (1), 297–305. (27)

Teo, E. Y.; Ong, S. Y.; Khoon Chong, M. S.; Zhang, Z.; Lu, J.; Moochhala, S.; Ho, B.; Teoh, S. H. Polycaprolactone-Based Fused Deposition Modeled Mesh for Delivery of Antibacterial Agents to Infected Wounds. Biomaterials 2011, 32 (1), 279–287.

(28)

Park, S. H.; Park, D. S.; Shin, J. W.; Kang, Y. G.; Kim, H. K.; Yoon, T. R.; Shin, J. W. Scaffolds for Bone Tissue Engineering Fabricated from Two Different Materials by the Rapid Prototyping Technique: PCL versus PLGA. J. Mater. Sci. Mater. Med. 2012, 23 (11), 2671–2678.

(29)

Worthington, K. S.; Wiley, L. A.; Kaalberg, E. E.; Collins, M. M.; Mullins, R. F.; Stone, E. M.; Tucker, B. A. Two-Photon Polymerization for Production of Human IPSC-Derived Retinal Cell Grafts. Acta Biomater. 2017, 55, 385–395.

(30)

Anseth, K. S.; Wang, C. M.; Bowman, C. N. Kinetic Evidence of Reaction Diffusion during the Polymerization of Multi (Meth) Acrylate Monomers. Macromolecules 1994, 27 (3), 650–655.

(31)

Lam, C. X.; Savalani, M. M.; Teoh, S. H.; Hutmacher, D. W. Dynamics of in Vitro Polymer Degradation of Polycaprolactone-Based Scaffolds: Accelerated versus Simulated Physiological Conditions. Biomed Mater 2008, 3 (3), 34108.

(32)

Worthington, K. S.; Green, B. J.; Rethwisch, M.; Wiley, L. A.; Tucker, B. A.; Guymon, C. A.; Salem, A. K. Neuronal Differentiation of Induced Pluripotent Stem Cells on Surfactant Templated Chitosan Hydrogels. Biomacromolecules 2016.

(33)

Sohn, E. H.; Jiao, C.; Kaalberg, E.; Cranston, C.; Mullins, R. F.; Stone, E. M.; Tucker, B. A. Allogenic IPSC-Derived RPE Cell Transplants Induce Immune Response in Pigs: A Pilot Study. Sci. Rep. 2015, 5, 1–10.

(34)

Worthington, K. S.; Wiley, L. a; Bartlett, A. M.; Stone, E. M.; Mullins, R. F.; Salem, A. K.; Guymon, 36 ACS Paragon Plus Environment

Page 36 of 38

Page 37 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

C. A.; Tucker, B. A. Mechanical Properties of Murine and Porcine Ocular Tissues in Compression. Exp. Eye Res. 2014, 121C, 194–199. (35)

Tuft, B. W.; Li, S.; Xu, L.; Clarke, J. C.; White, S. P.; Guymon, B. A.; Perez, K. X.; Hansen, M. R.; Guymon, C. A. Photopolymerized Microfeatures for Directed Spiral Ganglion Neurite and Schwann Cell Growth. Biomaterials 2013, 34 (1), 42–54.

(36)

Tuft, B. W.; Zhang, L.; Xu, L.; Hangartner, A.; Leigh, B.; Hansen, M. R.; Guymon, C. A. Material Sti Ff Ness E Ff Ects on Neurite Alignment to Photopolymerized Micropatterns. Biomacromolecules 2014, 15, 3717–3727.

(37)

Leigh, B. L.; Truong, K.; Bartholomew, R.; Ramirez, M.; Hansen, M. R.; Guymon, C. A. Tuning Surface and Topographical Features to Investigate Competitive Guidance of Spiral Ganglion Neurons. ACS Appl. Mater. Interfaces 2017, 9 (37), 31488–31496.

(38)

Trese, M.; Regatieri, C. V; Young, M. J. Advances in Retinal Tissue Engineering. Materials (Basel). 2012, 5, 108–120.

(39)

Williams, D. F. On the Mechanisms of Biocompatibility. Biomaterials 2008, 29 (20), 2941–2953.

(40)

Nijst, C. L. E.; Bruggeman, J. P.; Karp, J. M.; Ferreira, L.; Zumbuehl, A.; Bettinger, C. J.; Langer, R. Synthesis and Characterization of Photocurable Elastomers from Poly(Glycerol-Co-Sebacate). Biomacromolecules 2007, 8 (10), 3067–3073.

(41)

Andrzejewska, E. Photopolymerization Kinetics of Multifunctional Monomers. Prog. Polym. Sci. 2001, 26 (4), 605–665.

37 ACS Paragon Plus Environment

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

TOC Graphic 87x44mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 38 of 38