© 2019

Nathan Zachary Dreger

ALL RIGHTS RESERVED Amino Acid-Based Poly(ester urea)s for Soft-tissue Repair Applications

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Nathan Z. Dreger

February 2019 AMINO ACID-BASED POLY(ESTER UREA)S FOR SOFT-TISSUE REPAIR APPLICATIONS

Nathan Zachary Dreger

Dissertation

Approved: Accepted:

______Advisor Department Chair Dr. Matthew L. Becker Dr. Tianbo Liu

______Committee Member Dean of College Dr. Li Jia Dr. Ali Dhinojwala

______Committee Member Dean of the Graduate School Dr. Chrys Wesdemiotis Dr. Chand Midha

______Committee Member Date Dr. Darrell Reneker

______Committee Member Dr. Rebecca Willits

ii ABSTRACT

Preclinical Assessment of L-Valine-Based Poly(ester urea)s for Soft-Tissue

Applications. New degradable polymer materials are required to aid in the treatment of soft-tissue injuries. Designing materials with mechanical properties similar to that of the tissue at the implant site is often overlooked. To more fully meet the needs required for an implantable biomaterial, a series of amino acid-based linear and branched poly(ester urea)s (PEU)s with varied diol chain length spacers were synthesized that were found to have mechanical properties similar to clinically employed polymer materials used in the treatment of hernia-repair. Concomitantly, these materials were designed and found to be degradable in vivo in to relatively benign byproducts readily excreted indicated by a limited inflammatory response through a comprehensive histological assessment.

Amino Acid-based Poly(ester urea) Copolymer Films for Hernia-Repair

Applications. The use of synthetic degradable materials is needed to help bridge the gap between currently utilized materials to aid in the treatment of hernia-repair.

Readily available clinical materials include but are not limited to non-resorbable synthetic materials (i.e. polypropylene) which are mechanically robust however are permanent, and biologically derived tissues (i.e. small intestine submucosa extracellular matrix (SIS-ECM)) which provoke a limited inflammatory response yet are precluded by failure at low

iii strain. To improve upon current options, a series of amino acid-based poly(ester urea)

(PEU) copolymers comprised of L-valine and L-phenylalanine were synthesized and fabricated in to solvent cast stand-alone films or as composite films with SIS-ECM. Cell spreading, proliferation, viability, and an improved inflammatory response was observed in a rat hernia model for all PEU derivatives. Additionally, improved mechanical integrity of SIS-ECM was observed when combined as a composite film with PEUs.

Surface-functionalized Poly(ester urea)s as Adhesion Barriers in Hernia-Repair.

Regardless of material selected to aid in the treatment of hernia-repair, there is a significant risk of adhesion of the implant to the peritoneal space. This complication can lead to a fusion between the abdominal wall and other organs which often times requires further corrective surgery. A surface functionalized anti-adhesion poly(ester urea) (PEU)s with varied amounts of a scalable zwitterion thiol attached via thiol-ene chemistry was found to have limited cytotoxicity while also preventing adhesion of fibrinogen in vitro. Furthermore, a reduction of adhesion was observed for surface functionalized materials when compared to unfunctionalized controls in an intraabdominal rat adhesion model. DEDICATION

This body of work is dedicated in three parts. First to my wife, Jessica. You have always supported my endeavors with patience and love. This work would not have come to fruition without you. Second, to my parents, Eric and Theresa Dreger. You have created and fostered an environment of “life-long learning” that has undoubtedly facilitated where I have gone in life. For this I am eternally grateful.

Finally, to my brother Kyle. My relationship with you has helped me become a better thinker and, more importantly, a better person. Throughout this process, it has helped me keep perspective.

v ACKNOWLEDGEMENTS

This body of work could not have been completed without the contribution of many individuals. First and foremost, I want to thank my advisor, Dr. Matthew L.

Becker. You have fostered an environment where the team is more important than the individual and where the best idea has to win. This enables students of all ages, backgrounds, and skill levels to help and contribute; this has driven me to contribute more and more. Al Davis said it best, “Just win baby, win.”

I also want to thank my committee members Dr. Li Jia, Dr. Rebecca Willits, Dr.

Chrys Wesdemiotis, and Dr. Darrell Reneker for challenging me to be a better student in the classroom and a better researcher in the laboratory.

To all Becker group members, thank you for constantly creating a fun and collaborative work environment. On the road towards this degree, it has made all the difference.

Finally, I want to thank my family for the constant support. The completion of this body of work is as much theirs as it is mine.

vi TABLE OF CONTENTS

LIST OF TABLES ...... xi LIST OF FIGURES ...... xii CHAPTER ...... 1 II. INTRODUCTION ...... 1 1.1. Material design for diverse applications ...... 1

1.2. The role of polymers in the evolution of hernia-repair ...... 18

II. MATERIALS AND INSTRUMENTATION ...... 27 2.1. Materials ...... 27

2.2. Instrumentation ...... 27

III. PRECLINICAL IN VITRO AND INVIVO ASSESSMENT OF LINEAR AND BRANCHED L-VALINE- BASED POLY(ESTER UREA)S FOR SOFT TISSUE APPLICATIONS ...... 30 3.1. Abstract ...... 30

3.2. Introduction...... 31

3.3. Experimental ...... 34

3.3.1. Materials ...... 34 3.3.2. Characterization ...... 34 3.3.3. Synthesis of Poly(ester urea) monomers ...... 35 3.3.4. Synthesis of Poly(ester urea) polymers ...... 38 3.3.5. Mechanical Property Measurements ...... 41 3.3.6. In Vivo Implant Degradation ...... 42 3.3.7. Host-Implant Interaction ...... 43 3.4. Results ...... 43

vii 3.4.1. Synthesis...... 43 3.4.2. Physical Properties ...... 45 3.4.3. In Vivo Degradation ...... 47 3.4.4. Mechanical Properties ...... 51 3.4.5. Histology ...... 55 3.5. Conclusion ...... 60

3.6. Acknowledgement ...... 60

IV. AMINO ACID-BASED POLY(ESTER UREA) COPOLYMER FILMS FOR HERNIA-REPAIR APPLICATIONS ...... 62 4.1. Abstract ...... 62

4.2. Introduction...... 63

4.3. Experimental ...... 66

4.3.1. Materials ...... 66 4.3.2. Characterization ...... 67 4.3.3. Synthesis of Poly(ester urea) Copolymer Monomers ...... 68 4.3.4. Synthesis of Poly(ester urea) Copolymers ...... 69 4.3.5. Uniaxial Mechanical Property Measurements ...... 73 4.3.6. PEU-ECM Films ...... 73 4.3.7. PEU Free-Standing Films ...... 74 4.3.8. Film Water Uptake ...... 74 4.3.9. Contact Angle ...... 75 4.3.10. In Vitro Degradation ...... 75 4.3.11. Cell Viability ...... 76 4.3.12. In Vivo Animal Model ...... 77 4.3.13. Host-Implant Interaction ...... 78 4.4. Results ...... 79

4.4.1. Synthesis...... 79

viii 4.4.2. Physical Properties ...... 81 4.4.3. Water Uptake and Contact Angle ...... 82 4.4.4. Mechanical Properties ...... 84 4.4.5. Cell Culture Studies ...... 91 4.4.6. Histological Assessment ...... 94 4.4.7. Film Degradation ...... 100 4.5. Conclusion ...... 102

4.6. Acknowledgement ...... 102

V. ZWITTERIONIC AMINO ACID POLY(ESTER UREA)S SUPPRESS ADHESION FORMATION IN A RAT INTRA-ABDOMINAL CECAL ABRASION MODEL ...... 104 5.1. Abstract ...... 104

5.2. Introduction...... 105

5.2.1. Materials ...... 108 5.2.2. Characterization ...... 108 5.2.3. Synthesis of Poly(ester urea) monomers ...... 109 5.2.4. Lithium Phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) Photoinitiator Synthesis...... 111 5.2.5. Synthesis of 3-((3-((3-mercaptopropanoyl)oxy)propyl)dimethylammon- io)propane-1-sulfonate (Zwitterion-SH)...... 112 5.2.6. Synthesis of Poly(ester urea) Terpolymers...... 114 5.2.7. Terpolymer PEU Free-Standing Films ...... 115 5.2.8. FITC-PEG-Thiol Surface Functionalization of Terpolymer PEUs ...... 116 5.2.9. Zwitterion Surface Functionalization of Terpolymer PEUs...... 117 5.2.10. Protein Adsorption ...... 118 5.2.11. Cytotoxicity Assessment ...... 119 5.2.12. PEU Terpolymer Implantation ...... 120 5.2.13. Adhesion Histological Assessment ...... 121 5.2.14. Film Surface Topology ...... 122

ix 5.3. Results ...... 122

5.3.1. Synthesis...... 122 5.3.2. Physical Properties ...... 126 5.3.3. Mechanical Properties ...... 127 5.3.4. Surface Functionalization ...... 128 5.3.5. Protein Adsorption ...... 135 5.3.6. Cytotoxicity Assessment ...... 138 5.3.7. In Vivo Characterization ...... 140 5.4. Conclusion ...... 148

5.5. Acknowledgement ...... 149 VI. CONCLUSION ...... 151 6.1. Conclusion and Outlook ...... 151

6.1.1. Preclinical In Vitro and In Vivo Assessment of Linear and Branched Poly(ester urea)s for Soft tissue Applications ...... 151 6.1.2. Amino Acid-Based Poly(ester urea) Copolymer Films for Hernia-Repair Applications ...... 152 6.1.3. Poly(ester urea) Adhesion Barriers Aid in the Treatment of Hernia-Mesh Repair …………………………………………………………………………………………………………….154

REFERENCES ...... 156 APPENDIX A-SUPPORTING FIGURES ...... 178 APPENDIX B-SUPPORTING SCHEMES AND TABLES ...... 210

x LIST OF TABLES Table Page Number

Table 3.1. Molecular mass degradation determined following ethylene oxide sterilization and post-implantation by SEC. Physical properties of poly(1-VAL-8), poly(1- VAL-10), poly(1-VAL-12), poly[(1-VAL-8)0.98-co-(Triol-TYR)0.02], and poly[(1-VAL-10)0.98-co- (Triol-TYR)0.02] polymers analyzed in this study...... 46 Table 3.2. Mechanical properties comparison ...... 52 Table 3.3. H&E slide scoring scale ...... 59 Table 3.4. H&E slide scores for implanted materials at 3 months...... 59 Table 4.1. Physical properties of the PEU Copolymers ...... 82 Table 4.2. Uniaxial mechanical properties comparison ...... 85 Table 4.3. H&E slide modified scoring scale based on ISO 10993-6 Annex E ...... 97 Table 4.4. Histology slide scores ...... 97 Table 5.1. Physical properties of alloc-PEUs ...... 126 Table 5.2. FITC-PEG-SH Dye Attachment ...... 130 Table 5.3. Extent of adhesions to abraded cecum scoring scale ...... 143 Table 5.4. Tenacity of adhesions scoring scale ...... 143 Table 5.5. Molecular mass degradation values of 10% alloc-PEU derivatives ...... 148 Table 6.1. Fibrous capsule histology measurements ...... 210 Table 6.2. Burst-test PEU-ECM mechanical properties comparison ...... 211 Table 6.3. Free-standing film burst-test mechanical properties ...... 212

xi LIST OF FIGURES

Figure Page Number

Figure 1.1. A timeline of drugs and devices containing PLGA that have been approved for use in humans by the FDA. This list, while not comprehensive, illustrates the diverse applications that PLGA has been utilized within over the past several decades.5–13 ...... 4

Figure 1.2. Mechanical properties ranges (Young’s modulus and Ultimate Tensile Strength) of PLGA and tissues for which regenerative medicine strategies may be desirable on a logarithmic scale...... 5

Figure 1.3. (A) A spider diagram to outline a framework of properties that an ideal biomaterial would have and rating scales for the properties. (B) The ratings within this framework for several traditional and contemporary materials as examples (PLGA, poly(ester urea)s, aliphatic polycarbonates and thermoplastic polyurethanes, TPUs). An ideal material would simultaneously possess the five categories, in addition to having tunable properties across the range to meet a desired application...... 7

Figure 3.1. Scanning electron microscopy (SEM) was performed on all polymers at each time point to observe variations in the surface morphology. The surface topology of P(1-VAL-8) (A), p(1-VAL-10) (B), p(1-VAL-12) (C), PP (D), 2% branched p(1-VAL-8) (E), and 2% branched p(1-VAL-10) (F) after EtO sterilization (left) are compared after 2 (middle) and 3 months (right) of implantation. Images were captured at 750 x magnification and scale bars indicate 10 μm...... 50

Figure 3.2. Young’s modulus for implanted materials was extrapolated at each time- point through linear regression with R2 = 0.98. P(1-VAL-8), p(1-VAL-10), p(1-VAL-12), polypropylene, 2% branched p(1-VAL-8), and 2% branched p(1-VAL-10) moduli values were assessed through the 3 month time point (A). * or ** indicate a p value < 0.05 between a reference sample (first sample denoted with * or ** reading from left to right) and other samples sharing like symbols (n = 4-6 samples). For example, * indicates a significant difference between initial p(1-VAL-8) and post-EtO p(1-VAL-8), between initial p(1-VAL-8) and 2 month p(1-VAL-8), and between initial p(1-VAL-8) and 3 month p(1-VAL-8) moduli values. * does not indicate a significant difference between 2 month and 3 month p(1-VAL-8) moduli values. Yield stress (σy) was measured at the yield point for p(1-VAL-8), p(1-VAL-10), p(1-VAL-12), polypropylene, 2% branched p(1- xii VAL-8), and 2% branched p(1-VAL-10) samples through the 3 month time point (B). *, **, or *** indicate a p value < 0.05 between samples sharing similar symbols (n = 4-6 samples). Statistical difference can be discerned the same way as previously explained for moduli values. Yield strain (εy) was measured at the yield point for p(1-VAL-8), p(1- VAL-10), p(1-VAL-12), polypropylene, 2% branched p(1-VAL-8), and 2% branched p(1- VAL-10) samples through the 3 month time point (C). * or ** indicate a p value < 0.05 between samples sharing similar symbols (n = 4-6 samples). Statistical difference can be discerned the same way as previously explained for moduli values...... 54

Figure 3.3. Histology images of (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) olypropylene, (E) 2% branched P(1-VAL-8), and (F) 2% branched P(1-VAL-10) are from the cross-sectional area of polymer and surrounding tissue which was stained with hematoxylin and eosin. All images are from the 2 month timepoint at 20 x magnification with scale bars being equal to 1 mm...... 56

Figure 3.4. Capsule thickness values for 2 month samples (A) were measured to assess inflammatory response (* indicates p value < 0.01 between P(1-VAL-8) and P(1-VAL-12) and between P(1-VAL-8) and 2% branched P(1-VAL-8) samples. ** indicates p value < 0.01 between P(1-VAL-10) and P(1-VAL-12), and between P(1-VAL-10) and 2% branched P(1-VAL-8) samples. *** indicates p value < 0.01 between P(1-VAL-12) and polypropylene, and between P(1-VAL-12) and 2% branched P(1-VAL-10) samples. **** indicates p value < 0.01 between polypropylene and 2% branched P(1-VAL-8), and between polypropylene and 2% branched P(1-VAL-10) samples, n = 7). Capsule thickness values were also assessed for 3 month samples (B) (* indicates p value < 0.01 between P(1-VAL-8) and polypropylene samples. ** indicates p value < 0.01 between P(1-VAL-10) and polypropylene and between P(1-VAL-10) and 2% branched P(1-VAL-10) samples. *** indicates p value < 0.01 between P(1-VAL-12) and polypropylene samples. **** indicates p value < 0.01 between polypropylene and 2% branched P(1-VAL-8), and between polypropylene and 2% branched P(1-VAL-10) samples, n = 7)...... 57

Figure 4.1. Physical properties of all six copolymers were assessed utilizing several techniques. Thermogravimetric analysis (TGA) (A) was performed to determine the degradation temperature (Td) for each polymer and curves were reported. The Td for each copolymer was high enough to allow for these materials to be thermally processed through compression molding. Size-exclusion chromatography (SEC) (B) for each copolymer was performed with molecular mass traces indicating uniform dispersity between polymers with Đm values between 1.4-1.7. Differential scanning calorimetry (DSC) (C) curves indicate that the glass transition temperatures (Tg) for these materials are above physiological conditions. The Tg are significantly lower than the degradation temperature which allowed for these materials to be thermally processed through compression molding without degradation. Finally, all six copolymers and p(1-VAL-8) were assessed to determine how incorporation of L-phenylalanine and change in diol change length would affect water uptake in 1 × PBS through eight days (n = 3) (D). An

xiii increase in L-phenylalanine led to a decrease in water uptake for the PHE6 P(1-VAL-8) polymers. The opposite trend was observed for the PHE8 P(1-VAL-8) polymers as the diol chain length and disruption of interchain packing was the dominating factor...... 84

Figure 4.2. Uniaxial mechanical properties were assessed for all six copolymers and compared to the p(1-VAL-8) homopolymer and polypropylene. Tensile tests performed at 25 °C with a constant strain rate of 25.4 mm/min with all data extrapolated from the stress versus strain curves which are representative of 4-6 samples for each polymer (A). Young’s moduli values (B) for each polymer were extrapolated at 10% strain (*, **, ***, ****, ***** indicates a p value < 0.05 between a reference sample (first sample denoted with *, **, ***, or etc. reading from left to right) and other samples sharing like symbols (n = 4-6 samples)). For example, (* indicates p value < 0.05 between PP and 20% PHE8 P(1-VAL-8), between PP and 30% PHE8 P(1-VAL-8), and between PP and 30% PHE6 P(1-VAL-8)). * does not indicate a significant difference between 20% PHE8 P(1- VAL-8) and 30% PHE8 P(1-VAL-8) as * indicates a difference from PP (reference). Yield stress (σy) (C) for each polymer was measured at the yield point (*, **, ***, ****, or ***** indicates p value < 0.05 between samples sharing like symbols. Statistical difference can be observed as previously described for moduli values. Yield strain (εy) (D) for each polymer was measured at the yield point (*, **, or *** indicates a p value < 0.05 between samples sharing like symbols. Statistical difference can be observed as previously described for moduli values...... 87

Figure 4.3. Cell viability was determined using L-929 cells seeded on PEU copolymers for 48 h (A). All slides were compared to a glass control with no statistical difference being observed between the groups. Results indicate that blade coated PEU films are non- toxic towards mammalian cells. Cell spreading and attachment results for L-929 fibroblast connective tissue cell line on PEU copolymer surfaces were assessed. Cells cultured on all six copolymers were stained with rhodamine phalloidin (red-pink color) and DAPI (blue-purple color). All pictograms were collected using an IX 81 Olympus microscope at 40 × magnification (B-H). Scale bars are equal to 100 µm. Cell spreading was calculated using ImageJ software (I). 30% PHE6 P(1-VAL-8) had the greatest spreading area although no copolymers were significantly greater than glass (n = 5). Cell attachment values were normalized and statistically compared to the glass control slide with 30% PHE6 P(1-VAL-8) (E) having a p < 0.01 and 20% PHE8 P(1-VAL-8) having a p < 0.001 (J)...... 93

Figure 4.4. A rat hernia model was carried out where a ventral incision was made (A), and a hernia defect was created on either side of the midline incision (B). Disc implants were fixed using PGA sutures to cover the hernia defect (C) on either side of the ventral incision. After two implants were placed per animal, the incision was closed with 4-5 sutures (D). Histology pictograms were obtained by staining implanted films in hematoxylin and eosin. Pictograms are oriented with the material on the right and abdominal wall (AW) on the left. Polypropylene (A), SIS-ECM (B), and 30% PHE6 P(1-

xiv VAL-8) (C) were imaged at 7 days and again at 14 days (G, H, and I respectively). Pictograms were captured at 20 × magnification with scale bars being equal to 1 mm. 95

Figure 4.5. Film surface morphology for each material was characterized using scanning electron microscopy (SEM) at 200 × magnification with scale bars equal to 10 µm. Polypropylene SIS-ECM, and 30% PHE6 P(1-VAL-8) films were analyzed after ethylene oxide (EtO) sterilization (A, B, and C respectively). Polypropylene and 30% PHE6 P(1- VAL-8) are noticeably smooth while ECM has cavities and undulations expected with its porous structure. After 14 days in vivo implantation, samples were analyzed again with polypropylene looking unchanged (D), ECM having thick tissue accumulation (E), and 30% PHE6 P(1-VAL-8) having small defects start to appear (F)...... 101

Figure 5.1. Polymer 1H-NMR overlay of previously synthesized 0% Alloc-PEU79 and 5% alloc-PEU and 10% alloc-PEU. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine, the methylene resonances from L-phenylalanine denoted ‘l’ in blue, and the methylene resonances from the benzyl protected L-tyrosine denoted ‘v’ in green...... 124

Figure 5.2. Tensile testing mechanical properties of PEU terpolymers. The stress versus strain curves (A) indicate that there is greater stiffness (B) for the 10% alloc PEU when compared to the 5% alloc analogue. There is an increase in yield strain for the 5% alloc- PEUs when compared to the 10% alloc-PEUs (C). No statistical difference for yield stress was observed for between the PEU derivatives (D). An * indicates a p value < 0.05 (n = 3 samples)...... 128

Figure 5.3. FITC-PEG-SH fluorescence dye attachment for surface functionalization and physically adsorbed 5% and 10% alloc-PEU analogues at 550 nm endpoint emission. An increase in dye concentration is observed for the 10% alloc-PEU functionalized polymer when compared to 10% alloc-PEU physically adsorbed. The same trend is observed for the 5% alloc-PEU analogues...... 130

Figure 5.4. High resolution XPS of the N1s orbital for the 5% alloc-PEU functionalized, physically adsorbed, and blank materials were plotted (A). A slight broadening is observed with curve fitting indicating that there are two distinct peaks for the urea nitrogen (~399.9 eV) and quaternary ammonium nitrogen (~401.5 eV) for the 5% alloc- PEU functionalized material (integration 95.2:4.8) (B). High resolution XPS of the S2p orbital was also taken and two distinct sulfur peaks correlating to the zwitterion-S-C sulfur (~162 eV) and ring opened sultone sulfur (~168 eV) are observed (integration 48:52) (C) were observed. High resolution XPS of the N1s orbital for the 10% alloc-PEU functionalized, physically adsorbed, and blank materials were plotted (D). A broadening is observed with curve fitting for the urea nitrogen (~399.9 eV) and quaternary ammonium nitrogen (~401.5 eV) showing integration values of 65.5:35.5 (E). High resolution XPS of the S2p orbital of the 10% alloc-PEU functionalized material showed

xv two distinct sulfur peaks correlating to the zwitterion-S-C sulfur and ring opened sultone sulfur are again observed (integration 55:45) (F)...... 133

Figure 5.5. Contact angle of functionalized and blank alloc-PEU analogues. Alloc-PEUs were coated with deionized water (5 µL), allowed to equilibrate for 10 minutes, and finally imaged with a goniometer (A-D). Angles were measured using ImageJ software to afford the contact angle for each material (E). The contact angle of the 5% alloc-PEU blank material and the 5% alloc-PEU functionalized material is 52.3 ± 0.5 and 46.6 ± 3.6 respectively (n = 4-6). The contact angle of the 10% alloc-PEU blank and 10% alloc-PEU functionalized materials were 61.8 ± 3.1 and 56.3 ± 7.6 respectively (n = 4-6)...... 135

Figure 5.6. QCM for 5% and 10% alloc-PEU analogues with fibrinogen. QCM representative curves for fibrinogen (1.5 mg/mL) on alloc-PEU analogues from the 5th overtone are reported (A, B). The amount of initially attached fibrinogen to the surface is reported (C) for 5% alloc-PEU blank (0.123 ± 0.015 ng/cm2), 5% alloc-PEU functionalized (0.024 ± 0.002 ng/cm2), 10% alloc-PEU blank (0.031 ± 0.017 ng/cm2), and 10% alloc-PEU functionalized (0.017 ± 0.001 ng/cm2) materials (n = 3-4) (an * indicates a statistical difference between 5% alloc-PEU blank materials and any material that shares this symbol, p < 0.05). The amount of irreversibly attached fibrinogen to the surface is reported (D) for 5% alloc-PEU blank (0.105 ± 0.046 ng/cm2), 5% alloc-PEU functionalized (0.175 ± 0.009 ng/cm2), 10% alloc-PEU blank (0.007 ± 0.009 ng/cm2), and 10% alloc-PEU functionalized (0.011 ± 0.001 ng/cm2) materials (n = 3-4) (an * indicates a statistical difference between 5% alloc-PEU blank materials and any material that shares this symbol, p < 0.05)...... 137

Figure 5.7. In vitro cytotoxicity assessment of alloc-PEUs. 5% alloc-PEU blank (A), 5% alloc-PEU functionalized (B), 10% alloc-PEU blank (C), and 10% alloc-PEU functionalized (D) materials’ extracted media was cultured with NIH-3T3 cells for 48 hours. Cell culture plates were then imaged and cell death and distress were scored from 0-4 (0 = limited cell death and equivalent to negative control, 1 = < 25% cell death and distress, 2 = < 50% cell death and distress, 3 = < 75% cell death and distress, and 4 = > 75% cell death and distress). Three controls were used where cells cultured without serum acted as a positive control (E), cells cultured with serum acted as a negative control (F), and cells cultured with media from Latex cots acted as a positive material control (G). All groups were scored with all alloc-PEU materials scoring equivalent to the negative control (score = 0) (n = 3 for alloc-PEUs, n = 2 for negative control). Both positive controls elicited a cytotoxic score (score = 4) (n = 2 for positive controls)...... 139

Figure 5.8. Adhesion extent and tenacity of explanted materials. The overall extent of adhesion (includes adhesions between tissue and abraded cecum and adhesion of abraded cecum to the device) with the sham, adhesion positive control, 10% alloc-PEU blank, and 10% alloc-PEU functionalized material having scores of 0 ± 0, 2.3 ± 1.4, 2.4 ± 1.0, and 2.0 ± 1.1 respectively (A). Tenacity of adhesions to the device and underlying tissue were scored with the sham, adhesion positive control, 10% alloc-PEU blank, and xvi 10% alloc-PEU functionalized material having scores of 0.0 ± 0.0, 2.4 ± 0.5, 2.6 ± 0.5, and 2.3 ± 0.7 respectively (B). The extent of adhesions exclusively attached to the implanted device and abraded cecum were scored with minimally observed adhesions for the 10% alloc-PEU functionalized material (0.3 ± 0.5) and the 10% alloc-PEU blank material (0.9 ± 0.7) (C). No tenacity of adhesion was observed for the 10% alloc-PEU functionalized material with slight tenacity being shown for the 10% alloc-PEU blank material (1.1 ± 0.9) (D). No statistical differences were observed between groups on a p < 0.05 confidence interval (n = 7-8)...... 144

Figure 5.9. Tissue samples were stained with hematoxylin and eosin (H&E) or Masson’s trichrome (tri) to identify collagen deposition and cell types present surrounding the adhesion areas. Representative images are shown for the groups under study. Sham control groups displayed a normal cecum without abnormal cellular densities (A-B). All other groups displayed hyperplastic lymphoid tissue (C-D) and blue-stained bundles of collagen fibers containing numerous fibroblasts and several giant cells (E-F)...... 146

Figure 5.10. Scanning electron microscopy (SEM) images of solvent cast films post-EtO sterilization (A), and of 10% alloc-PEU blank (B) and 10% alloc-PEU functionalized (C) films after 21 days of implantation...... 147

Figure 5.11. Molecular mass degradation of 10% alloc-PEU derivatives using size- exclusion chromatography-gel permeation chromatography (SEC-GPC) in THF...... 148

Figure 6.1. A) 1H-NMR overlay shows the spectra of linear monomers 1-VAL-8, 1-VAL- 10, and 1-VAL-12. The resonances “b” indicated in the light blue highlighted regions denote the methylene resonances for each of the three monomers. B) 1H-NMR of the Triol-TYR is shown (top spectrum), and successful Boc-deprotection (bottom spectrum) was identified by the disappearance of the methylene resonance at 1.28 ppm and the appearance of the broad amine proton resonance between 8.72-8.78 ppm, both of which are highlighted in blue...... 178

Figure 6.2. A) 1H-NMR overlay spectra of linear poly(ester urea)s poly(1-VAL-8), poly(1- VAL-10), and poly(1-VAL-12). The highlighted resonances “b” indicate the methylene resonances from the diol chain lengths.for each of the three polymers. B) 1H-NMR spectra of branched poly[(1-VAL-8)0.98-co-(Triol-TYR)0.02] and C) poly[(1-VAL-10)0.98-co- (Triol-TYR)0.02] are shown. The degree of branching was determined by integration of the six methylene protons denoted “e” from the Triol-TYR monomer and comparing them to the twelve methyl l-valine protons denoted “n” from the linear monomers. . 180

Figure 6.3. Thermogravimetric analysis (TGA) of the linear (A) and branched polymers (B) shows that the degradation temperatures for these materials is well above the temperature required for compression molding...... 181

xvii Figure 6.4. Differential scanning calorimetry (DSC) of the linear and branched polymers shows that the glass transition temperatures of these materials are near physiological conditions. The glass transition temperatures are significantly lower than the degradation temperature which allowed for these materials to be processed...... 182

Figure 6.5. Size-exclusion chromatographs for each PEU synthesized in this study comparing the initial molecular mass, post EtO sterilization, and the 2 and 3 month time points in vivo for (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) 2% branched P(1- VAL-8), and (E) 2% branched P(1-VAL-10)...... 183

Figure 6.6. Stress-strain curves for (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) PP, (E) 2% branched P(1-VAL-8), and (F) 2% branched P(1-VAL-10) were obtained from tensile tests performed at 25 °C with an extension rate of 25.4 mm/min. All mechanical data were extrapolated from the curves which represent an average of 4-6 samples. 184

Figure 6.7. 1H-NMR monomers 1-PHE-8 monomer shows successful synthesis based on the characteristic L-phenylalanine aromatic resonances denoted ‘d, e, f’ and the p- toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties...... 185

Figure 6.8. 1H-NMR monomers 1-PHE-6 monomer shows successful synthesis based on the characteristic L-phenylalanine aromatic resonances denoted ‘d, e, f’ and the p- toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties...... 186

Figure 6.9. 1H-NMR of 1-VAL-8 monomer shows successful synthesis based on the characteristic L-valine methyl resonance denoted ‘k’ and p-toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties...... 187

Figure 6.10. 1H-NMR overlay of P(1-VAL-8) and PHE6 poly(1-VAL-8) PEU polymers show successful synthesis. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine and the methylene resonances from L-phenylalanine denoted ‘l’ in blue. As the molar composition rises from 10-30% more of the L-phenylalanine resonances can be observed...... 188

Figure 6.11. 1H-NMR overlay of P(1-VAL-8) and PHE8 poly(1-VAL-8) PEU polymers show successful synthesis. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine and the methylene resonances from L-phenylalanine denoted ‘l’ in blue. As the molar composition rises from 10-30% more of the L-phenylalanine resonances can be observed...... 189

xviii Figure 6.12. 13C-NMR of PEU copolymers shows successful synthesis. The characteristic L-valine methyl resonances are observed between 18-20 ppm while the L-phenylalanine ring resonances can be seen with between 125-136 ppm...... 190

Figure 6.13. Water contact angle was used to determine surface properties of the six copolymers and how they compared to p(1-VAL-8) (n = 3) (D). An increase in water contact was observed for all copolymers with incorporation of L-phenylalanine content...... 191

Figure 6.14. Relative stiffness of PEU-ECM composite films was recorded by dividing the force at break by the extension at break (A). There was no significant difference among any samples which indicates that the films force and extension at break is proportional to that of free standing ECM. Force at break for ECM and PEU-ECM composite films were recorded from the force versus extension curves (* indicates p value < 0.05 between ECM and 10% PHE8 P(1-VAL-8) and between ECM and 30% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8) and between 10% PHE6 P(1-VAL-8) and 30% PHE8 P(1-VAL-8). *** indicated p value < 0.05 between 10% PHE8 P(1-VAL-8) and 20% PHE8 P(1-VAL-8). **** indicates p value < 0.05 between 20% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (B). Extension at break for ECM and PEU-ECM composite films were recorded from the force versus extension curves (* indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (C). Relative stiffness of free-standing films was recorded calculated the same way as previously described (* indicates p value < 0.05 between ECM and all six copolymers. ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (D). Force at break for free-standing films were recorded from the force versus extension curves (* indicates p value < 0.05 between ECM and 10% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8) and between 10% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (E). Extension at break for free-standing films were recorded from the force versus extension curves (* indicates p value < 0.05 between ECM and 10% PHE6 P(1-VAL-8), between ECM and 20% PHE6 P(1-VAL-8), between 30% PHE6 P(1-VAL-8), and between ECM and 30% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 20% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8) and between 20% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (F)...... 192

Figure 6.15. Sprague-Dawley rat showing abdominal bulge indicating that the abdominal incisional model can induce a hernia...... 194

Figure 6.16. Hydrolytic degradation rates for the fabricated films, SIS-ECM, 30% PHE6 P(1-VAL-8)-ECM in 0.025 M NaOH (A) and 30% PHE6 P(1-VAL-8) films in 0.1 M NaOH (n = 3)...... 195 xix Figure 6.17. Size-exclusion chromatography was used to obtain the molecular mass of the 30% PHE6 P(1-VAL-8) copolymer prior to implantation (red) and after 14 day implantation (black)...... 196

Figure 6.18. 1H-NMR of 1-TYR-2 Alloc monomer shows successful synthesis based on the successful deprotection noted by the disappearance of the boc peak and appearance of the broad amine...... 197

Figure 6.19. 1H-NMR of 3,3’-dithiodipropionyl chloride...... 198

Figure 6.20. 1H-NMR of bis(3-(dimethylamino)propyl) 3,3'-disulfanediyldipropionate.199

Figure 6.21. 1H-NMR of zwitterion disulfide...... 200

Figure 6.22. 13C-NMR of zwitterion disulfide...... 201

Figure 6.23. 1H-NMR of 3-((3-((3-mercaptopropanoyl)oxy)propyl)di-methylammon- io)propane-1-sulfonate (zwitterion-SH)...... 202

Figure 6.24. 13C-NMR of 3-((3-((3-mercaptopropanoyl)oxy)propyl)di-methylammon- io)propane-1-sulfonate (zwitterion-SH)...... 203

Figure 6.25. Size-exclusion chromatography of 5 and 10% alloc-PEUs in THF...... 204

Figure 6.26. Differential scanning calorimetry of 5 and 10% alloc-PEUs...... 205

Figure 6.27. Thermogravimetric analysis of 5 and 10% alloc-PEUs...... 206

Figure 6.28. 1000 MW FITC-PEG-SH calibration curve with concentration values reported in triplicate and mean intensity of emission at 550 nm recorded...... 207

Figure 6.29. 1H-NMR of lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP)...... 208

Figure 6.30. Extent of adhesion images for implanted 10% alloc-PEU materials after three weeks of implantation. Highlighted areas show adhesions that circumvented the device and adhesions that formed to the sutures used to retain the 10% alloc-PEU blank device (A). The highlighted area shows adhesions circumventing the device to form on the abraded cecum underneath for the 10% alloc-PEU functionalized device (B)...... 209

xx LIST OF SCHEMES

Scheme Page Number

Scheme 3.1. General synthetic scheme for L-valine monomers with diol-chain length varied between 8, 10, and 12 methylene units. Poly(1-VAL-8), poly(1-VAL-10), and poly(1-VAL-12) were synthesized using interfacial polymerization with triphosgene. Branched PEUs were synthesized using 1-VAL-8 and 1-VAL-10 with a 2% molar feed ratio of Triol-TYR to afford 2% branched poly(1-VAL-8) and 2% branched poly(1-VAL-10), respectively...... 45 Scheme 3.2. General tensile bar implantation. Tensile bars (A) were cut with a dye- cutter and subcutaneously implanted in to the backs of female Sprague-Dawley rats (B). Basic surgical procedures included subcutaneous incision (C) with surgical blade, subcutaneous pocket tunneling (D) with hemostats followed by polymer tensile bar insertion (E) and final incision closure (F) with Michel clips...... 47 Scheme 4.1. A general synthetic route for the monomer synthesis of 1-VAL-8, 1-PHE-6, and 1-PHE-8 carried out via a Fischer esterification between varying diol chain lengths and amino acids. In total six copolymers were synthesized with two of the monomers combining 1-VAL-8 and 1-PHE-6 in three different stoichiometric ratios to form 10% PHE6 P(1-VAL8), 20% PHE6 P(1-VAL-8), and 30% PHE6 P(1-VAL-8) or combining 1-VAL-8 and 1-PHE-8 to form 10% PHE8 P(1-VAL-8), 20% PHE8 P(1-VAL-8), and 30% PHE8 P(1- VAL-8). All polymers were synthesized using triphosgene to form urea linkages...... 80 Scheme 4.2. Roll-to-roll processing film fabrication set up by first dissolving polymer pellets in an acetone solution (A). Solutions were allowed to dissolve overnight before being fed in to the solution well with the doctor blade. A polyethylene terephthalate (PET) substrate, coating direction, and processing speed (150 cm/min) were set on the Bio-Hybrid processing line (B & C). For composite films, ECM was further adhered to the PET substrate and coated to afford the PEU-ECM film (D) while the free-standing PEU films were simply coated on PET...... 89 Scheme 4.3. Burst testing set up for composite PEU-ECM and free-standing PEU copolymer films. Burst-testing was adapted from ASTM D 3787-07 standards. Burst-test clamp surface was (A) cleaned prior to film fixation using tissue-paper. PEU-ECM composite films were submerged in 1 × PBS for 5 min prior to being placed on the bottom clamp (B). Films were then fastened (C) with a top clamp with two fixation

xxi screws. Burst-testing was performed (D) at an extension rate of 25.4 mm/min. The force versus extension curves were recorded for composite PEU-ECM films (E) and for free-standing films (F) (n = 3)...... 91 Scheme 6.1. General synthetic scheme for PEU terpolymer analogues where the monomer synthesis of 1-VAL-8, 1-PHE-6, and 1-TYR-2 alloc were carried out via a Fischer esterification or a DIC coupling between varying diol chain lengths and amino acids. In total two terpolymers were synthesized by combining 1-VAL-8 and 1-PHE-6 and 1-TYR-2 alloc in two stoichiometric amounts to form 5% alloc-PEU and 10% alloc-PEU. Both polymers were formed through interfacial polymerization with triphosgene to afford urea units. Polymers were further functionalized with a thiol-ene reaction between a zwitterion-SH and the alkene functionality...... 125

xxii CHAPTER I

INTRODUCTION

In part, this work has been submitted to Nature Materials as Dove, A. P.; Dreger,

N. Z.; Becker, M. L., Bioresorbable Polymers – Lost in Translation, 2018, Submitted.

1.1. Material design for diverse applications

Poly(lactic-co-glycolic acid) (PLGA) materials and other polyester formulations are found in numerous medical devices and drug delivery systems. As a material, PLGA dominates the landscape of bioresorbable polymers, but despite wide ranging exploration and study there are several unmet medical needs that could be advanced with the right polymer. As the materials community is beginning to transition to studying de novo resorbable biomaterials, this introduction seeks to address why PLGA remains the “go to” bioresorbable polymer and understand how new materials can be translated to tackle unsolved medical problems.

Resorbable materials are advantageous in many device, drug delivery and regenerative medicine applications because they naturally degrade in vivo to leave behind native tissue which promotes a more complete healing response. This stands in contrast to many nonresorbable materials which may elicit a prolonged inflammatory response and permanently alter the site of implantation. In addition, devices fabricated from

1 degradable materials are classified as temporary devices, and regulatory assessments of safety and efficacy over their lifetime are more straightforward than if permanent nonresorbable materials are used. However, given the relative ease of materials design that researchers enjoy today afforded by numerous advances in synthesis and fabrication, the diversity of materials used in medical applications is significantly less than may be expected.

The use of degradable polymers for medical applications is not a new concept.

First reported in the use of resorbable sutures1 in 1974, poly(lactic acid) (PLA) and subsequently poly(glycolic acid) (PGA), poly(caprolactone) (PCL) and their respective copolymers such as poly(lactic-co-glycolic acid) (PLGA) have been used in applications ranging from solid implantable devices ranging from resorbable bone screws to injectable devices for drug elution.2,3 Despite known limitations, over the past 30 years or more, this small library of materials has dominated the market in products that have gained regulatory approval for commercial use in humans. Considering the advances in polymer science, it is surprising that the diversity of biodegradable materials approved for use is not greater. In principle, the design process for new materials is straight forward; monomer diversity, stoichiometry, architecture and topology are each able to tailor the macroscopic properties of a material to meet targeted performance requirements for a given application. From the perspective of a materials chemist, the design space is limited only by creativity, time, and resources and hence, surely there are materials better suited to in vivo applications that remain to be discovered or translated to clinical use.

2 The barrier commercially and therefore clinically is risk management and the very low level of risk that, not unreasonably, a pharmaceutical company or device manufacturer is willing to tolerate to bring a new degradable polymer system through the regulatory process.

There is an overwhelming preference to use degradable polymers that have been used in other approved applications when pursuing new devices or drug delivery applications. In the rigorous review processes of the US Food and Drug Administration

(FDA), and other regulatory organizations, it is important to understand that no degradable polymer has blanket approval for use in humans. Instead, implantable materials are reviewed as a part of a drug, device or combination product for a specific application. Notably, the regulatory approval process for devices has also evolved significantly over time. Repeated use, approval and favorable clinical outcomes have led to some materials being generally regarded as safe (GRAS). However, a widespread misconception is that devices that use GRAS materials will avoid the normal regulatory review process and scrutiny for safety and efficacy on account of their extensive use. Each device is reviewed independently for a specific application. Using GRAS materials lowers some aspects of the risk barrier in the approval process of a drug or device but does not entirely remove them from regulatory scrutiny.

PLA and PLGA were first used in devices with preamendment status, which refers to devices marketed in the US before May 28, 1976 that have not been significantly changed or modified and for which a regulation requiring a premarket approval

3 application has not been published by the FDA. For example, PLGA was used in degradable sutures marketed as Dexon® and Vicryl®, which were designed to degrade over weeks and in which the mechanical properties of the material were not significantly critical to their success. More applications have subsequently demanded a wider range of physical properties and degradation rates. For example, PLGA Inion Hexalon™ bone screws are often still present after 12 months whereas PLGA-derived Ozurdex™ used as intravitreal implants, degrades over 3 months and has had many complications.4

These early successes of PLA and PLGA in resorbable implants led to industry investigations using these materials across a number of device and disease indications because of the commercial availability of current good manufacturing practice (cGMP) grade materials. Drug and device companies are keen to source cGMP materials to protect their downstream innovations, yet, crucially for the development of new

Figure 1.1. A timeline of drugs and devices containing PLGA that have been approved for use in

humans by the FDA. This list, while not comprehensive, illustrates the diverse applications that PLGA has been utilized within over the past several decades.5–13

4 degradable polymers in this field, it is prohibitively expensive for materials companies to protect their innovations and commercialize cGMP materials without dedicated customers. As if to accentuate the problem, the FDA established the device master file

(MAF) system that protects the proprietary data of material manufacturers (including polymers) while facilitating regulatory review. Although device manufacturers do not have direct access to this information, an applicant sourcing polymer from an entity is able to rely on the MAF file for their own device submission. As such, degradable polymers with established MAFs have a significant commercial advantage and as a result are the

“go to” materials for device manufacturers.

Figure 1.2. Mechanical properties ranges (Young’s modulus and Ultimate Tensile Strength) of PLGA and tissues for which regenerative medicine strategies may be desirable on a logarithmic scale.

5 While the PLGA family has found, and will continue to find much success, especially in the area of drug delivery (Figure 1.1), this practice drives companies to focus using these existing and available polyester materials for an ever-increasing range of applications. In many cases however, the materials have mechanical and physical properties that do not adequately fit those of the tissues that they are being designed to treat (Figure 1.2), which from a materials perspective, makes it feel like trying to fit a square peg in a round hole.

To increase the diversity of materials available to the community, the first step is to understand the limitations of the current ‘gold standard’ materials. Then, the key aspects to be considered within a new framework of understanding must be identified.

This advance in our knowledge should facilitate design of new materials, meet application-specific needs and mitigate the risk for new materials. Herein, we propose a framework that is divided into five categories: byproduct toxicity; inflammation response; erosion profile; polymer crystallinity; and elasticity (Figure 1.3A-B). Here, we highlight these categories, discuss the state-of-the-art materials as well as opportunities for innovation. While each tissue regeneration niche will have specific needs, thus meaning there is no single ‘ideal’ material, the family of materials that are required need to display the same features to present the step-change to PLGA that is required to justify bridging the regulatory gap between materials innovation and translation. There will of course be additional application specific considerations that will require further editing of the material, however the fundamental materials requirements will have been met.

6 Figure 1.3. (A) A spider diagram to outline a framework of properties that an ideal biomaterial would have and rating scales for the properties. (B) The ratings within this framework for several traditional and contemporary materials as examples (PLGA, poly(ester urea)s, aliphatic polycarbonates and thermoplastic polyurethanes, TPUs). An ideal material would simultaneously possess the five categories, in addition to having tunable properties across the range to meet a desired application.

The goal of a resorbable material is to serve its purpose for as long as it is required.

During the degradation and resorption process, it is critical that the local and systemic concentrations of the byproducts are controlled to avoid inflammation and toxicity.

Ideally, resorbable materials leave behind near-native tissue in contrast to nonresorbable materials which are often prone to encapsulation and prolonged inflammatory responses.

In an ideal world, a degradable polymer should have non-toxic byproducts that are readily solubilized so that they can be metabolized or excreted, and do not initiate inflammatory responses or systemic toxicity upon degradation (Figure 3A). In reality, all implanted 7 materials will elicit some foreign body response, however the goal should be to find a balance between degradation, resorption and healing over the lifetime of the implant while it is required to perform a specific task.

Byproduct toxicity and inflammation response are inextricably linked. An immediate consideration for a resorbable polymer is the toxicity of the degradation products, including their relative acidity and terminal oxidation products, which can largely be predicted based on chemical structure. What is often overlooked however, is how the degradation rate of the material will determine local and systemic concentrations. This relationship is very clearly exemplified in the degradation of PLGA.

The primary mode of degradation for PLGA is hydrolysis which releases lactic and glycolic acid. Both compounds are oxidized in the body and lactic acid is readily metabolized via the Cori cycle. Conversely, the terminal oxidation product of glycolic acid is oxalic acid, which is the same product resulting from ethylene glycol oxidation in vivo. However, PLGA has found acceptance in high surface area, small volume devices, such as for drug delivery and sutures in which its slow degradation in vivo leads to systemic concentrations being low at any one time. As such, the biodistribution, bioaccumulation and excretion pathways are each essential in addition to the chemical nature of the degradation products in the evaluation of a new resorbable biomaterial. Although the overall systemic concentration of the degradation products of PLGA may not be cytotoxic, the byproducts are prone to facilitating local acidic environments, termed acidolysis, which does cause local inflammation. This has limited the application of PLGA to small volume devices such

8 as microspheres for drug delivery where it has excelled as the degradation process does not generate sufficient concentrations of residuals to initiate systemic toxicity and inflammation.14

The degradation profile of a material is critical to its performance in vivo and is essentially defined by the erosion mechanism and crystallinity of a material.15 Swelling, water transport, and degradation occur over several steps depending on the chemistry, crystallinity and topology of a resorbable material in vivo. In an implantable device, its success hinges on a delicate balance between degradation profile and tissue remodeling.

If degradation occurs too quickly then the device may not meet the mechanical requirements of the application, and if degradation occurs too slowly it can inhibit rapid tissue remodeling and healing.

Materials have the potential to be deceiving upon initial analysis in vivo.

Degradation products may appear benign with limited inflammation and early mechanical properties may be promising. However, for many applications, these positive traits are negated if the ultimate degradation profile follows a bulk erosion mechanism. Bulk erosion occurs when water penetration into the material is faster than hydrolytic degradation.16 This leads to anisotropic degradation throughout the material which makes changes in local acidity and the mechanical properties difficult to predict. Such a profile is exacerbated when the material is semi-crystalline because crystalline and amorphous domains possess wildly different water transport properties that depend on both domain size and volume fraction of the domains. The combination of bulk erosion

9 profiles and crystallinity lead to loss of mechanical properties throughout the degradation process as well as a burst release of degradation byproducts which, at best, make device failure challenging to predict (particularly for load bearing applications) but, at worst, preclude the material from use. Crystallinity also affects the shelf life of a material with aging leading to altered physical properties including shifts to higher glass transition temperatures, increased moduli values and alterations in drug distribution. While the use of racemic PDLLA copolymer greatly suppresses challenges associated with crystallization, the mechanical properties are consequently drastically weakened.

Again, these characteristics that epitomize PLGA and its analogues. Degradable polyesters are susceptible to degradation through several mechanisms: hydrolysis

(cleavage from water at the ester moiety), enzymatic (bond scission driven by resident enzymes), oxidative (inflammatory response resulting in peroxide radical generation), and mechanical (breakdown from stresses and strains exerted on the material).16 Of these degradation mechanisms, materials that primarily degrade through enzymatic and oxidative pathways present a challenge to use in a device because these two degradation rates will likely vary from patient to patient which makes a consistent degradation profile hard to observe.17 PLGA is prone to bulk degradation which leads to a non-linear loss in mechanical properties and a residual scaffold which limits prolonged tissue regeneration.

As noted above, the slow rate of hydrolysis of PLGA is critical to avoiding inflammation and while the ratio of lactic to glycolic acid can be controlled to increase the ultimate degradation rate (compared to either homopolymer) through limiting homocrystallite

10 formation, attempts to slow the degradation rate of PLGA have been limited. Typically, these have focused on altering the crystallinity of the polymer, incorporating additives, or a combination of the two approaches that have limited water uptake and in turn, slowed material degradation.18 As may be expected from its semi-crystalline nature, the ambient temperature storage and shelf life of PLGA derivatives are complicated by ageing of the material which leads to properties that change over time to differ from the initially manufactured material. Anisotropy in biomaterials is not desirable and therefore, designing an amorphous material that can undergo a controlled and linear degradation in vivo should be a target. While we note that all materials display a balance between bulk and surface erosion (the latter is typically achieved by having high hydrophobicity to prevent water ingress into the polymer), surface eroding materials enable device predictability from fabrication to resorption.

The available materials are largely brittle (for example, PLGA has a Young’s modulus, E, ~ 1 to 3 GPa) and prone to permanent deformation or are soft, highly swollen hydrogel networks. These materials certainly have their own niches, indeed there are some such as drug delivery where this is not of significant importance. There are however a large number of applications where less brittle, more flexible materials are required

(Figure 2) for example, regeneration of skin, vascular, esophageal, cardiac, ligament, tendon and cartilage tissues in which relaxation and fatigue of the material is required to be predictable. There is no shortage of elastic materials like natural rubber, polydimethylsiloxane, and thermoplastic polyurethanes which elicit the desired

11 mechanical properties. However, these polymers are unable to degrade in a meaningful timeframe under physiological conditions, if at all.

The major innovation space is to find materials that are concomitantly elastic and degradable. Although thermoplastic polyurethanes with degradable ‘soft’ blocks, PCL and

PLGA, have been widely studied, the gains in elastic behavior are still offset by the observation of bulk erosion profiles and anisotropic degradation caused by the phase separation between the hard and soft blocks of the materials.18 Creative strategies that look to use these properties concomitantly are required. Drawing on inspiration from both groups of materials might be the key – for example, coupling of the elastic component of a natural rubber derivative polymer with a degradable functional group such as a peptide, oligonucleotide or protein could unlock the door to elastic degradables.

There is definitely space to be creative and explore strategies to couple previously orthogonal properties.

A number of degradable polymers have emerged, primarily in academic laboratories with the aim of overcoming some,19,20 or all, of the problems highlighted with

PLGA and related polyesters.16,21 There clearly remains scope for the development of the polyester family. This is perhaps most notable in the area of poly(propylene fumarate)

(PPF) and polydioxanone (PDO).19,20,22 While PDO is largely amorphous, the ether linkages that are in the backbone lead to an undesirable swelling profile and bulk erosion. PDO has been used widely in high surface area materials such as sutures and electrospun meshes and more recently in exploratory vascular studies.22 It has been widely studied as a

12 biomaterial and if innovative chemical strategies can be applied to increase the range of mechanical properties that are available as well as control its degradation and swelling, it could become a platform degradable biomaterial.23 PPF is a glassy but amorphous polymer on account of the regioirregularity of the 2-propyl unit and the presence of double bonds from the fumarate unit in the polymeric backbone. Significantly, the double bonds in the backbone are reactive towards radicals and hence can be crosslinked thermally or photochemically. Crosslinked PPF-based materials show slow water ingress thus leading to surface erosion behaviour. Despite these beneficial characteristics, crosslinked PPF lacks elasticity and hence only limited applications are possible. Recent chemical advances to more readily enable its synthesis have resulted in commercialization of this promising polymer.

Stepping away from polyester-based materials, polyphosphoesters are being examined as degradable biomaterials based around the enormous synthetic versatility that is readily available in their design and their increased resistance to hydrolytic degradation compared to polyesters. However, the release of phosphate and highly acidic phosphoric acid byproducts will most likely limit their application as bulk materials in larger medical devices. The considerable advances in their synthesis and the flexibility with regard to functional group incorporation make them appealing to look at again in the context of smaller devices and targeted drug delivery. Similarly, polyphosphazenes have great synthetic versatility and have been studied for a wide range of (bio)materials applications.24 By contrast, their degradation products are milder (phosphates and

13 ammonia) and some materials have shown good in vivo biocompatibility. Typically, these materials are slow to degrade in mildly basic conditions and hence achieving control of degradability as well as increasing the elasticity of the materials are areas of future development. The full extent of their chemical potential has not yet been realized and hence polyphosphazenes are promising candidates for further development to make them both more easily accessible and to advance the control over their properties.

Some of the most promising candidates widely studied to date are poly(ester urea)s and aliphatic polycarbonates which possess more diverse physico-chemical properties than poly(esters). Poly(ester urea)s are a diverse library of amino acid-based polymers and copolymers that in turn enables them to have mechanical and degradation properties that are tuneable over a wide range.25,26 A few variants are semi-crystalline, but in general polymers in this family are amorphous, lack acidic byproducts during resorption and possess shape memory properties. These properties are useful for many, but not all, potential devices in which resorbable polymers are beneficial and hence tailoring the materials design to areas of potential need will no doubt unlock useful approaches to deliver new medical devices. Aliphatic polycarbonates are also amorphous materials that enable significant versatility in their chemical design that has been exploited to incorporate a wide range of functional groups with degradation products that are largely benign (although depend on the polymer functionality). Moreover, the most commonly applied example, poly(trimethylene carbonate), displays a surface erosion profile and is more elastic than PLGA, although also degrades much more slowly (Figure

14 1D).27,28 Aliphatic polycarbonates do however, with a few exceptions, remain materials with low glass transition temperatures and thus lack bulk mechanical properties that are suitable for building medical devices. To some extent, this can be overcome through crosslinking and the further development of these polymers could readily yield materials that are suitable for further study.

Other leading candidates are polyanhydrides, polyorthoesters and polyacetals.21,29–31 While polyanhydrides and polyacetals have had a large amount of study devoted to them, to date their translation has been limited. They are all largely amorphous and display highly desirable surface erosion profiles and arguably remain underdeveloped but the low toxicity, in many cases, of their degradation byproducts would make them worthy of expanded study in order to develop them as versatile materials platforms.21 Perhaps one of the main limitations of these materials is their slow degradation in basic conditions, these groups degrade faster when catalyzed by acid.

Additionally, and especially with polyorthoesters, they are also typically very challenging to access synthetically and the diversity of potential monomers and functional groups is relatively small. In turn, this not only limits the number of laboratories and commercial suppliers that are able/willing to provide them in sufficient quantity for large scale biological studies but limits the control of structural alterations to improve their less desirable properties such as swelling and poor bulk mechanical properties. While there has been a significant amount of study with polyanhydrides in particular, there remain significant hurdles to provide high quality materials with predictable and controllable

15 mechanical and physical properties that are suitable for translation across a wider spectrum. Notably, some recent progress in the synthesis of polyorthoesters has been made,31 but much more study of the synthetic routes to these classes of material is required to increase the molecular weights, diversify chemical structure and hence control a range of properties to unlock their full potential.

The need for innovation is clear - tomorrow’s problems will not be solved by yesterday’s materials. Degradable polyester materials, such as PLGA, remain commercially relevant and will continue to find new application areas. However, significant gaps in the physical, biological and chemical properties will continue to limit innovation in medical devices, drug delivery and regenerative medicine. Of course, the limited tolerance for risk by industry will remain a major limitation to the implementation of a new material in a delivery system or medical device. If a family of material other than

PLGA wasn’t the accepted standard, it seems unlikely that, given its limitations, it would be developed and relied upon as it is now. Whatever had been that breakthrough material would remain in that position instead. It is very difficult for a company to pursue new materials and new devices or applications simultaneously. Repeatedly, most choose to use existing materials and pursue novel applications as this presents the best return on investment in an acceptable time scale. The necessary, if overbearing, regulatory approval process that limits the risk tolerance of device companies is rightly focused more on safety and less on effectiveness. Bridging the innovation gap to bring a new material to a device is both time consuming, expensive and requires multidisciplinary study to

16 design and prove its potential worth. Critically, as we have noted earlier, the approval of devices and not materials means that a new material must be tailored to a device before being advanced through the approval process, adding yet more cost and time. New solutions must provide a sufficient step change in behavior to warrant the investment required to go out of the academic lab and produce large scale materials by cGMP methodologies to really further the claims in clinical trials.

The solution to this problem requires innovation in materials science performed with a specific focus on solving a medical problem. All too often, especially in academic laboratories, new materials are synthesized because they can be, rather than by design with an end goal or application in mind. Although this free innovation is essential to defining new materials space, overcoming the regulatory roadblock will only be achieved by application-driven innovation for unmet medical needs. In itself, this will only be achieved with an open-minded multidisciplinary approach to solving unmet medical needs that includes removing preconceptions based on the failures of previous iterations of materials that have failed to have impact in the field. It is likely that through the standard route of academic research, such change will not occur. Instead, disruptive start-up companies will be needed to advance these innovations and begin to establish supply routes and regulatory approval for devices that contain new materials so that they will be subsequently studied more widely. There is still more work to be done, and much is still unknown for many of these applications. This research area can be a rewarding

17 endeavor, with such an open landscape, a novel new material has the potential to be the next PLGA of our time.

1.2. The role of polymers in the evolution of hernia-repair

Over the past two centuries, medicine and technology have coupled together to account for an increase in the quality of care and treatment available. Medicine has advanced from improved mechanistic understanding of the chemistry involved with the human body and how it can interact with emerging technologies.34,35 Diseases, ailments, and injuries that were once considered fatal, can now through surgery, utilize materials to augment and prevent loss of form and function.35–38 There are thousands of examples ranging from open heart surgery to immunomodulatory therapy that show how technology has improved health and wellbeing; however, the major advancements in hernia-repair pale in comparison.35–40 This physical condition dating back centuries still provides challenges to surgical units today. This introduction seeks to focus on the evolution of surgical materials and techniques associated with hernia-repair.

Herniation of soft-tissue can occur in a variety of ways including but not limited to, cyclic fatigue, blunt force trauma, surgical incision, and/or genetic predisposition.41

The given structural weakness or tear allows for inner organs or tissues to migrate away from their designed locations. Hernias are divided in to categories based on the anatomical position where they occur. Two of the most common hernias include inguinal

(groin) and ventral (abdominal).39,41 In ventral hernias, there are several subclassifications varying on location which include umbilical (navel), incisional (location

18 of surgical incision), and epigastric (upper-central abdomen). Inguinal hernias are most commonly associated as a chronic repetitive sports injury.42 While the type of hernia observed can vary by location, the current treatment methods tend to be even more diverse. Surgical techniques range from open surgery to minimally invasive techniques like laparoscopic and robotic hernia-repair.43–46 The reason for such variation is likely attributed to no clear consensus on treatment management. While there are many qualified reviews that discuss a variety of surgical techniques for hernia-repair, this intro will focus on improving hernia-repair from a materials point of view.

To understand how current materials used for hernia repair have come to use, it is important to understand the history of hernia diagnosis and treatment. The classic hernia and related symptoms from soft-tissue damage is not new with the first documented hernia injury dating back to the Egyptians.39,43 Diagnostic techniques for herniation are often times hands on and predominantly visual which enables this class of injury to be readily diagnosed without sophisticated imaging techniques. The rudimentary diagnostics are undoubtedly why hernias were identifiable, however, no major progress was made in treating the structural defect before aseptic technique and sterile surgical procedures were heavily adopted. Surgery on hernia defects was largely limited to tension techniques which involved direct suture administration to pull the defect back together.47,48 The tension from direct suturing creates a distortion of the surrounding muscle and tissues which can lead to patient pain and discomfort.

Additionally, the defect site is contracted compared to its native shape which lends itself

19 to further tears. While tension techniques work well for many skin applications, these techniques were precluded from finding success in hernia-repair because of unreasonably high recurrence rates. A transition to tension free techniques was first pioneered in the early 20th century and later perfected in the late 20th century with the use of synthetic materials.36,47,49 A polymer material could be used as a patch where it simply covered the defect and provided mechanical support without the need for the muscles to be pulled back together. Upon the introduction of tension free techniques, the recurrence rates were reduced after 5 years of initial hernia surgery with some rates being reported at less than 5%.40,43,50,51 While this technique was a straightforward improvement for simple hernia-repair cases, a large majority of hernia-repairs are neither simple nor straightforward. Complex hernia-repair is almost exclusively associated with a decrease in quality of life and recurrent surgical complications.52 Many complex hernia repair cases propagate from surgical incision; as there is not expected to be a suppression in the number of abdominal surgeries, it becomes important to find an effective treatment method for consequential hernia cases.

With the introduction of polymer materials indicated for the treatment of herniated tissue, a significant amount of time and research has been conducted to design the best hernia-repair device. The implants employed can primarily be varied based on material composition (permanent or temporary) and device design (mesh, film, weave, etc.). The dominant hernia implants initially were nonresorbable materials like polypropylene and polytetrafluoroethylene.53,54 These materials had robust mechanical

20 properties to be able to handle the inner abdominal pressures throughout the healing process and were found to significantly reduce the stresses placed on the muscles surrounding the herniated site. As the understanding of tissue remodeling and healing cascade has improved, the device design for these materials has been altered over the past 40 years to improve upon tissue ingrowth.55–57 The most commonly used design is a lightweight mesh as it is largely porous, thus limiting the amount of material implanted that could elicit a foreign body response while also facilitating facile tissue integration at the repair site.58,59 Over time, these factors have led to polypropylene being the go-to material with the mesh design being preferred for most ventral hernia applications.

If short-term recurrence rates (< 5 years) of hernia-repair were all that were understudy, then polypropylene meshes would likely not require replacement for this common procedure; however, this is far from the case. Since the first implantation of a tension free technique material, there has been the ongoing challenge that the implanted material is nonresorbable. While there are many applications where a nonresorbable material would be ideal, this is often times only the case when the mechanical properties of the material and implant site tissue are a match.60,61 For any implanted material, there is a buildup of fibrous scar tissue as inflammatory cells attempt to engulf and break down an implanted material through phagocytosis and release of acidic granules.62,63 As polypropylene is impervious to this attempted assault, this fibrous capsule scar tissue will persist and can leave the surrounding tissue in a prolonged state of inflammation.64 As scar tissue is mechanically weaker than native healthy tissue, and as the material is

21 located at a weakened area, recurrence and complications are possible for a patient who does not experience recurrence in the first several years. When reoccurrence occurs, a second surgery is required and thus the cycle of material implantation repeats. This vicious cycle propagates with multiple recurring repair surgeries with the amount of scar tissue ever increasing on each iteration.40 After several surgeries, there may not be enough viable tissue to perform additional surgeries. Additionally, the hernia surgery is often as a consequence of other underlying issues which can play a role in limiting a proper healing response.65,66 There have been several different attempts to mitigate this issue with the strategies employed being divided in to two categories: 1) modulation of nonresorbable materials to mitigate negative host-immune response to prevent scar tissue buildup or 2) use of resorbable materials that degrade over a given timeframe.55,67

When modulating a material to limit an immune response, it is important to consider the cellular activity of the implant surface. Any implanted material will interact with blood, plasma, fibrinogen, and a whole host of cells within the first minutes of being implanted; thus, it is important that the extent to which cells respond with inflammation is limited.68–70 A certain level of initial inflammation is unavoidable and thought to be helpful to the healing process as it can kickstart the remodeling process, however it would be ideal for this response to transition from a static state of inflammation towards remodeling, healing, and quiescence.62,63,71–73 A common theme in literature has been to coat polypropylene with an exterior layer that is designed to provide these exact characteristics and mitigate inflammatory issues. The coatings vary with one group

22 employing a titanium coating on the surface which was found to limit the patient discomfort likely attributed to a suppression of adhesion, however there was no significant difference in performance from polypropylene controls.74 Still other research has explored the use of biologically derived coatings on polypropylene to facilitate a more compatible cellular response. It has been observed that collagen coatings can improve cell attachment and adhesion for fibroblasts while not reducing the mechanical stability of the underlying polymer.75 This response has been further improved upon with the clever addition of extracellular matrix (ECM) on the surface which is composed of collagen, proteoglycans, cellular transmembrane proteins, and other cell signaling factors. Many ECM coated meshes display a reduction in foreign body response indicated by a suppression of multinucleated giant cells found at the implant site.55,75 While these efforts are to be applauded and certainly are a step in the right direction, the coated polypropylene surface is often readily degraded in vivo with degradation occurring in some coated meshes as quickly as 35 days.76 While the immune response has successfully been modulated over this abbreviated timeframe, when considering the lifetime of the implant is likely years and even decades, this modulation feels more like a small step rather than a huge stride.

At the end of the day, it seems like there is a material selection issue with nonresorbable implants for long-term hernia applications. The surface modulatory design processes feel much like a stop gap rather than a permanent solution. An alternative strategy that might prove fruitful is to explore resorbable materials for hernia-

23 repair. When considering common analgesic drug administration, it is common place for drug to be administered orally, broken down in the body, provide a desired response with minimal side effects, and then readily excreted over a desired timeframe. While the parallels for drug administration and large biomaterial implants may not be perfectly in line, the design criteria for a gold standard hernia-repair material overlap nicely.

Resorbable materials would be precluded from a long-term foreign body response as the design process could allow for mitigation of the host-immune response over the short term while also utilizing degradation to promote complete tissue integration and repair.

If the mechanical properties were properly tuned, then a resorbable material would provide support until it is excreted. Several groups have attempted this line of thinking and have turned to two main classes of degradable materials: 1) biologically derived materials or 2) biologically degradable synthetic polymers. Biologically derived materials like small intestine submucosa extracellular matrix (SIS-ECM) have clear advantages in cellular activity as was previously observed with the coated materials. These materials have gained clinical use in certain hernia applications where robust mechanical properties are less necessary.77 These materials are limited by failure at low strain (< 10%) but match the mechanical properties of soft-tissue better than previously used nonresorbable synthetics.78 One challenge with SIS-ECM and other biologics is that there is a necessary decellularization process prior to implantation to ensure that there is not a serious host- immune response from a piece of tissue from a foreign species.77,79 This cost does not preclude these materials from coming to market however, it does carry an added step

24 which makes them potentially more expensive to manufacture than an optimized synthetic polymer.

There have been several classes of degradable synthetic materials that utilize ester degradable moieties in the backbone. Degradation rates can be controlled based on several factors including crystallinity, polymer side chain, and additive composition.

One material that is not PLGA, PCL, or derivatives thereof is poly-4-hydroxybutyrate

(P4HB). This material was approved for use as a thermoplastic degradable suture in 2007 and has since been indicated for its use in hernia-repair.67 It maintains mechanical properties similar to that of currently used polypropylene materials and is readily

67,80 degraded through the Kreb’s cycle being processed in to CO2 and water. Other materials of interest are amino acid-based degradable synthetics. These materials are of interest as their degradable byproducts are amino-acids which are known to be nontoxic.81–84 Two common synthetics are amino acid-based polycarbonates and amino acid-based poly(ester urea)s. While polycarbonates have been more fully explored over a range of applications, poly(ester urea)s could prove advantageous for hernia-repair applications. Polycarbonates tend to degrade slowly under physiological conditions which wouldn’t promote an optimized tissue regeneration rate.80,82 Poly(ester urea)s can be tuned to have mechanical properties similar to that of soft tissue depending on the diol chain length and amino acid selected during synthesis.79,85–87 Similarly, their degradation times can vary based on chemical structure (i.e. amino acid and diol selected) and fabrication method (i.e. compression molding, solvent casting, extrusion)

25 selected.79,88 The synthesis, characterization, cellular viability, fabrication, and in vivo implantation of poly(ester urea)s were explored for their use as materials to aid in the treatment of hernia-repair.

26

CHAPTER II

MATERIALS AND INSTRUMENTATION

2.1. Materials

1,6-hexanediol, 1,8-octanediol, 1,10-decanediol, 1,12-dodecanediol, triphosgene, sodium carbonate, 1,1,1-tris(hydroxymethyl)ethane, 3-allyloxy-1,2-propanediol, 1,3- diisopropyl carbodiimide, and p-toluenesulfonic acid monohydrate were purchased from

Sigma Aldrich (Milwaukee, WI). Toluene, chloroform, reagent grade alcohol, eosin Y, hematoxylin, FITC-PEG-thiol (1000 MW), fibrinogen (from human plasma), ethyl acetate, hexane, acetone, and N,N-dimethylformamide were purchased from Fisher Scientific

(Pittsburgh, PA). Boc-o-benzyl tyrosine, L-valine, and L-phenylalanine were purchased from Acros (Pittsburgh, PA) and Bachem (Torrance, CA) respectively. The commercially available 2.0 1-1 LL SIS-ECM (Single-layer lyophilized ECM) was provided by Cook Biotech and used as provided. All solvents were reagent grade and all chemicals were used without further purification unless otherwise noted.

2.2. Instrumentation

1H-NMR and 13C-NMR spectra were collected using a 300 MHz and 500 MHz Varian

NMR spectrophotometer, respectively. Chemical shifts are reported in ppm (δ) and referenced to DMSO-d6 and D2O (2.50 ppm and 4.79 ppm respectively). Multiplicities

27 were explained using the following abbreviations: s = singlet, d = doublet, t = triplet, br = broad singlet, and m = multiplet.

Size exclusion chromatography (SEC) was performed using an EcoSEC HLC-

8320GPC (Tosoh Bioscience, LLC) equipped with a TSKgel SuperH-RC 6.0 mm ID. × 15 cm mixed bed column and refractive index (RI) detector. The number average molecular mass (Mn), weight average molecular mass (Mw), and molecular mass distribution (Đm) for each sample was calculated using a calibration curve determined from linear polystyrene standards (PStQuick MP-M standards, Tosoh Bioscience, LLC) with either 0.01 M LiBr in

DMF as the eluent flowing 1.0 mL/min at 50 °C or in THF as the eluent flowing 0.35 mL/min at 35-50 °C.

Differential scanning calorimetry (DSC) was performed using a TA Q2000 with heating and cooling cycle ramps of 10 °C/min in the temperature range of 0-100 °C. The glass transition temperature (Tg) was determined from the midpoint of the second heating cycle endotherm.

Thermogravimetric analysis (TGA) was performed using a TA Q500 with heating ramps of 20 °C/min in the temperature range from 0-500 °C/min. The degradation temperature (Td) was determined from 10% mass loss.

Surface topology images were obtained using scanning electron microscopy

(SEM). Using a JEOL USA SEM, samples were sputter-coated with gold and scanned at 2.0 kV excitation between 50-750 × magnification.

28

Histology images were obtained using a Keyence BZ-X700 between 2-20 × magnification under brightfield conditions.

A Rame-Hart Contact Angle Goniometer was used to determine contact angle for

5 µL of deionized water on a sample surface.

The XPS spectra were obtained using a VersaProbe II Scanning XPS Microprobe from Physical Electronics (PHI), under ultrahigh vacuum conditions with a pressure of 2.0

µPa. Automated dual beam charge neutralization was used during the analysis of the samples to provide accurate data. The analyzer pass energy was 117.4 eV for the survey spectra and 23.5 eV for the high-resolution scans in the N1s and S2p regions. The XPS high resolution spectra of N1s were decomposed into two components using the curve fitting routine in PHI MultiPak. Each spectrum was collected using a monochromatic (Al Kα) x- ray beam (E = 1486.6 eV) over a 100 μm × 1400 μm probing area with a beam power of

100 W.

Quartz crystal microbalance with dissipation (QCM-D) was used to obtain adsorption profiles (Qsense E4, Biolin Scientific AB). Samples were spin coated onto QCM chips (Qsense, 335 SiO2) using a 1 % (wt/wt) solution of polymer in acetone at 8000 rpm for 1 min. QCM profiles were obtained with 1 × PBS and fibrinogen (1.5 mg/mL) with solutions flowed at 0.150 mL/min at 37 °C. Adsorption to the surface was calculated from the 5th overtone using the Sauerbrey equation.

29

CHAPTER III

PRECLINICAL IN VITRO AND INVIVO ASSESSMENT OF LINEAR AND BRANCHED L-VALINE-

BASED POLY(ESTER UREA)S FOR SOFT TISSUE APPLICATIONS

In part this work has been reprinted with permission from Dreger, N. Z.; Wandel,

M. B.; Robinson, L. L.; Luong, D.; Søndergaard, C. S.; Hiles, M.; Premanandan, C.; Becker,

M. L., Preclinical in Vitro and in Vivo Assessment of Linear and Branched L-Valine-Based

Poly(ester urea)s for Soft-tissue Applications, ACS Biomater. Sci. Eng., 2018, 4 (4), pp

1346–1356. Copyright 2018 American Chemical Society.

3.1. Abstract

New polymers are needed to address the shortcomings of commercially available materials for soft-tissue repair. Herein, we investigated a series of L-valine-based poly(ester urea)s (PEU)s that vary in monomer composition and the extent of branching as candidate materials for soft-tissue repair. The pre-implantation Young’s moduli (105 ±

30 - 269 ± 12 MPa) for all the PEUs are comparable to polypropylene (165 ± 5 MPa) materials currently employed in hernia-mesh repair. The 2% branched poly(1-VAL-8) maintained the highest Young’s modulus following 3 months of in vivo implantation (78 ±

34 MPa) when compared to other PEU analogs (20 ± 6 – 45 ± 5 MPa). Neither the linear and branched PEUs elicited a significant inflammatory response in vivo as noted by less

30 fibrous capsule formation after 3 months of implantation (80 ± 38 – 103 ± 33 µm) relative to polypropylene controls (126 ± 34 µm). Mechanical degradation in vivo through 3 months, coupled with limited inflammatory response, suggest L-valine based PEUs are translationally-relevant materials for soft-tissue applications.

3.2. Introduction

Synthetic polymers have been used in medical devices for more than 50 years.54,64,89,90 Hernias are one medical malady that employ polymers in device design to facilitate clinical outcomes. A hernia arises from a structural defect in tissue or muscle that allows for organs or tissues to protrude from their natural location. The location of the defect can vary across the body with the most common types occurring at the inner groin (inguinal) and the abdomen wall (ventral).91,92 In the 1800s, sutures were used to close the herniated tissue and unsurprisingly, post-operative recurrence rates were high.93 Polymers have since been utilized to augment the structural defect, which has led to a significant drop in recurrence rates, with some inguinal hernia rates being reported in less than 15% of cases.94 Recurrence rates vary greatly depending on multiple factors including hernia type, surgical complexity, and pre-existing patient conditions.94–96

Despite advances in surgical techniques and higher success rates, much is left to be desired from a material standpoint.

New polymers for hernia-mesh repair are of interest because of unmet needs from current materials. Polypropylene (PP) meshes have been used widely to aid in the treatment of ventral hernias.97 PP meshes provide strong reinforcement to the affected

31 area, which has helped reduce the rate of recurrence from previous surgical methods.98,99

Despite vast improvements from previous surgical techniques, the rigidity of PP promotes the deposition of dense, fibrous scar-tissue which is foreign to the injury site.100,101

Additionally, PP does not degrade, which leaves the implant permanently in the patient.

Long term complications include shrinkage, erosion and risk of infection, which may occur several years after implantation.102 Therefore, the immediate recurrence prevention that this material provides comes at the cost of long term comfort and structural integrity of the surgical site.

The unfavorable long-term performance of PP has prompted research into alternative materials that address these problems. Synthetic polymers currently used clinically include PP, poly(ε-caprolactone) (PεCL), polylactides (PLA), polyvinylidene fluoride (PVDF), poly(glycolic acid) (PGA), polyurethanes (PU), and copolymers thereof.56,57,64,103–106 Regardless of the material chosen for study, there are several ideal design criteria a material should address to meet the needs for a hernia injury: limit inflammatory response, sufficient reinforcement to the affected area, promote native tissue regeneration at the wound site, and if possible degrade over time to prevent recurrence and patient discomfort.90,100,104,107

Unlike PP, homopolymers and copolymers consisting of PεCL, PLA, and PGA are degradable, and in general, they maintain mechanical properties that are comparable to

PP at the time of implantation.56,57,103,105,108 Despite improved processability and degradation rates, the acidic byproducts from degradation of the ester bond often

32 promote an undesired inflammatory response at the surgical site.109 To mitigate these issues, biologic meshes of xenogeneic and allogeneic materials have been developed

(porcine, human skin grafts, etc.)with variable success rates.66,110,111 Xenogeneic and allogeneic materials have each found clinical utility.112–115 Extracellular matrix (ECM) materials promote healing at the surgical site with limited prolonged inflammatory response while being less prone to infection and erosion. However, the mechanical properties deteriorate quickly in vivo which carries the risk of hernia recurrence if tissue regeneration and mesh resorption is not adequately balanced.100,113,115 ECM materials also are expensive to manufacture compared to synthetic meshes thereby placing the former at a disadvantage.

Amino acid-based poly(ester urea)s (PEU)s are a novel class of materials that we are targeting for hernia repair, as they have previously demonstrated tunable mechanical properties and degradation rates, while eliciting limited inflammatory response in vivo.26,116,117 Properties of α-amino acid-based PEUs are tuned based on monomer diol chain length, amino acid, and extent of branching.84,87 Degradation byproducts of α- amino acid-based PEUs have been previously shown to have no observable local inflammatory response.85 PEUs can be broken down in to readily excretable small molecules which could include short chain diols, amino acids, and urea byproducts. The in vivo degradation mechanism of action could be attributed to hydrolysis, enzymatic cleavage, or mechanical scission. In this study, we sought to investigate l-valine based

PEUs as hernia repair materials in a preclinical rat model. Herein we describe the

33 synthesis and characterization of linear and branched PEUs. The potential application of these materials for soft tissue repair applications was then investigated through in vivo degradation and immune response studies.

3.3. Experimental

3.3.1. Materials

1,8-octanediol, 1,10-decanediol, 1,12-dodecanediol, triphosgene, sodium carbonate, 1,1,1-tris(hydroxymethyl)ethane and p-toluenesulfonic acid monohydrate were purchased from Sigma Aldrich (Milwaukee, WI). Toluene, chloroform, and N,N- dimethylformamide were purchased from Fisher Scientific (Pittsburgh, PA). Boc-o-benzyl tyrosine and l-valine were purchased from Acros (Pittsburgh, PA) and Bachem (Torrance,

CA) respectively. All solvents were reagent grade and all chemicals were used without further purification unless otherwise noted.

3.3.2. Characterization

1H NMR spectra were collected using a 300 MHz Varian NMR spectrometer.

Chemical shifts are reported in ppm (δ) and referenced to DMSO-d6 (2.50 ppm).

Multiplicities were explained using the following abbreviations: s = singlet, d = doublet, t

= triplet, br = broad singlet, and m = multiplet. Size exclusion chromatography (SEC) was performed using an EcoSEC HLC-8320GPC (Tosoh Bioscience, LLC) equipped with a TSKgel

SuperH-RC 6.0 mm ID. × 15 cm mixed bed column and refractive index (RI) detector. The number average molecular mass (Mn), weight average molecular mass (Mw), and molecular mass distribution (Đm) for each sample was calculated using a calibration curve

34 determined from linear polystyrene standards (PStQuick MP-M standards, Tosoh

Bioscience, LLC) with 0.01 M LiBr in DMF as the eluent flowing 1.0 mL/min at 50 °C.

Differential scanning calorimetry (DSC) was performed using a TA Q2000 with heating and cooling cycle ramps of 10 °C/min in the temperature range of 0-100 °C. The glass transition temperature (Tg) was determined from the midpoint of the second heating cycle endotherm. Thermogravimetric analysis (TGA) was performed using a TA Q500 with heating ramps of 20 °C/min in the temperature range from 0-500 °C/min. The degradation temperature (Td) was determined from 10% mass loss. Surface topology images were obtained using scanning electron microscopy (SEM). Using a JEOL USA SEM, samples were sputter-coated with gold and scanned at 2.0 kV excitation at 750 × magnification. Histology images were obtained using a Keyence BZ-X700 at 20 × magnification. Statistical analysis was performed using a Tukey post-hoc ANOVA.

3.3.3. Synthesis of Poly(ester urea) monomers

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-valine)-Octane 1,8-Diester

Monomer (1-VAL-8). Synthesis of di-p-toluenesulfonic acid salts of bis(L-valine)-octane

1,8-diester (1-VAL-8) was carried out following previously published procedures.84

Briefly, 1,8-octanediol (43.8 g, 0.3 mol, 1 eq.), L-valine (73.8 g, 0.63 mol, 2.3 eq.), p- toluenesulfonic acid monohydrate (131.3 g, 0.69 mol, 2.4 eq.), and toluene (1300 mL) were added to a 3 L 3-neck round bottom flask and mixed using a stir bar. A Dean-Stark

Trap was attached to the round bottom flask and the reaction was heated to reflux for 24 h. The reaction was cooled to ambient temperature, and the resulting white precipitate

35 was isolated by vacuum filtration using a Buchner funnel. The product was recrystallized by dissolving in boiling water (2 L), vacuum filtering hot, and cooling to room temperature to afford a white solid precipitate. The precipitate was collected via filtration and the recrystallization process was performed three times to maximize purity (166 g, 79% yield).

1 H NMR (300 MHz, DMSO-d6): δ = 0.95-0.99 (m, 12H, -CH(CH3)2), 1.24-1.35 (s, 8H, -

COOCH2CH2(CH2)4-), 1.55-1.65 (m, 4H, -COOCH2CH2(CH2)4CH2-), 2.06-2.22 (m, 2H,

(CH3)2CH-), 2.26-2.31 (s, 6H, -CH3Ar-), 2.50 (s, DMSO), 3.33-3.38 (s, H2O), 3.88-3.90 (d, J =

+ 4.3 Hz, 2H, NH3CHCOO-), 4.08-4.24 (m, 4H, -COOCH2CH2(CH2)4-), 7.10-7.14 (d, J = 8.2 Hz,

+ 4H, aromatic H ), 7.48-7.50 (d, J = 8.1 Hz, 4H, aromatic H), 8.25-8.33 (br, 6H, -NH3 ).

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-valine)-Decane 1,10-Diester

Monomer. (1-VAL-10). Synthesis of di-p-toluenesulfonic acid salts of bis(L-valine)-decane

1,10-diester (1-VAL-10) was carried out using the method described above (154 g, 71%

1 yield). H NMR (300 MHz, DMSO-d6): δ = 0.93-1.00 (m, 12H, -CH(CH3)2-), 1.22-1.33 (s, 12H,

-COOCH2CH2(CH2)6-), 1.55-1.64 (m, 4H, -COOCH2CH2(CH2)4CH2-), 2.09-2.21 (m, 2H,

(CH3)2CH-), 2.28-2.31 (s, 6H, -CH3Ar-), 2.50 (s, DMSO), 3.30-3.35 (s, H2O), 3.87-3.91 (d, J =

+ 4.5 Hz, 2H, NH3CHCOO-), 4.08-4.24 (m, 4H,-COOCH2CH2(CH2)6-), 7.10-7.13 (d, J = 7.8 Hz,

+ 4H, aromatic H), 7.47-7.51 (d, J = 7.8 Hz, 4H, aromatic H), 8.27-8.31 (br, 6H, -NH3 ).

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis-(L-valine)-Dodecane 1,12-

Diester Monomer (1-VAL-12). Synthesis of di-p-toluenesulfonic acid salts of bis(L-valine)- dodecane 1,12-diester (1-VAL-12) was carried out using the method described above (106

1 g, 82% yield). H NMR (300 MHz, DMSO-d6): δ = 0.90-0.98 (m, 12H,-CH(CH3)2), 1.22-1.27

36

(s, 16H, -COOCH2CH2(CH2)8-), 1.53-1.63 (m, 4H, -COOCH2CH2(CH2)8CH2-), 2.07-2.18 (m,

+ 2H, (CH3)2CH -), 2.27-2.29 (s, 6H, -CH3Ar-), 2.50 (s, DMSO), 3.29-3.33 (s, H2O), 3.87-3.90

+ (d, J = 4.3 Hz, 2H, NH3CHCOO-), 4.06-4.22 (m, 4H, -COOCH2CH2(CH2)8-), 7.08-7.11 (d, J =

+ 7.9 Hz, 4H, aromatic H), 7.45-7.49 (d, J = 8.1 Hz, 4H, aromatic H), 8.25-8.28 (br, 6H, -NH3 ).

Synthesis of Hydrochloric Acid Salts of Tri-O-benzyl-L-tyrosine-1,1,1- trimethylethane Triester Monomer (Triol-TYR). Synthesis of hydrochloric acid salts of Tri-

O-benzyl-L-tyrosine-1,1,1-trimethylethane triester monomer (Triol-TYR) was carried out following previously published procedures.87 The branched monomer was formed through the esterification between 1,1,1-tri(hydroxylmethyl)ethane and Boc-O-benzyl-L- tyrosine. In a 500 mL RBF, 1,1,1-tri(hydroxylmethyl)ethane (2.0 g, 16 mmol, 1.0 eq.), Boc-

O-benzyl-L-tyrosine (22.2 g, 60 mmol, 3.75 eq.), and 4-(N,N-dimethylamino)puridinium 4- toluenesulfonate (DPTS, 3.00 g, 10 mmol, 0.6 eq.) were dissolved in a minimum amount of DMF. Once dissolved, the reaction was placed in an ice bath for 10 minutes followed by syringe addition of 1,3-diisopropyl carbodiimide (DIC, 10.14 mL, 80 mmol, 5 eq.). The reaction was allowed to gradually come to ambient temperature while stirring for 24 h, and a yellow precipitate formed. DMF was removed under reduced pressure using a vacuum transfer and the remaining solid was dissolved in a minimal amount of chloroform. The solution was washed (3 ×) with sodium bicarbonate and the organic solution was concentrated for column chromatography purification. Silica gel was used as the stationary phase with hexane/ethyl acetate (4:1 v/v) mobile phase and all fractions were collected for rotary evaporation. The solvent was removed by evaporation and a

37

1 yellow solid was obtained (12.2 g, 73%). H NMR (300 MHz, DMSO-d6): δ = 0.81-0.87 (s,

3H, -CCH3), 1.26-1.30 (s, 27H, CH3 in Boc protecting group), 2.50 (s, DMSO), 2.72-2.94 (m,

+ 6H, -CHCH2-Ar-), 3.92-3.94 (m, 6H, -COOCH2C-), 4.08-4.16 (m, 3H, - NH3CHCOO-), 5.00-

5.03 (s, 6H, -Ar-OCH2-Ar-), 6.86-7.42 (m, 27H, aromatic H). The boc-protected yellow solid was dissolved in HCl/dioxane (4 M) and allowed to stir under nitrogen for 3 h. The yellow

1 solid was freeze-dried to remove solvent (11.5 g, 69%). H NMR (300 MHz, DMSO-d6): δ

= 0.65-0.67 (s, 3H, -CCH3), 2.50 (s, DMSO), 2.98-3.18 (m, 6H, -CHCH2-Ar-), 3.81-3.96 (m,

+ 6H, -COOCH2C-), 4.16-4.22 (m, 3H, - NH3CHCOO-), 5.02-5.04 (s, 6H, -Ar-OCH2-Ar-), 6.91-

+ 7.42 (m, 27H, aromatic H), 8.72-8.78 (br s, 9H, NH3-).

3.3.4. Synthesis of Poly(ester urea) polymers

Synthesis of Linear Poly(ester urea)s. The syntheses of linear poly(ester urea)s were based on previously published procedures.84,87 In short, interfacial polymerization of di- p-toluenesulfonic acid salts of bis(L-valine) monomers 1-VAL8, 1-VAL-10, and 1-VAL-12 was performed by dissolving the desired monomer and sodium carbonate in distilled water (0.1 M) in a 3 L 3-neck round-bottom flask. The solution was placed in a 40 °C water bath with overhead mechanical stirring until clear. The mixture was then cooled to 0 °C.

In a separate container, additional sodium carbonate (1.05 eq.) was dissolved in distilled water and added to the reaction flask and the solution was allowed to stir until clear.

Separately, triphosgene (0.35 eq.) was dissolved in distilled chloroform (0.6 M) and subsequently added to the reaction flask using an addition funnel. The solution turned white immediately and was allowed to stir for 30 minutes. An additional aliquot of

38 triphosgene (0.08 eq.) dissolved in distilled chloroform (0.6 M) was added to solution dropwise (~1 drop/second) using the addition funnel. The reaction was stirred for 3 hours and then transferred to a separatory funnel and washed with water (3 ×). The organic phase was collected and precipitated in hot water to remove impurities. The product was cooled and dried under reduced pressure. The white polymer was thus collected (82-92% yield).

1 Poly(1-VAL-8). H NMR (300 MHz, DMSO-d6): δ = 0.77-0.89 (m, 12H, -CH(CH3)2), 1.24-1.33

(s, 8H, -COOCH2CH2(CH2)4-), 1.50-1.58 (m, 4H, -COOCH2CH2(CH2)4CH2-), 1.94-2.04 (m, 2H,

+ -(CH3)2CHCHNH3 -), 2.50 (s, DMSO), 3.29-3.33 (s, H2O), 3.94-4.10 (m, 6H, -

CHCOOCH2CH2(CH2)4-), 6.37-6.41 (s, 2H, -NH-). (Mw = 71 kDa, Mn = 42 kDa, Đm = 1.7, Tg =

42 °C, Td = 310 °C)

1 Poly(1-VAL-10). H NMR (300 MHz, DMSO-d6): δ = 0.78-0.90 (s, 12H, -CH(CH3)2), 1.20-1.29

(s, 12H, -COOCH2CH2(CH2)6-), 1.49-1.56 (m, 4H, -COOCH2CH2(CH2)6CH2-), 1.91-2.00 (m, 2H,

(CH3)2CH-), 2.50 (s, DMSO), 3.30-3.34 (s, H2O), 3.97-4.11 (m, 6H, -CHCOOCH2CH2(CH2)6-),

6.32-6.42 (s, 2H, -NH-). (Mw = 71 kDa, Mn = 46 kDa, Đm = 1.6, Tg = 34 °C, Td = 339 °C)

1 Poly(1-VAL-12). H NMR (300 MHz, DMSO-d6): δ = 0.81-0.87 (s, 12H, -CH(CH3)2), 1.21-1.27

(s, 17H, -COOCH2CH2(CH2)8-), 1.50-1.56 (m, 4H, -COOCH2CH2(CH2)8CH2-), 1.92-2.05 (m, 2H,

(CH3)2CH-), 2.50 (s, DMSO), 3.28-3.31 (s, H2O), 3.95-4.11 (m, 6H, -CHCOOCH2CH2(CH2)8-),

6.32-6.42 (s, 2H, -NH-). (Mw = 75 kDa, Mn = 51 kDa, Đm = 1.4, Tg = 29 °C, Td = 205 °C)

Synthesis of Branched Poly(ester urea)s. The syntheses of the branched poly(ester urea)s were based on previously published procedures.84,87 In short, interfacial polymerization

39 was performed by dissolving the di-p-toluenesulfonic acid salt of bis(L-valine) monomers

1-VAL-8 or 1-VAL-10 with the hydrochloric acid salt of Triol-TYR in a molar ratio of 98:2 respectively (1.0 eq. in total), as well as sodium carbonate anhydrate (2.1 eq.) in distilled water (0.1 M) in a 3 L 3-neck round bottom flask. The solution was placed in a 40 °C water bath with overhead mechanical stirring until clear. Ice was added to the water bath until the temperature reached 0 °C. Separately, additional sodium carbonate (1.05 eq.) was dissolved in distilled water and the solution was added to the reaction flask and stirred until clear. Triphosgene (0.35 eq.) was dissolved in distilled chloroform (0.6 M) and subsequently added to the reaction flask using an addition funnel. The solution turned white immediately and was allowed to stir for 30 minutes. An additional aliquot of triphosgene (0.08 eq.) dissolved in distilled chloroform (0.6 M) was added to solution dropwise (~1 drop/second) through the addition funnel and stirred for 3 hours before transferring to a separatory funnel. The reaction mixture was washed with water (3 ×) and the organic phase was collected, then recrystallized in hot water. The product was allowed to cool, filtered, and dried under reduced pressure. The white polymer was then collected (79-88% yield).

Branched PEU-2% (Bis(L-valine)-Octane 1,8-Diester Monomer and Tri-O-benzyl-L-tyrosine-

1,1,1-trimethylethane Triester Monomer with a Molar Ratio of 98:2). 1H NMR (300 MHz,

DMSO-d6): δ = 0.80-0.90 (m, 12H -CH(CH3)2), 1.22-1.34 (s, 8H, -COOCH2CH2(CH2)4-), 1.52-

1.58 (m, 4H, -COOCH2CH2(CH2)4CH2-), 1.95-2.02 (m, 2H, (CH3)2CH-), 2.50 (s, DMSO), 3.33-

3.38 (s, Dioxane), 3.98-4.08 (m, 6H, -CHCOOCH2CH2(CH2)4-), 5.00-5.02 (s, -Ar-OCH2-Ar-),

40

6.37-6.42 (d, J=8.9 Hz, 2H, -NH-), 6.88-7.42 (aromatic H, branched monomer ), 8.25-8.33

(s, -NH-, branched monomer). (Mw = 410 kDa, Mn = 126 kDa, Đm = 3.3, Tg = 35 °C, Td = 301

°C)

Branched PEU-2% (Bis(L-valine)-Decane 1,10-Diester Monomer and Tri-O-benzyl-L- tyrosine-1,1,1-trimethylethane Triester Monomer with a Molar Ratio of 98:2). 1H NMR

(300 MHz, DMSO-d6): δ = 0.78-0.90 (m, 12H, -CH(CH3)2), 1.21-1.31 (s, 12H, -

COOCH2CH2(CH2)6-), 1.51-1.58 (m, 4H, -COOCH2CH2(CH2)6CH2-), 1.93-2.04 (m, 2H,

(CH3)2CH-), 2.50 (s, DMSO), 3.29-3.41 (s, Dioxane), 3.97-4.08 (m, 6H, -

CHCOOCH2CH2(CH2)6-), 5.01-5.04 (s, -Ar-OCH2-Ar-), 6.35-6.43 (d, J=8.8 Hz, 2H, -NH-), 6.87-

7.45 (aromatic H, branched monomer ), 8.30-8.38 (s, -NH-, branched monomer). (Mw =

137 kDa, Mn = 68 kDa, Đm = 2.0, Tg = 31 °C, Td = 320 °C)

3.3.5. Mechanical Property Measurements

To compression mold PEU films, polymers were pulverized into a fine powder using a Strand Mill Grinder (Model # S101DS). Each polymer was funneled in a mold (5 cm x 5 cm x 0.5 mm) and then placed in a vacuum compression instrument (TMP Technical

Products Corp). The polymers were melted (163 °C) and allowed to equilibrate for 30 minutes followed by degassing cycles (1000 psi). The polymer molds were pressed at 69

MPa, 103 MPa, and 138 MPa. The mold was then rapidly cooled to ambient temperature to afford the respective amorphous polymer films which were then cut into tensile bars

(4.76 mm x 38.1 mm x 0.5 mm). Elastic moduli, yield stress (σy), and yield strain (εy) were determined using tensile tests (Instron 5543 Universal Testing Machine) at 25 °C. The dimensions of each specimen were measured using calipers to ensure accurate 41 measurement. The elastic linear region was determined using linear regression with R2 values ≥ 0.98. The yield stress and yield strain were subsequently measured after the linear region. Statistical analyses were performed using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.05 was considered significant.

3.3.6. In Vivo Implant Degradation

An animal model was used to assess performance of this PEU series in vivo; primarily monitoring mechanical properties and degradation. All procedures and animal handling were in accordance with the University of Akron Institutional Animal Care and

Use Committee (IACUC Protocol Number 16-02-5-BRD) standards. Tensile bars were sterilized using ethylene oxide gas (EtO) and any loss of molecular mass was assessed using SEC. The sterilized PEU tensile bars and polypropylene (PP) control (n = 7 for each polymer) were then subcutaneously implanted into the back of adult female Sprague-

Dawley rats (n = 22). All rats received an anesthetic drug cocktail (ketamine, xylazine, acepromazine, 29.6:5.95:0.53 mg/kg respectively). Isoflurane (1.0% in 100% oxygen) was additionally administered to each rat through a nose-cone throughout the surgical procedure to maintain anesthetized state. A scalpel was used to create four dorsal incisions (1 cm in length) equidistance apart from the spine. Hemostats were then used to tunnel and create a subcutaneous pocket followed by polymer implantation with tweezers. The incisions were then closed with Michel clips. Survival rate was 100%

(22/22) for all time points (2 and 3 month). PEU polymer implants were collected

42 postmortem for each time-point and subjected to further characterization and mechanical testing.

3.3.7. Host-Implant Interaction

Polymer samples and surrounding tissue were collected post-mortem, fixed in a paraformaldehyde solution, and then embedded in paraffin wax for processing.

Embedded samples were sectioned (5 μm thick) and placed on microscope slides. All slides were stained in hematoxylin and eosin (H&E) and then fixed in DPX histology mount. Slides were then taken for imaging and the fibrous capsule thickness was measured at the 2 and 3 month time points to assess the host-immune response.

Statistical analyses were done using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.01 was considered significant. Additionally, each slide was assessed for inflammatory cell infiltrate based on a modified scoring system outlined by the

International Organization for Standardization (ISO 10993-6 Annex E) by a board-certified veterinary pathologist. The numbers of inflammatory cells were estimated in a 400X field by light microscopy and a score was assigned for each inflammatory cell type as denoted in table 3. The most severely affected region of the evaluated tissue was utilized to assign a score. The severity of necrosis was judge by the percentage of the fibrous capsule exhibiting evidence of necrosis (pyknosis, karyorrhexis or karyolysis) not including any inflammatory cell infiltrate.

3.4. Results

3.4.1. Synthesis

43

Linear and branched monomers were synthesized and characterized using 1H-

NMR spectrometry (Appendix A Figure 6.1). Monomers were synthesized via an esterification with 1,8-octanediol, 1,10-decanediol, or 1,12-dodecanediol and the carboxylic acid of L-valine using p-toluenesulfonic acid to afford 1-VAL-8, 1-VAL-10, and 1-

VAL-12. The linear monomers have similar 1H NMR spectra with the only variation coming from the integration that corresponds to the methylene resonances in the varying diol- chain lengths, shown between 1.22-1.35 ppm. The branched monomer (Triol-TYR) was synthesized via an esterification reaction between 1,1,1-tri(hydroxylmethyl)ethane and

Boc-O-benzyl-L-tyrosine using DIC as the coupling reagent. Urea byproducts were removed via silica gel chromatography. The final product was afforded following Boc- deprotection with 4 M HCl/dioxane, as demonstrated by the disappearance of the singlet at 1.28 ppm and appearance of the broad amine resonances at 8.72-8.78 ppm.

The PEUs were synthesized via interfacial polymerization using (1-VAL-8),

(1-VAL-10), or (1-VAL-12), and triphosgene (Scheme 3.1).84,87 The L-valine amino acid was chosen for its less rigid side chain when compared to previously studied L-phenylalanine and L-tyrosine which affords greater chain flexibility. Polymer synthesis was confirmed through 1H-NMR spectrometry (Appendix A Figure 6.2). The polymer spectra are discernible from the variable intensity of the methylene resonances highlighted in blue and denoted “b”. Branched PEUs were prepared using (1-VAL-8) or (1-VAL-10) with (Triol-

TYR) (Scheme 3.1) in a molar feed ratio of 98:2 respectively. Successful synthesis was confirmed through 1H-NMR spectrometry (Appendix A Figure 6.2). The extent of

44 branching was confirmed by comparing the integration of the six methylene protons denoted “e” from the Triol-TYR monomer to the twelve methyl L-valine protons denoted

“n” from the linear monomers. Post precipitation molecular mass and molecular mass distributions for all five polymers are listed (Table 3.1). All Mw values are greater than 71 kDa with Đm 1.7-3.3. Linear PEUs have Đm less than the theoretical value 2.0 because some of the lower molecular mass chains are fractionated during the precipitation process. The 2% branched polymers exhibit higher molecular mass because the Mn and

Mw values were obtained from a linear polystyrene standard.

Scheme 3.1. General synthetic scheme for L-valine monomers with diol-chain length varied between 8, 10, and 12 methylene units. Poly(1-VAL-8), poly(1-VAL-10), and poly(1- VAL-12) were synthesized using interfacial polymerization with triphosgene. Branched PEUs were synthesized using 1-VAL-8 and 1-VAL-10 with a 2% molar feed ratio of Triol- TYR to afford 2% branched poly(1-VAL-8) and 2% branched poly(1-VAL-10), respectively.

3.4.2. Physical Properties

45

Thermogravimetric analyses (TGA) for linear PEUs and 2% branched PEUs

(Appendix A Figure 6.3) (Table 1) show high degradation temperatures that make it suitable for compression mold processing. Poly(1-VAL-12) shows a broader degradation temperature which is consistent with previously published work and could be attributed

118 to greater chain flexibility which allows for more degradation processess. Values of Td are significantly higher than the reported Tg values (Appendix A Figure 6.4). As the methylene units within the polymer backbone increase, chain flexibility increases concomitantly, resulting in suppression of the Tg with values ranging between 29-42 °C

(Table 3.1). When the branching unit is incorporated, a drop in the Tg is further suppressed when compared to linear counterparts. This can be attributed to the branching unit interrupting interchain packing and hydrogen bonding between the urea groups.

Table 3.1. Molecular mass degradation determined following ethylene oxide sterilization and post-implantation by SEC. Physical properties of poly(1-VAL-8), poly(1-VAL-10), poly(1-VAL-12), poly[(1-VAL-8)0.98-co-(Triol-TYR)0.02], and poly[(1-VAL-10)0.98-co-(Triol- TYR)0.02] polymers analyzed in this study.

Polymer Initial Post-EtO 2 Month 3 Month Tg Td

Mn Mw Đm Mn Mw Đm Mn Mw Đm Mn Mw Đm (°C) (°C)

P(1-VAL-8) 42 71 1.7 53 79 1.5 59 79 1.3 94 105 1.1 42 310

P(1-VAL-10) 46 71 1.6 46 67 1.4 36 61 1.7 96 106 1.1 34 339

P(1-VAL-12) 51 75 1.4 59 78 1.3 66 85 1.2 102 111 1.1 29 205

2% Branched P(1-VAL-8) 126 410 3.3 61 113 1.9 65 247 3.8 67 210 3.1 35 301

2% Branched P(1-VAL-10) 68 137 2.0 63 150 2.4 44 94 2.2 46 85 1.9 31 320

Polypropylene ------384 46

Molecular masses were measured using SEC for each of the polymers before and after EtO sterilization and after each time point in vivo. The Mn, Mw, and Đm are reported. The physical properties of L-valine PEUs were assessed prior to sterilization and in vivo implantation. The Tg, Td, Mn, Mw, and Đm were all recorded.

3.4.3. In Vivo Degradation

In vivo polymer tensile bar implantation was performed using melt pressed ASTM standard tensile bars which were implanted subcutaneously into the dorsum of female

Sprague-Dawley rats (Scheme 3.2 A-B). A small incision was made followed by subcutaneous tunneling with hemostats, leading to polymer implantation and final incision closure (Scheme 3.2 C-F). Tracking molecular mass degradation from sterilization to in vivo implantation is important as mechanical failure in any soft-tissue device is likely to accompany molecular mass degradation (Table 3.1) (Appendix A Figure 6.5).

Scheme 3.2. General tensile bar implantation. Tensile bars (A) were cut with a dye-cutter and subcutaneously implanted in to the backs of female Sprague-Dawley rats (B). Basic surgical procedures included subcutaneous incision (C) with surgical blade, subcutaneous pocket tunneling (D) with hemostats followed by polymer tensile bar insertion (E) and final incision closure (F) with Michel clips.

The molecular masses of the PEUs were maintained throughout ethylene oxide

(EtO) sterilization which overcomes a crucial hurdle for the commercialization of these materials. Interestingly, an increase in Mn and Mw and a decrease in Đm values is observed

47 post-implantation. This change is attributed to lower molecular mass polymer chains having greater mobility and degrading faster. As a result, higher mass chains are left behind, affording a lower Đm. This trend holds when comparing the poly(1-VAL-8), poly(1-

VAL-10), and poly(1-VAL-12) molecular mass values before and after implantation. The molecular mass for the 2% branched polymers show less of a clear trend, however, the molecular mass distribution does narrow between the initial and 3 month time points likely due to the degradation of shorter chains. Surface topology images of the PEUs and polypropylene (Figure 3.1) illustrate the in vivo degradation of the PEUs post implantation. All PEU analogues elicit a surface eroding morphology which is consistent with previously studied PEU materials.84,32 After EtO sterilization, all samples have comparably smooth surfaces with limited defects (Figure 3.1 A-F Left). Through 2 months, a noticeable change can be seen for the PEU analogues where surface roughness and cavities start to appear (Figure 3.1 A-F Middle). Surface roughness and defects are not observed for PP (Figure 3.1 D Middle). After 3 months, the morphology change is more noticeable for the PEUs. Based on SEM surface morphology, poly(1-VAL-8) (Figure 3.1 A

Right) shows larger cavities and surface defects than poly(1-VAL-10) (Figure 3.1 B Right) and poly(1-VAL-12) (Figure 3.1 C Right) which display intermediate degradation. The initial smooth surface topology of the PP does not change with only minor cracks visible

(Figure 3.1 D Right) which is indicative of the non-resorbable nature of the material. 2% branched poly(1-VAL-8) (Figure 3.1 E Right) shows limited surface erosion relative to its

PEU linear analogue which aligns with the in vivo mechanical degradation results. This

48 result is attributed to the covalent crosslinking and hydrophobicity of the branching unit which help repel water and subsequent hydrolytic surface erosion.85 The 2% branched poly(1-VAL-10) (Figure 3.1 F Right) shows more degradation than 2% branched poly(1-

VAL-8). This is attributed to the increased flexibility of the polymer backbone which allows for greater water penetration and increased surface erosion. 2% poly(1-VAL-10) appears to have greater surface erosion than its linear counterpart, which could be attributed to the branching unit disrupting chain packing. This result correlates well with the observed in vivo mechanical degradation results as 2% poly(1-VAL-10) elicits the greatest amount of in vivo mechanical degradation through 3 months (Table 3.2).

49

. Figure 3.1. Scanning electron microscopy (SEM) was performed on all polymers at each time point to observe variations in the surface morphology. The surface topology of P(1- VAL-8) (A), p(1-VAL-10) (B), p(1-VAL-12) (C), PP (D), 2% branched p(1-VAL-8) (E), and 2% branched p(1-VAL-10) (F) after EtO sterilization (left) are compared after 2 (middle) and 3 months (right) of implantation. Images were captured at 750 x magnification and scale bars indicate 10 μm.

50

3.4.4. Mechanical Properties

Tensile testing was performed on the PEU and PP tensile bars prior to implantation, after sterilization, and at each in vivo time point. The stress and strain curves were recorded

(Appendix A Figure 6.6) and all extrapolated values were reported (Table 3.2). The

Young’s modulus was extrapolated from the linear region of the stress and strain curves

(Figure 3.2 A). For the linear and branched PEU analogues, a decrease in diol chain length manifests in an increase in Young’s modulus. This observed trend corroborates the hypothesis that increasing diol chain length will increase the polymer chain flexibility and decrease the material stiffness. Once exposed to EtO sterilization poly(1-VAL-8) and poly(1-VAL-12) moduli increased slightly. The increase is attributed to EtO having a plasticizing effect on the polymers and a corresponding increase in hydrogen bonding among the urea groups.25,119 The sustained mechanical properties after sterilization are ideal for future clinical application of these materials. When comparing sterilized samples to in vivo samples, a modulus drop was observed across all samples except for PP. This decrease in modulus is indicative of the hydrolytic and enzymatic PEU degradation in vivo, which reduces material stiffness and PP’s non resorbable nature. Of the linear and branched PEUs studied, the 2% branched poly(1-VAL-8) maintained the largest moduli values through the 3 month time point.

51

Table 3.2. Mechanical properties comparison

Polymer Initial Post-EtO 2 Month 3 Month Modulus (MPa) 193 ± 14 255 ± 11 70 ± 3 45 ± 5 P(1-VAL-8) σy (MPa) 34.4 ± 2.1 37.3 ± 7.7 7.2 ± 1.7 5.6 ± 1.2 0.2 0.2 0.1 0.1 εy (mm/mm) Modulus (MPa) 183 ± 11 223 ± 2 60 ± 38 38 ± 6 P(1-VAL-10) σy (MPa) 22.2 ± 1.4 30.9 ± 0.3 7.3 ± 8.0 8.6 ± 1.5 0.1 0.2 0.1 0.2 εy (mm/mm) Modulus (MPa) 105 ± 30 175 ± 17 48 ± 24 40 ± 5 P(1-VAL-12) σy (MPa) 10.3 ± 1.0 17.6 ± 1.0 3.9 ± 3.3 1.1 ± 0.4 0.2 0.1 0.1 < 0.1 εy (mm/mm) Modulus (MPa) 269 ± 12 251 ± 12 86 ± 22 78 ± 34 2% Branched σy (MPa) 53.0 ± 4.2 43.1 ± 4.6 14.1 ± 5.8 6.5 ± 1.0 P(1-VAL-8) 0.2 0.2 0.2 0.1 εy (mm/mm) Modulus (MPa) 140 ± 71 196 ± 9 8 ± 2 20 ± 6 2% Branched σy (MPa) 18.2 ± 8.2 13.3 ± 8.1 0.1 ± < 0.1 0.1 ± 0.1 P(1-VAL-10) 0.2 0.1 < 0.1 < 0.1 εy (mm/mm) Modulus (MPa) 165 ± 5 194 ± 10 200 ± 12 190 ± 9 Polypropylene σy (MPa) 27.4 ± 0.2 25.4 ± 1.0 26.3 ± 1.3 22.4 ± 1.9 0.2 0.2 0.2 0.1 εy (mm/mm) The Young’s modulus, stress at yield (σy) and strain at yield (εy) were measured and recorded. Values reported are an average of 4-6 samples.

This was expected as the 2% branched poly(1-VAL-8) has the shortest diol chain length and a hydrophobic branching unit which slows degradation when compared to the linear counterparts. The yield stress (σy) and yield strain (εy) were subsequently measured after the linear region. The σy (Figure 3.2 B) trends downward for all PEUs post- implantation. As the polymer chains are hydrolytically and enzymatically cleaved, the σy naturally decreases as the number of chains with molecular masses above chain entanglement diminishes. This trend is not observed for PP samples as degradation is negligible. The εy across samples do not show a clear trend (Figure 3.2 C). For the 2% branched poly(1-VAL-8) samples, a decrease in εy is observed after three months.

Alternatively, the εy values for poly(1-VAL-10) increase while with other samples like

52 poly(1-VAL-8) and poly(1-VAL-12) exhibited no change. Variation in εy was minimal between samples as all samples fell between 0.0-0.2 mm/mm. This indicated that although there was no observable trend, there was a well-defined εy range for these materials.

53

Figure 3.2. Young’s modulus for implanted materials was extrapolated at each time-point through linear regression with R2 = 0.98. P(1-VAL-8), p(1-VAL-10), p(1-VAL-12), 54 polypropylene, 2% branched p(1-VAL-8), and 2% branched p(1-VAL-10) moduli values were assessed through the 3 month time point (A). * or ** indicate a p value < 0.05 between a reference sample (first sample denoted with * or ** reading from left to right) and other samples sharing like symbols (n = 4-6 samples). For example, * indicates a significant difference between initial p(1-VAL-8) and post-EtO p(1-VAL-8), between initial p(1-VAL-8) and 2 month p(1-VAL-8), and between initial p(1-VAL-8) and 3 month p(1-VAL-8) moduli values. * does not indicate a significant difference between 2 month and 3 month p(1-VAL-8) moduli values. Yield stress (σy) was measured at the yield point for p(1-VAL-8), p(1-VAL-10), p(1-VAL-12), polypropylene, 2% branched p(1-VAL-8), and 2% branched p(1-VAL-10) samples through the 3 month time point (B). *, **, or *** indicate a p value < 0.05 between samples sharing similar symbols (n = 4-6 samples). Statistical difference can be discerned the same way as previously explained for moduli values. Yield strain (εy) was measured at the yield point for p(1-VAL-8), p(1-VAL-10), p(1-VAL-12), polypropylene, 2% branched p(1-VAL-8), and 2% branched p(1-VAL-10) samples through the 3 month time point (C). * or ** indicate a p value < 0.05 between samples sharing similar symbols (n = 4-6 samples). Statistical difference can be discerned the same way as previously explained for moduli values. 3.4.5. Histology

Histology images for PEUs and the PP control (Figure 3.3 A-F) are H&E stained cross-sectional areas of paraffin embedded polymer and surrounding tissue postmortem.

Implanted biomaterials characteristically induce a foreign body response and can elicit subsequent formation of a collagenous fibrous capsule as fibroblasts attempt to encapsulate the implanted material via the production of collagenous matrix.120 The formation of a fibrous capsule is an indication of sustained inflammatory response and a trademark of non-resorbable biomaterials.121 One of the challenges with using non- resorbable polymers as hernia repair materials is that sustained chronic inflammation can prevent fibrous tissue remodeling and eventual quiescence. Over time this can lead to tissue weakness and ultimately hernia recurrence. The fibrous capsule is identified as the circumferentially arranged eosinophilic collagenous matrix surrounding the polymer implant with interspersed fibroblasts containing elongated basophilic nuclei. Capsule 55 thickness was measured around the perimeter of each polymer implant at 2 and 3 month time points (Figure 3.4 A-B) (Appendix B Table 6.1).

Figure 3.3. Histology images of (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) olypropylene, (E) 2% branched P(1-VAL-8), and (F) 2% branched P(1-VAL-10) are from the cross-sectional area of polymer and surrounding tissue which was stained with hematoxylin and eosin. All images are from the 2 month timepoint at 20 x magnification with scale bars being equal to 1 mm.

Branched PEU analogues and poly(1-VAL-12) show a significantly thinner fibrous capsule thickness through 2 months when compared to PP. This is likely an indication of a decreased fibrous tissue proliferative response to chronic inflammation. No significant difference is noted between p(1-VAL-8) or p(1-VAL-10) and PP through 2 months.

56

However, at 3 months, all five PEUs exhibit significantly smaller fibrous capsule thickness than PP. This change is attributed to the remodeling process differences between PEUs and PP at the extruded timepoint. As PEUs degrade, cellular infiltration can occur which reduces chronic inflammation and promotes tissue remodeling. This shift towards native tissue deposition through the tissue remodeling process is evidenced by the reduction of fibrous capsule thickness. Remodeling is ideal for a hernia repair material as native tissue has greater mechanical integrity than fibrous capsule scar tissue. The improved inflammatory response over time for the L-valine based PEUs compared to PP make these materials potential candidates for soft-tissue applications.

Figure 3.4. Capsule thickness values for 2 month samples (A) were measured to assess inflammatory response (* indicates p value < 0.01 between P(1-VAL-8) and P(1-VAL-12) and between P(1-VAL-8) and 2% branched P(1-VAL-8) samples. ** indicates p value < 0.01 between P(1-VAL-10) and P(1-VAL-12), and between P(1-VAL-10) and 2% branched P(1- VAL-8) samples. *** indicates p value < 0.01 between P(1-VAL-12) and polypropylene, and between P(1-VAL-12) and 2% branched P(1-VAL-10) samples. **** indicates p value 57

< 0.01 between polypropylene and 2% branched P(1-VAL-8), and between polypropylene and 2% branched P(1-VAL-10) samples, n = 7). Capsule thickness values were also assessed for 3 month samples (B) (* indicates p value < 0.01 between P(1-VAL-8) and polypropylene samples. ** indicates p value < 0.01 between P(1-VAL-10) and polypropylene and between P(1-VAL-10) and 2% branched P(1-VAL-10) samples. *** indicates p value < 0.01 between P(1-VAL-12) and polypropylene samples. **** indicates p value < 0.01 between polypropylene and 2% branched P(1-VAL-8), and between polypropylene and 2% branched P(1-VAL-10) samples, n = 7).

Further histological characterization on the H&E stained samples was performed to assess cell infiltration and inflammation. Samples were scored from 0-4 for neutrophils, lymphocytes, plasma cells, macrophages, multinucleated giant cells, and necrosis (Table 3.3). All sections contained a clear region in the subcutaneous tissue where the implant was removed. This region was surrounded by a thin circular rim of reactive tissue comprised primarily of eosinophilic fibrillary extracellular material (fibrous connective tissue) containing numerous spindle shaped cells (fibroblasts) varying from immature (larger nuclei and visible cytoplasm) to mature (small hyperchromatic nuclei with minimal visible cytoplasm). Profiles of capillaries were observed in the fibrous capsule were observed in all sections. Sections containing the polypropylene implant exhibited a higher degree of inflammation as assessed by the scoring scale described above (Table 3.4). All inflammation scores were lower than the PP implant scores with only the 2% branched P(1-VAL-8) implant exhibiting a similar inflammation score (Table

3.4). In addition, multinucleated giant cells (a classic finding associated with foreign body responses) were observed in the fibrous capsule associated with the PP implant.

Multinucleated giant cells were not a feature of other implants with the exception of one

58

2% branched P(1-VAL-8) implant. Necrosis of fibroblasts comprising the capsule was only observed in the PP implant group.

Table 3.3. H&E slide scoring scale

Cell 0 1 2 3 4 Type/Response

Rare, 1- 5-10/400X Heavy Neutrophils 0 Packed 5/400X field field Infiltrate

Rare, 1- 5-10/400X Heavy Lymphocytes 0 Packed 5/400X field field Infiltrate

Rare, 1- 5-10/400X Heavy Plasma Cells 0 Packed 5/400X field field Infiltrate

Rare, 1- 5-10/400X Heavy Macrophages 0 Packed 5/400X field field Infiltrate

Rare, 1- 3-5/400X Heavy Multinucleated 0 Sheets Giant Cells 5/400X field field Infiltrate

Minimal Mild Moderate Severe Necrosis None (<25%) (25-50%) (50-75%) (75-100%) Modified scoring system outlined by the International Organization for Standardization (ISO 10993-6 Annex E)

Table 3.4. H&E slide scores for implanted materials at 3 months

P(1-VAL-8) P(1-VAL-10) P(1-VAL-12) PP 2% 2% Branched Branched P(1-VAL-8) P(1-VAL-10) Neutrophils 1.0 1.5 0.5 1.5 1.0 1.7 Lymphocytes 1.0 1.5 1.5 1.0 2.5 0.7 Plasma Cells 0.5 0 0 0 0 0 Single 0.5 1.5 0.5 2.5 2.5 0.7 Macrophages Multinucleated 0 0 0 1.0 0.5 0.0 Giant Cells Necrosis 0 0 0 1.5 0 0.0 Score Totals 3.0 4.5 2.5 7.5 6.5 3.1 Reported scores are the average of 4-6 sections. 59

3.5. Conclusion

Non-resorbable polymeric materials such a polypropylene are commonly used in medical devices for soft tissue repair due to their high mechanical strength and durability.

However, these materials are burdened by higher risk of infection, dense fibrous capsule formation and shrinkage. Such drawbacks may be potentially remedied by long term resorbable polymers. We therefore explored a series of L-valine based PEUs as alternatives for soft-tissue repair applications. The PEUs tested show Young’s moduli comparable to PP (105 ± 30 - 269 ± 12 MPa) with the 2% branched poly(1-VAL-8) maintaining the greatest mechanical properties measured up to 3 month following in vivo implantation. Notably, all PEUs generated a limited inflammatory response through three months relative to polypropylene. A limited inflammatory response through 3 months along with tunable mechanical properties make L-valine based PEUs an attractive material to be explored for soft-tissue repair. Further investigation in a clinically relevant hernia model are ongoing.

3.6. Acknowledgement

The authors gratefully acknowledge financial support from Cook Biotech, The NSF

Research Experience for Undergraduates (DMR 1359321) in the College of Polymer

Science and Polymer Engineering and the W. Gerald Austen Endowed Chair in Polymer

Science and Polymer Engineering from the Knight Foundation. Additionally, the authors

60 would like to sincerely thank Keyence for the use of their Keyence BZ-X700 microscope for histological imaging.

61

CHAPTER IV

AMINO ACID-BASED POLY(ESTER UREA) COPOLYMER FILMS FOR HERNIA-REPAIR

APPLICATIONS

In part this work has been reprinted with permission from Dreger, N. Z.; Fan, Z.;

Zander, Z. K.; Tantisuwanno, C.; Haines, M. C.; Waggoner, M.; Parsell, T.; Søndergaard,

C. S.; Hiles, M.; Premanandan, C.; Becker, M. L., Amino Acid-Based Poly(ester urea)

Copolymer Films for Hernia-Repair Applications, Biomaterials, 2018, 182, pp 44-57.

Copyright 2018 Elsevier Ltd.

4.1. Abstract

The use of degradable materials is required to address current performance and functionality shortcomings from biologically-derived tissues and non-resorbable synthetic materials used for hernia mesh repair applications. Herein a series of degradable L-valine- co-L-phenylalanine poly(ester urea) (PEU) copolymers were investigated for soft-tissue repair. Poly[(1-VAL-8)0.7-co-(1-PHE-6)0.3] showed the highest uniaxial mechanical properties (332.5 ± 3.5 MPa). Additionally, L-valine-co-L-phenylalanine poly(ester urea)s were blade coated on small intestine submucosa extracellular matrix (SIS-ECM) and found to enhance the burst test mechanical properties of SIS-ECM in composite films (force at break between 102.6 ± 6.5 – 151.4 ± 11.3 N). Free standing films of L-valine-co-L-

62 phenylalanine PEUs were found to have superior extension at break when compared to

SIS-ECM (averages between 1.2-1.9 cm and 1.2 cm respectively). Cellular spreading and proliferation were observed in vitro with a reduced inflammatory response for poly[(1-

VAL-8)0.7-co-(1-PHE-6)0.3] when compared to polypropylene in an in vivo hernia rat model.

These results support the use of PEU copolymers as free-standing films or as composite materials in soft-tissue applications for hernia-repair.

4.2. Introduction

A number of synthetic polymers and biologically derived materials have been employed as meshes for the repair and treatment of hernias.56,57,89,103,105,122,123 Of these, polypropylene is the most widely used clinically for ventral hernias, yet the recent clinical challenges, litigation and outcomes leave much to be desired. Polypropylene is non- resorbable and often leads to prolonged foreign body responses that result in long-term complications years after the initial surgery.54,59,64,100,104,124 Additionally, polypropylene has vastly different mechanical properties compared to native tissue causing the implant to feel rigid with high incidence of patient discomfort.125,126 Ultimately, polypropylene is overengineered for many of the soft-tissue hernia locations.

To combat the shortcomings of polypropylene, resorbable materials are being explored as alternatives for hernia mesh repair. The ideal material would simultaneously possess tunable degradation rates to match the needs of the injury, sufficient mechanical support for the duration of the healing process to prevent long-term recurrence, and a limited foreign body response to prevent pain and discomfort. Degradable polyesters

63 have been employed as alternative devices to improve tissue ingrowth however, they are prone to local acidic degradation byproducts and have not been shown to decrease patient pain or discomfort.56,59,126 Current synthetic derivatives used include DEXON

(polyglycolic acid), VICRYL (poly(lactic-co-glycolic acid), PHASIX (poly-4-hydroxybutyrate), and GORE (Poly(glycolide-co-trimenthylene carbonate) with degradation times varying from 2-18 months.67 Premature loss in mechanical properties, inflammation, or lack of sufficient clinical data have precluded a clear frontrunner in degradable implants.

One alternative to synthetic materials is a xenogeneic, decellularized small- intestine submucosa extracellular matrix (SIS-ECM), which concurrently possesses sufficient mechanical properties for hernia-repair and promotes cellular integration.75,78,101,127 SIS-ECM has been used in the treatment of hernia repair with a varying degree of clinical success that depends on efficient cellular integration into the device to compensate for the rapid loss in mechanical support as the mesh degrades.100,113,128,129 100,113,128,129 If SIS-ECM sustains the mechanical properties necessary for the given application throughout the remodeling process the clinical outcomes are excellent as native tissue is replaced and the defected area is remodeled.65,97,111,122 However, if the degradation rate and subsequent loss of mechanical properties occurs prematurely then recurrence can occur prior to proper healing and remodeling leading to device failure.94,127,129 Although SIS-ECM has high cell activity it lacks elasticity and fails at low strain, whereas polypropylene elicits prolonged inflammation but is mechanically robust.54,102,104,107 Some strategies have been

64 attempted to couple the positive components from each material utilizing ECM as a coating for polypropylene.130, 55 While these strategies have found noticeable improvement in mitigating the initial innate immune response, the non-resorbable polypropylene remains after ECM degradation, leading to the aforementioned complications.

There is a clear need for some material to bridge the gap between the mechanical competence of polypropylene and the resorptivity and cellular integration of SIS-ECM.

One class of materials gaining interest are amino acid-based polyester derivatives which include poly(ester amide)s, poly(ester urea)s, and poly(ester urethane)s.131–134 These materials have been shown to meet a variety of biomedical applications by having readily cleavable hydrolytic or enzymatic units with their applications of use ranging from drug delivery to bone-tendon repair, to vascular tissue regeneration.131,133–135 What makes these amino acid-based materials so appealing is that their degradation byproducts are essential amino acids which promotes limited cytotoxicity. Poly(ester urea)s in particular have been shown to have a limited inflammatory response in vivo with tunable mechanical properties.84,85,87 A series of linear and branched L-valine based poly(ester urea)s (PEU)s have previously been assessed in vitro and in vivo as possible alternative materials for soft-tissue repair applications.88 However, for hernia-mesh repair applications L-valine based PEUs in vivo degradation rates are too rapid and cause the mechanical properties to decline more quickly than desired. Additionally, branched L- valine based PEUs led to solubility challenges which precluded them from roll-to-roll

65 fabrication methods and future scalability. In contrast, the degradation rates of L- phenylalanine based PEUs are substantially slower than that observed for L-valine based

PEU with mechanical properties comparable to bone.136 To meet the needs for hernia repair, bolstered initial mechanical properties could help prevent device failure over a longer period of time.85,87 To combine the tunable mechanical properties and limited inflammatory response with sustained mechanical properties and eventual degradation a series of resorbable copolymers consisting of L-valine and L-phenylalanine monomers were fabricated and processed in a pilot roll to roll processing method as free standing films and as composite films with SIS-ECM. Herein, we report the characterization, mechanical properties, in vitro cell viability, and inflammatory response of L-valine and L- phenylalanine copolymers and their composites with SIS-ECM in a hernia repair rat model.

4.3. Experimental

4.3.1. Materials

1,8-octanediol, 1,6-hexanediol, sodium carbonate, p-toluenesulfonic acid monohydrate, and triphosgene were all purchased from Sigma Aldrich (Milwaukee, WI).

Toluene, chloroform, acetone, and N,N-dimethylformamide were purchased from

Fischer Scientific (Pittsburgh, PA). L-valine and L-phenylalanine were purchased from

Acros (Pittsburgh, PA). The commercially available 2.0 1-1 LL SIS-ECM (Single-layer lyophilized ECM) was provided by Cook Biotech and used as provided. All solvents were

66 reagent grade and all chemicals were used without further purification unless otherwise stated.

4.3.2. Characterization

1H-NMR and 13C-NMR spectra were collected using a 300 MHz and 500 MHz Varian

NMR spectrophotometer, respectively. Chemical shifts are reported in ppm (δ) and

1 referenced to residual solvent resonances ( H-NMR, DMSO-d6: 2.50 ppm). Multiplicities were explained using the following abbreviations: s = singlet, d = doublet, t = triplet, br = broad singlet, and m = multiplet. Size exclusion chromatography (SEC) was performed using an EcoSEC HLC-8320GPC (Tosoh Bioscience, LLC) equipped with a TSKgel SuperH-RC

6.0 mm I.D. × 15 cm mixed bed column and refractive index (RI) detector. The number average molecular mass (Mn), weight average molecular mass (Mw), and molecular mass distribution (ĐM) for each sample was calculated using a calibration curve determined from poly(styrene) standards (PStQuick MP-M standards, Tosoh Bioscience LLC) with 0.01

M LiBr in DMF as the eluent flowing 1.0 mL/min at 50 °C. Differential scanning calorimetry

(DSC) was performed using a TA Q200 scanning a temperature range from 0-100 °C with heating and cooling ramps of 20 °C/min. The glass transition temperature (Tg) was determined from the midpoint of the second heating curve. Thermogravimetric analysis

(TGA) was performed using a TA Q500 with heating ramps of 20 °C/min in the temperature range from 0-500 °C/min. The degradation temperature (Td) was determined from 10% mass loss. Statistical analysis was performed using a post-hoc

Tukey ANOVA test. Surface topology images were obtained using scanning electron

67 microscopy (SEM) with a JEOL USA SEM. Samples were sputter-coated with gold and scanned at 2.0 kV excitation at 200 × magnification. Histology images were obtained using a Keyence BZ-X700 at 20 × magnification. Statistical analyses on mechanical properties were performed using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.05 was considered significant. Additionally, statistical analyses on in vitro cell results were performed using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.01 and p < 0.001 was considered significant.

4.3.3. Synthesis of Poly(ester urea) Copolymer Monomers

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-valine)-Octane 1,8-Diester

Monomer. (1-VAL-8). Synthesis of di-p-toluenesulfonic acid salts of bis(L-valine)-octane

1,8-diester (1-VAL-8) was carried out following previously published procedures.88

Briefly, 1,8-octanediol (43.8 g, 0.30 mol, 1.0 eq.), L-valine (73.8 g, 0.63 mol, 2.3 eq.), p- toluenesulfonic acid monohydrate (131.3 g, 0.69 mol, 2.4 eq.), and toluene (1300 mL) were added to a 1 neck flask and equipped with a stir bar. A Dean-Stark trap was fastened to the round bottom flask and the reaction was heated to 110 °C to reflux for 24 h. The reaction was cooled to room temperature, and the resulting white precipitate was isolated by vacuum filtration using a Buchner funnel. The product was dissolved in boiling water (2 L), hot filtered, and cooled to room temperature to further purify the white solid precipitate. The precipitate was collected via filtration and the recrystallization process was performed three times for purity (166.0 g, 79% yield). 1H-NMR (300 MHz, 303 K,

DMSO-d6): δ = 0.94 (m, 12H), 1.28 (s, 8H), 1.59 (m, 4H), 2.07-2.18 (m, 2H), 2.27 (s, 6H),

68

3 2.50 (s, DMSO), 3.33-3.38 (s, H2O), 3.89 (d, JH-H = 3.0 Hz, 2H), 4.07-4.23 (m, 4H), 7.07-7.23

3 3 (d, JH-H = 8.2 Hz, 4H, aromatic H ), 7.45-7.48 (d, JH-H = 8.1 Hz, 4H, aromatic H), 8.25 (br,

6H) ppm.

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-phenylalanine)-Hexane 1,6-

Diester Monomer. (1-PHE-6). Synthesis of di-p-toluene sulfonic acid of bis(L- phenylalanine)-hexane 1,6-diester (1-PHE-6) was carried out using the method and molar

1 equivalents described above (153.0 g, 73% yield). H-NMR (300 MHz, 303 K, DMSO-d6):

3 δ =1.06 (s, 4H), 1.38 (m, 4H), 2.27 (s, 6H), 2.50 (s, DMSO), 2.96−3.17 (m, 4H), 4.01 (t, JH-H

3 = 9.0 Hz , 4H), 4.28 (t, JH-H=6.0 Hz, 2H), 7.08−7.11 (d, 4 H), 7.20−7.35 (m, 10H), 7.45−7.48

(d, 4H), 8.37 (s, 6H) ppm.

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-phenylalanine)-Octane 1,8-

Diester Monomer. (1-PHE-8). Synthesis of di-p-toluene sulfonic acid of bis(L- phenylalanine)-octane 1,8-diester (1-PHE-8) was carried out using the method and molar

1 equivalents described above (157.0 g, 75% yield). H-NMR (300 MHz, 303 K, DMSO-d6):

3 δ =1.14 (s, 8H) 1.41 (m, 4H), 2.27 (s, 6H), 2.50 (s, DMSO), 2.96−3.17 (m, 4H), 4.02 (t, JH-H

3 6.0 Hz,, 4H), 4.28 (t, JH-H=6.0 Hz, 2H) 7.08−7.11 (d, 4H) 7.20−7.35 (m, 10H) 7.48−7.49 (d,

4H) 8.36 (s, 6H) ppm.

4.3.4. Synthesis of Poly(ester urea) Copolymers

The synthesis was carried out according to methods published previously.84,87

Interfacial polymerization of p-toluenesulfonic acid salts of bis(L-valine) and p- toluenesulfonic acid salts of bis(L-phenylalanine) monomers 1-VAL-8, 1-PHE-6, and 1-PHE-

69

8 was performed by dissolving the desired monomers with desired molar equivalents (1 eq. total) with sodium carbonate (3.4 eq.) in distilled water (0.1 M, 35 °C) in a 2 L 2-neck round-bottom flask. The flask was attached with an overhead mechanical stir rod and allowed to stir until clear. Triphosgene (0.35 eq.) was dissolved in distilled chloroform

(0.6 M) and subsequently added to the reaction vessel through an addition funnel. The solution turned white upon addition and the solution was stirred for 1 h before another aliquot of triphosgene (0.08 eq.) dissolved in distilled chloroform was added to help push the reaction to completion. The reaction was stirred for an additional 2 h when the reaction solution was transferred to a separatory funnel. The organic phase containing the desired PEU product was precipitated into boiling water to remove chloroform and residual starting material. White polymer was obtained and lyophilized (90-95 % yield).

1 Poly(1-VAL-8). (P(1-VAL-8)). H-NMR (300 MHz, 303 K, DMSO-d6): δ = 0.77-0.89

(m, 12H, -CH(CH3)2), 1.24-1.33 (s, 8H, -COOCH2CH2(CH2)4-), 1.50-1.58 (m, 4H, -

+ COOCH2CH2(CH2)4CH2-), 1.94-2.04 (m, 2H, -(CH3)2CHCHNH3 -), 2.50 (s, DMSO), 3.29-3.33

(H2O), 3.94-4.10 (m, 6H, -CHCOOCH2CH2(CH2)4-), 6.37-6.41 (s, 2H, -NH-) ppm. (Mw = 71 kDa, Mn = 42 kDa, Đm = 1.7, Tg = 42 °C, Td = 310 °C)

1 Poly[(1-VAL-8)0.7-co-(1-PHE-6)0.3]. (30% PHE6 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

70

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 18.29, 19.52, 25.53,

28.72, 29.13, 30.90, 39.95, 54.68, 57.86, 64.55, 126.75, 128.45, 129.43, 137.41, 157.61,

157.84, and 172.86 ppm. (Mw = 95 kDa, Mn = 57 kDa, Đm = 1.7, Tg = 29 °C, Td = 340 °C)

1 Poly[(1-VAL-8)0.8-co-(1-PHE-6)0.2]. (20% PHE6 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 18.23, 19.49, 25.86,

28.65, 29.00, 31.01, 40.00, 54.58, 58.29, 64.64, 127.22, 128.61, 130.02, 137.86, 157.66,

158.10, 172.96 ppm. (Mw = 88 kDa, Mn = 56 kDa, Đm = 1.6, Tg = 27 °C, Td = 278 °C)

1 Poly[(1-VAL-8)0.9-co-(1-PHE-6)0.1]. (10% PHE6 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 18.75, 19.81, 26.16,

29.01, 29.75, 31.44, 40.00, 55.15, 58.68, 65.08, 127.20, 129.65, 130.04, 138.13, 157.90,

158.63, and 173.12 ppm. (Mw = 60 kDa, Mn = 44 kDa, Đm = 1.4, Tg = 28 °C, Td = 268 °C)

1 Poly[(1-VAL-8)0.7-co-(1-PHE-8)0.3]. (30% PHE8 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

71 DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 17.76, 19.15, 25.91,

28.03, 28.66, 30.78, 40.00, 54.48, 57.96, 64.42, 126.85, 128.66, 129.43, 137.15, 156.92,

157.97, and 173.21 ppm. (Mw = 108 kDa, Mn = 69 kDa, Đm = 1.6, Tg = 28 °C, Td = 278 °C)

1 Poly[(1-VAL-8)0.8-co-(1-PHE-8)0.2]. (20% PHE8 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 17.36, 18.74, 24.77,

27.60, 28.27, 30.07, 40.00, 54.09, 57.56, 64.34, 126.45, 127.92, 128.99, 137.05, 156.52,

157.90, and 172.39 ppm. (Mw = 105 kDa, Mn = 68 kDa, Đm = 1.5, Tg = 36 °C, Td = 287 °C)

1 Poly[(1-VAL-8)0.9-co-(1-PHE-8)0.1]. (10% PHE8 P(1-VAL-8)). H-NMR (300 MHz, 303

K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -NHCH(CH(CH3)2)C(O)O-), 2.50 (s,

DMSO), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -NHCH(CH2Ph) C(O)O-),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05 (all

13 remaining protons) ppm. C-NMR (125 MHz, 303 K, DMSO- d6): δ = 18.52, 19.55, 25.83,

72 28.20, 28.91, 31.04, 40.00, 54.60, 58.37, 64.69, 127.12, 128.87, 129.91, 137.20, 157.37,

158.00, and 173.00 ppm. (Mw = 84 kDa, Mn = 49 kDa, Đm = 1.7, Tg = 34 °C, Td = 297 °C)

4.3.5. Uniaxial Mechanical Property Measurements

To compression mold PEU films, the polymers were milled into a fine powder using a Strand Mill Grinder. Each polymer powder was funneled in a mold (5 cm x 5 cm x 0.5 mm) and placed into a vacuum compression machine (TMP Technical Products Corp).

Each polymer was heated to melting temperature and equilibrated for 30 min under pressure (7 MPa). Air bubbles were removed by degassing cycles. The polymer molds were consecutively pressed at 70 MPa, 100 MPa, and 140 MPa. The mold was then rapidly cooled to room temperature to afford the respective amorphous polymer films, which were then cut into tensile bars (4.76 mm × 38.1 mm × 0.5 mm). Elastic moduli, yield stress (σy), and yield strain (εy) were determined using tensile tests (Instron 5543

Universal Testing Machine) at 25 °C with a strain rate of 25.4 mm/min. The dimensions of each specimen were measured using calipers to ensure accurate measurement. The viscoelastic linear region was determined using linear regression at 10% strain. The yield stress and yield strain were subsequently measured after the linear region at 10% strain.

4.3.6. PEU-ECM Films

Burst-test mechanical properties were obtained by blade coating PEU solutions on polyethylene terephthalate (PET) substrates. PEU copolymer analogues were dissolved in acetone at 5% weight and then filtered with 5 micrometer syringe filters to remove impurities. The solutions were then concentrated to 33 wt.% polymer. Small-intestine

73 submucosa extracellular matrix (Cook Biotech SIS Single-layer lyophilized ECM) was secured to PET with tape on the edges to prevent slippage during coating. Polymer solutions were then blade coated using a Bio-hybrid casting line (8 cm blade width, gap height 300 µm) on the ECM and allowed to air dry for 24 h. The PEU-ECM films were then further dried under reduced pressure to remove residual solvent. PEU-ECM films were cut into 5 × 5 cm sheets and submerged in 1× PBS (pH = 7.4) for five min. Films were then fastened in an ASTM D 3787-07 standard ball-burst apparatus (2.54 cm burst ball and 4.45 cm circular diameter) and burst with a constant rate of traverse (25.4 mm/min). Force at break (FB), extension at break (EB), and relative stiffness (RS) were recorded until film failure, as well as the location of the failure.

4.3.7. PEU Free-Standing Films

PEU copolymer free-standing films were prepared by blade coating with adaptations from the procedure described above. In short, PEU copolymers were dissolved in acetone and concentrated to 35 wt.%. Polymer solutions were then blade coated (8 cm blade width, 500 µm gap height, 150 cm/min) on PET and allowed to air dry for 24 h. The PEU films were then further dried under reduced pressure to remove residual solvent. Films were then cut into 5 × 5 cm sheets and submerged in PBS (pH =

7.4) for five min. Films were then fastened in an ASTM D 3787-07 standard ball-burst apparatus and burst with a constant rate of traverse (25.4 mm/min). FB, EB, and RS were recorded until film failure, as well as the location of the failure.

4.3.8. Film Water Uptake

74

Water uptake (WU) was carried out following previously published work.85

Samples prepared from blade-coating as described above (35 wt.% polymer) were weighed to determine an initial mass (Wi). Samples were then placed in PBS (pH = 7.4, 25

°C) for one week. Samples were then removed, blotted dry with a paper towel and immediately weighed to determine swollen weight (Wt). The percentage of water uptake was calculated from equation 4.1.

푊 −푊 푊푈(%) = 푡 푖 × 100% (4.1) 푊푖

4.3.9. Contact Angle

Contact angle was conducted using a Rame-Hart Contact Angle Goniometer on the blade coated

PEU copolymers to assess surface hydrophilicity. In brief, 5 µL of deionized water administered on the sample surface and the contact angle was immediately measured (n = 3). Images were analyzed using

ImageJ (NIH) software and the angles were thus reported.

4.3.10. In Vitro Degradation

SIS-ECM, 30% PHE6 P(1-VAL-8)-ECM and 30% PHE6 P(1-VAL-8) films were assessed in accelerated in vitro degradation studies. SIS-ECM, 30% PHE6 P(1-VAL-8)-ECM, and 30%

PHE6 P(1-VAL-8) were cut in to 0.5 x 1 cm films, weighed, and then placed in 0.025 M

NaOH PBS solution. Films were incubated at 37 °C for the duration of the study. When collected, films were lyophilized to remove residual water, reweighed, and mass loss observed and calculated at several time points through 48 h. These conditions were not harsh enough to observe a mass loss for the 30% PHE6 P(1-VAL-8) films through 48 h. To observe mass loss, 30% PHE6 P(1-VAL-8) was subjected to a 0.1 M NaOH PBS solution.

Mass loss was tracked through 14 d and recorded as described above. 75 4.3.11. Cell Viability

In order to evaluate PEU copolymer biocompatibility, cell viability, attachment and stretching were quantified on fabricated PEU films. Fibroblast phenotype mouse cells

(ATCC CRL-1658) were used to assess cell viability (passage 2) on the blade coated PEU films previously fabricated. Samples were sterilized by short-wave UV sterilization for 30 minutes prior to cell seeding. Cells were seeded at a cell density of 130,000 cells/cm2 and cultured in an incubator (48 hours, 37 °C). Samples were then removed and stained with a LIVE/DEAD™ Viability/Cytotoxicity Kit (ThermoFisher Scientific L3224). Slides were analyzed with a Keyence BZ-X700 microscope at 20× magnification and analyzed with

ImageJ software (n = 4-5) with all groups compared to a glass slide control. For cell survival and spreading, mouse L-929 cells (ATCC) (fibroblast phenotype cell line harvested from subcutaneous connective tissue) were used. Cells were cultured in Eagle's Minimum

Essential Medium (EMEM, ATCC) with 10% horse serum and a 1% solution of antibiotics

o (penicillin and streptomycin). Cells were cultured at 37 C, 21% O2, and 5% CO2. Cell attachment was investigated by staining L-929 cells cultured on 6 various copolymers and glass cover slips. In brief, polymeric substrates and glass coverslips were sterilized under

UV for 30 min followed by 4 h wetting within culture medium. L-929 cells, ranging from passage 6-8, were seeded on the respective substrate with a cell density of 1.3 × 104 cells/cm2 and cultured for 48 h in the incubator with same culture conditions. Harvested cell/substrates were fixed by 4% formaldehyde for 30 min followed by a PBS rinse (3×).

Cells were further blocked by 5% donkey serum in PBS supplemented with 0.5% TritonX-

76

100 for 1 h followed by another PBS rinse (3×). The actin filaments and cell nucleus were stained with rhodamine phalloidin (1:200) and DAPI (1:000 dilution in PBS) antibody dilution buffer for 40 min. The stained samples were washed 3 times with PBS before imaging with an IX 81 microscope (Olympus, Center Valley, PA) with 40 × objectives. The stretched area was quantified by measuring cell area via ImageJ. A CYQUANT assay

(Invitrogen) was utilized to determine the cell number on the various substrates. In brief, the lysis buffer was diluted 20 times and added to harvested cell/substrate samples. The cell lysis product was stored in -80 oC overnight before quantification. The frozen lysis sample was thawed to ambient temperature and mixed with the same volume of diluted

CYQUANT GR dye (1: 200). Five replicates were measured in each group. The fluorescence intensities of the experimental groups were normalized to the glass control group.

4.3.12. In Vivo Animal Model

Animal studies were adapted from several rat animal models and were used to assess the inflammatory response for 30% PHE6 P(1-VAL-8) compared to an ECM, and PP control.137,138 All procedures and animal handling were in accordance with The University of Akron Institutional Animal Care and Use Committee (IACUC Protocol Number 17-12-

16-BRD) standards. PP and PEU films were fabricated through compression molding and blade coating respectively. SIS-ECM (Single-layer lyophilized ECM), PP, and PEU films were cut in to circular disks (1 cm diameter). All films were sterilized using ethylene oxide gas (EtO). Animals were anesthetized with isoflurane ((range 1-4%)/ 100% O2), to maintain an anesthetized state. Upon anesthesia, animals were given a subcutaneous

77 shoulder injection of buprenorphineSR (ZooPharm Inc.) (0.7 mg/kg) for analgesia and incision site injection of Lidocaine (5mg/mL at 7mg/kg). Once the animals were anesthetized, the abdomen was trichromatized (divided in to three sections running parallel from tail to head) and a small incision of 1.4-1.6 cm in length was made at the umbilical region of the abdomen. An abdominal muscular hernia defect was created using a scalpel (~0.5 cm in length) on either side of the incision. Two film implants were placed per animal with one on the right and one on the left sections of the abdomen separated by 2 cm to prevent film interaction. PEU, ECM or PP were then fixed with 4-0 PGA sutures

(Ethicon) covering the hernia defects in an onlay position and the incision was closed with simple sutures. All animals were given a post-operative injection of saline (5 mL) for any fluid loss during the surgery. Animals were outfitted with small Elizabethan collars (Kent

Scientific) to prevent access to the incision for the first seven days. Survival rate was 100%

(12/12) for all time points (7 and 14 days). Film implants were collected postmortem for each time-point and subjected to further characterization.

4.3.13. Host-Implant Interaction

Films were collected post-mortem and fixed in paraformaldehyde for 24 h under ambient conditions. Samples were then processed and embedded in paraffin wax.

Embedded samples were sectioned, placed on microscope slides, fixed with PermountTM histology mounting media, and finally stained with Hematoxylin and Eosin (H&E). Slides were also analyzed by a board-certified veterinary pathologist for cellular infiltration.

78

Analyses were based on a modified scoring system outlined by the International

Organization for Standardization (ISO 10993-6 Annex E).

4.4. Results

4.4.1. Synthesis

Amino acid-based monomers were synthesized and characterized using 1H-NMR

(Appendix A Figure 6.7-6.9). 1,6-hexanediol and 1,8-octanediol were coupled to the carboxylic acid of L-valine or L-phenylalanine via an esterification using p-toluenesulfonic acid to prevent amidation (Scheme 4.1). The resulting monomers were named based on their diol chain length and amino acid; (1-VAL-8) formed from 1,8-octanediol and L-valine,

(1-PHE-6) formed from 1,6-hexanediol and L-phenylalanine, and (1-PHE-8) formed from

1,8-octanediol and L-phenylalanine. 1-PHE-6 and 1-PHE-8 can be differentiated from the integration of the methylene resonances at 1.06-1.14 ppm. 1-VAL-8 synthesis was confirmed based on 1H-NMR identification and integration of the characteristic L-valine methyl resonances at 0.96 ppm.

Poly(ester urea)s were synthesized via interfacial polymerization by combining 1-

VAL-8 with one of the two L-phenylalanine monomers (1-PHE-6 or 1-PHE-8) in varying molar ratios of 90:10, 80:20, and 70:30 (Scheme 4.1) adapted from previously published work.84,87 The incorporation of 1-PHE-6 or 1-PHE-8 monomers were chosen for stiffer mechanical properties and increased hydrophobicity compared to 1-VAL-8, as it is predicted that this enhanced initial mechanical properties will prevent device failure with in vivo degradation compared to previously studied L-valine PEUs.88 Polymer synthesis

79 was confirmed through 1H-NMR (Appendix A Figure 6.10-11) and 13C-NMR (Appendix A

Figure 6.12). The copolymer spectra (30% PHE6 P(1-VAL-8), 20% PHE6 P(1-VAL-8), 10%

PHE6 P(1-VAL-8), 30% PHE8 P(1-VAL-8), 20% PHE8 P(1-VAL-8), and 10% PHE8 P(1-VAL-8) were compared to P(1-VAL-8). Monomer incorporation and ratios were confirmed by 1H-

NMR integration of the methyl L-valine protons denoted “a” and L-phenylalanine methylene protons denoted “l” (Appendix A Figure 6.10-11).

Scheme 4.1. A general synthetic route for the monomer synthesis of 1-VAL-8, 1-PHE-6, and 1-PHE-8 carried out via a Fischer esterification between varying diol chain lengths and amino acids. In total six copolymers were synthesized with two of the monomers combining 1-VAL-8 and 1-PHE-6 in three different stoichiometric ratios to form 10% PHE6 P(1-VAL8), 20% PHE6 P(1-VAL-8), and 30% PHE6 P(1-VAL-8) or combining 1-VAL-8 and 1-PHE-8 to form 10% PHE8 P(1-VAL-8), 20% PHE8 P(1-VAL-8), and 30% PHE8 P(1- VAL-8). All polymers were synthesized using triphosgene to form urea linkages.

80

4.4.2. Physical Properties

Thermal properties of PEU copolymers were determined using thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) (Table 4.1). TGA curves (Figure

4.1 A) show high degradation temperatures (Td) which allow for these materials to be thermally processed and opens the door for other processing techniques such as compression molding and 3D printing. This is ideal as a robust material should have sufficient thermal stability for processing and storage/shelf-life. The observed mass loss for 30% PHE6 P(1-VAL-8) near 100-150 ºC is attributed to water. Size-exclusion chromatography (SEC) was performed to determine the molecular mass of the copolymer analogues (Figure 4.1 B). Molecular mass values (Mn and Mw) were calculated from linear polystyrene standards (Table 4.1) and were found to be comparable among the copolymers. The molar mass distributions (Đm) are less than the theoretical value of 2 because the values reported are post-fractionation. DSC was used to characterize the Tg,

(Figure 4.1 C). PHE6 P(1-VAL-8) analogues display a higher Tg than the PHE8 P(1-VAL-8) derivatives with values falling between 48-57 °C and 42-44 °C respectively. This was expected as the 1-PHE-6 analogues have a shorter diol-chain length which leads to decreased chain mobility and an observed increase in the Tg. Furthermore, all L- phenylalanine containing copolymers have Tg values equal to or greater than the p(1-val-

8) homopolymer (42 °C) as the aromatic side chain restricts chain mobility.88

81

Table 4.1. Physical properties of the PEU Copolymers

Tg Td Mn Mw Polymer Đm (°C) (°C) (kDa) (kDa) 10% PHE8 P(1-VAL-8) 42 297 49 84 1.7 20% PHE8 P(1-VAL-8) 44 287 68 105 1.5 30% PHE8 P(1-VAL-8) 42 278 69 108 1.6 10% PHE6 P(1-VAL-8) 53 268 44 60 1.4 20% PHE6 P(1-VAL-8) 57 278 56 88 1.6 30% PHE6 P(1-VAL-8) 48 340 57 95 1.7

Tg values are the glass transition temperatures reported from the TA Q200 thermal analysis instrument. Td was assessed at 10% mass loss for all polymer analogues. Number average molecular weight (Mn) and weight average molecular weight (Mw) are reported with the molar mass distribution (Đm) being calculated from those values.

4.4.3. Water Uptake and Contact Angle

Water uptake studies were conducted to predict hydrolytic degradation of PEU copolymers. Previously published work has shown that the rate of water uptake correlates with PEU degradation as the primary degradation mechanism is through hydrolytic cleavage of the ester bond.85 P(1-VAL-8) showed in vivo degradation that was more rapid than desired for hernia repair applications as more than a 50% loss in mechanical properties was observed over 3 months.88 Consequently, the incorporation of the hydrophobic L-phenylalanine side chain was hypothesized to slow degradation when compared to L-valine. All six PEU copolymer analogues and a P(1-VAL-8) control were studied to assess the effects 1-PHE-6 and 1-PHE-8 incorporation have on P(1-VAL-8) and how that relates to water uptake (Figure 4.1 D). Of the PHE6 P(1-VAL-8) analogues, the 20% PHE6 P(1-VAL-8) and 30% PHE6 P(1-VAL-8) display lower water uptake (5.03 ±

0.92 and 2.24 ± 2.66 % respectively) when compared to P(1-VAL-8) (13.99 ± 3.14 %). This

82 was ideal as the incorporation of 1-PHE-6 monomer slows water uptake and consequently hydrolytic degradation which should lead to sustained mechanical properties compared to P(1-VAL-8) for hernia-repair applications. Only the 10% PHE8 P(1-VAL-8) of the PHE8

P(1-VAL-8) analogues display lower water uptake (9.09 ± 0.87 %) when compared to the

P(1-VAL-8) control. While the L-phenylalanine plays a role in water uptake, diol chain length is a second competing factor. Increasing the diol chain length between amino acids increases chain mobility. The increase in chain mobility could manifest itself with an increase in water uptake. This was observed in the 10% PHE6 P(1-VAL-8), 20% PHE8 P(1-

VAL-8), and 30% PHE8 P(1-VAL-8) (25.46 ± 5.18, 25.61 ± 4.63, and 37.22 ± 4.94 % respectively). As more 1-PHE-8 is incorporated an increase in water uptake is observed as the interchain flexibility and packing is disrupted from the L-phenylalanine unit. These results show tunable water uptake properties for PEUs which is appealing as degradation rates will vary accordingly with water penetration into the sample. Further scattering characterization would be required to confirm the mechanism behind the observed water uptake trends.

To compare the surface properties of the six copolymers to previously studied P(1-VAL-8), water contact angle was measured (Appendix A Figure 6.13). An increase in the average angle was observed for all copolymers when compared to P(1-

VAL-8) (64.5 ± 5.0 – 71.5 ± 4.0° and 62.9 ± 3.1° respectively). This result was expected as the incorporation of the L-phenylalanine monomer increased the hydrophobic character of the film surface.

83

Figure 4.1. Physical properties of all six copolymers were assessed utilizing several techniques. Thermogravimetric analysis (TGA) (A) was performed to determine the degradation temperature (Td) for each polymer and curves were reported. The Td for each copolymer was high enough to allow for these materials to be thermally processed through compression molding. Size-exclusion chromatography (SEC) (B) for each copolymer was performed with molecular mass traces indicating uniform dispersity between polymers with Đm values between 1.4-1.7. Differential scanning calorimetry (DSC) (C) curves indicate that the glass transition temperatures (Tg) for these materials are above physiological conditions. The Tg are significantly lower than the degradation temperature which allowed for these materials to be thermally processed through compression molding without degradation. Finally, all six copolymers and p(1-VAL-8) were assessed to determine how incorporation of L-phenylalanine and change in diol change length would affect water uptake in 1 × PBS through eight days (n = 3) (D). An increase in L-phenylalanine led to a decrease in water uptake for the PHE6 P(1-VAL-8) polymers. The opposite trend was observed for the PHE8 P(1-VAL-8) polymers as the diol chain length and disruption of interchain packing was the dominating factor. 4.4.4. Mechanical Properties

84

4.4.4.1. Uniaxial Tensile Testing

Tensile testing was performed on all PEU copolymers, P(1-VAL-8), and PP with representative stress and strain curves (Figure 4.2 A) being reported. The Young’s moduli values were taken at 10% strain (Figure 4.2 B). Stress at yield (σy) (Figure 4.2 C) and strain at yield (εy) (Figure 4.2 D) were measured from the end of the linear elastic region and all values were reported (Table 4.2).

Table 4.2. Uniaxial mechanical properties comparison

Modulus σy εy Polymer (MPa) (MPa) (mm/mm) PP 216.1 ± 3.2 27.4 ± 0.2 0.2 P(1-VAL-8) 254.0 ± 3.9 34.4 ± 2.6 0.2 10% PHE8 P(1-VAL-8) 207.0 ± 24.9 27.2 ± 4.2 0.2 20% PHE8 P(1-VAL-8) 52.1 ± 63.0 4.4 ± 6.2 0.1 30% PHE8 P(1-VAL-8) 110.6 ± 89.8 11.6 ± 9.2 0.1 10% PHE6 P(1-VAL-8) 283.2 ± 17.5 31.6 ± 3.9 0.1 20% PHE6 P(1-VAL-8) 298.5 ± 28.2 29.1 ± 7.6 0.1 30% PHE6 P(1-VAL-8) 332.5 ± 3.5 40.5 ± 6.2 0.1 The Young’s modulus, stress at yield (σy) and strain at yield (εy) were measured and recorded from the stress-strain curves from uniaxial tensile testing. Values reported are an average of 4-6 samples.

When comparing moduli values 30% PHE6 P(1-VAL-8) was significantly higher compared to PP, highlighting the effect of the addition of 1-PHE-6 on mechanical properties. The increased initial Young’s modulus is ideal as degradation for these materials is expected upon implantation, which leads to a gradual decline in mechanical properties. For degradable devices, temporarily bolstered mechanical properties can be utilized as a way to maintain the required mechanical properties throughout the healing process prior to material degradation. 10% PHE8 P(1-VAL8), 20% PHE8 P(1-VAL-8), and

85

30% PHE8 P(1-VAL-8) each exhibit brittle behavior after compression molding manifested with a drop in Young’s modulus when compared to P(1-VAL-8). The relatively larger error bars observed for the 20% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL-8) species could be attributed to sample inhomogeneity from the compression molding method as inefficient interchain hydrogen bonding could have occurred in the temperature quenching step. All

PHE8 P(1-VAL-8) polymers exhibit lower moduli values than the PHE6 P(1-VAL-8) copolymers with 20% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL-8) being significantly lower than PHE6 P(1-VAL-8) counterparts. This was an expected result as PHE8 copolymer analogues have longer diol-chain lengths than the PHE6 copolymers resulting in a decrease in material stiffness. Similar trends were observed when comparing σy values for the PHE8 P(1-VAL-8) and PHE6 P(1-VAL-8) polymers with 20% PHE8 P(1-VAL-8) and

30% PHE8 P(1-VAL-8) being significantly lower than PHE6 P(1-VAL-8) analogues. While not significant, the average σy values for 10% PHE6 P(1-VAL-8) and 30% PHE6 P(1-VAL-8) are greater than PP. The εy for all PEU copolymers are all lower than PP with 20% PHE8

P(1-VAL-8), 30% PHE8 P(1-VAL-8), and 20% PHE6 P(1-VAL-8) being significant. The εy values for all PEU copolymers are not different from P(1-VAL-8) except for 20% PHE8 P(1-

VAL-8) which is significantly less. Increasing the stiffness of PEUs through the incorporation of PHE6 or PHE8 monomers without altering the εy values is favorable as maintaining linear viscoelasticity under physiologically relevant conditions is important for device success.

86

Figure 4.2. Uniaxial mechanical properties were assessed for all six copolymers and compared to the p(1-VAL-8) homopolymer and polypropylene. Tensile tests performed at 25 °C with a constant strain rate of 25.4 mm/min with all data extrapolated from the stress versus strain curves which are representative of 4-6 samples for each polymer (A). Young’s moduli values (B) for each polymer were extrapolated at 10% strain (*, **, ***, ****, ***** indicates a p value < 0.05 between a reference sample (first sample denoted with *, **, ***, or etc. reading from left to right) and other samples sharing like symbols (n = 4-6 samples)). For example, (* indicates p value < 0.05 between PP and 20% PHE8 P(1-VAL-8), between PP and 30% PHE8 P(1-VAL-8), and between PP and 30% PHE6 P(1- VAL-8)). * does not indicate a significant difference between 20% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL-8) as * indicates a difference from PP (reference). Yield stress (σy) (C) for each polymer was measured at the yield point (*, **, ***, ****, or ***** indicates p value < 0.05 between samples sharing like symbols. Statistical difference can be observed as previously described for moduli values. Yield strain (εy) (D) for each polymer was measured at the yield point (*, **, or *** indicates a p value < 0.05 between samples sharing like symbols. Statistical difference can be observed as previously described for moduli values.

87

4.4.4.2. Burst-Test Mechanics Composite Films

Small-intestine submucosa extra cellular matrix (SIS-ECM) is a currently employed material for hernia repair. While it is heralded for its promotion of native tissue growth

SIS-ECM suffers from degradation which can lead to mechanical property loss and hernia recurrence (e.g. SIS typically is resorbed in less than 9-12 months with some gradual loss of mechanic strength over time).66,75,102 Burst testing has previously been described as a relevant testing method for devices used in hernia repair.139 To enhance SIS-ECM, PEU copolymers were dissolved in acetone (Scheme 4.2 A) and blade coated onto SIS-ECM

(Scheme 4.2 B-C) to produce composite materials with improved mechanical properties

(Scheme 4.2 D). A clamp was outfitted with the hydrated composite PEU-ECM film, fixed with screw fixtures, and subjected to burst-testing (Scheme 4.3 A-D). Force versus extension curves were recorded (Scheme 4.3 E). Relative stiffness was defined as the ratio of force to extension at break for each sample. No increase in stiffness was observed when ECM was coated with the PEU copolymers (Appendix A Figure 6.14 A). This was attributed to the successful integration of the PEU solution into the SIS-ECM which is ideal as film stiffness may correlate with patient discomfort.140 Relative stiffness, force at break and extension at break were recorded (Appendix B Table 6.2). Force at break

(Appendix A Figure 6.14 B) is greater for all PEU-ECM copolymer analogues compared to

SIS-ECM with 10% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL8) being significantly greater.

Another improvement is observed in the extension required to rupture the composite films. Extension at break (Appendix A Figure 6.14 C) was recorded and although not

88 significant, all PEU-ECM copolymers have equal or greater extension compared to SIS-

ECM. PEU-ECM composite analogues possess enhanced elastic properties without increased stiffness to the device. This is significantly different from standalone SIS-ECM as the composite materials can undergo more strain prior to device failure. Blade coating

PEU copolymer films on SIS-ECM enhances the mechanical properties of SIS-ECM without increasing the relative stiffness, making these composite films an attractive candidate for hernia repair.

Scheme 4.2. Roll-to-roll processing film fabrication set up by first dissolving polymer pellets in an acetone solution (A). Solutions were allowed to dissolve overnight before being fed in to the solution well with the doctor blade. A polyethylene terephthalate (PET) substrate, coating direction, and processing speed (150 cm/min) were set on the Bio-Hybrid processing line (B & C). For composite films, ECM was further adhered to the 89

PET substrate and coated to afford the PEU-ECM film (D) while the free-standing PEU films were simply coated on PET.

4.4.4.3. Burst-Test Mechanics Free-Standing Films

While enhancing SIS-ECM films in a multi-material composite system is an attractive option, creating a new single material stand-alone film (Scheme 4.2 E) that fulfills hernia repair requirements would be preferred as manufacturing cost could be reduced. Burst testing was performed as previously described and the force versus extension curves are shown (Scheme 4.3 F) with relative stiffness, force at break and extension at break being reported (Appendix B Table 6.2). Relative stiffness (Appendix A

Figure 6.14 D) is decreased when compared to SIS-ECM which could make these films more comfortable to the patient. For the PEU free-standing films (Appendix A Figure 6.14

E) force at break does not change significantly from SIS-ECM; however, extension at break

(Appendix A Figure 6.14 F) for PEU free-standing films is far superior to that of SIS-ECM.

This is thought to be ideal as improved extendibility could prevent device failure as seen in SIS-ECM.

90

Scheme 4.3. Burst testing set up for composite PEU-ECM and free-standing PEU copolymer films. Burst-testing was adapted from ASTM D 3787-07 standards. Burst-test clamp surface was (A) cleaned prior to film fixation using tissue-paper. PEU-ECM composite films were submerged in 1 × PBS for 5 min prior to being placed on the bottom clamp (B). Films were then fastened (C) with a top clamp with two fixation screws. Burst- testing was performed (D) at an extension rate of 25.4 mm/min. The force versus extension curves were recorded for composite PEU-ECM films (E) and for free-standing films (F) (n = 3).

4.4.5. Cell Culture Studies

Cell viability is of utmost importance for any material seeking to enjoy success as a new material for soft-tissue hernia repair. PEU blade coated copolymers were seeded with a fibroblast phenotype mouse cell line harvested from subcutaneous connective tissue (L-929 cells ATCC) to assess cell viability. This cell line was selected as it is similar 91 to the type of connective tissue present that an implanted material would interact with during many hernia repair applications. All samples display good viability through 48 h of culture (all samples greater than 70% when compared to glass) with no statistical difference observed between the films and the glass control (Figure 4.3 A). To further understand cellular interaction with the copolymer films, cell attachment and spreading was measured on the PEU copolymer substrates and compared to a glass slide control with a cell density of 1.3 × 104 cells/cm2 for 48 h (Figure 4.3 B-H) with averages being greater when compared to the glass slide control (223 ± 61 - 365 ± 94 µm and 145 ± 38

µm respectively); however, none were significantly different on a 95% confidence interval

(Figure 4.3 I). A CYQUANT assay (Invitrogen) was utilized to determine the number of cells adhered to the substrates. Samples were stained with CYQUANT GR dye and fluorescence intensity was used to assess attachment compared to the glass slide control

(n = 5). All six copolymers showed good cell attachment when compared to glass (ratio determined by cell attachment on copolymer normalized by attachment on glass)

(attachment ratios of 0.8 ± 0.6 – 1.3 ± 0.9 and 1.0 ± 0.6 respectively) (Figure 4.3 J). 30%

PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8) showed a statistical difference between the glass slide control with confidence intervals of p < 0.01 and p < 0.001 respectively. These results indicate that roll-to-roll processable films are not only tolerated but rather provide a hospitable environment for cellular processes.

92

Figure 4.3. Cell viability was determined using L-929 cells seeded on PEU copolymers for 48 h (A). All slides were compared to a glass control with no statistical difference being observed between the groups. Results indicate that blade coated PEU films are non-toxic towards mammalian cells. Cell spreading and attachment results for L-929 fibroblast connective tissue cell line on PEU copolymer surfaces were assessed. Cells cultured on all six copolymers were stained with rhodamine phalloidin (red-pink color) and DAPI (blue- purple color). All pictograms were collected using an IX 81 Olympus microscope at 40 × magnification (B-H). Scale bars are equal to 100 µm. Cell spreading was calculated using ImageJ software (I). 30% PHE6 P(1-VAL-8) had the greatest spreading area although no copolymers were significantly greater than glass (n = 5). Cell attachment values were

93 normalized and statistically compared to the glass control slide with 30% PHE6 P(1-VAL- 8) (E) having a p < 0.01 and 20% PHE8 P(1-VAL-8) having a p < 0.001 (J).

4.4.6. Histological Assessment

A rat hernia model was used to assess the inflammatory response of 30% PHE6

P(1-VAL-8) compared to polypropylene and SIS-ECM. Surgical procedures and animal handling complied with The University of Akron Institutional Animal Care and Use

Committee (IACUC Protocol Number 17-12-16-BRD). The 30% PHE6 P(1-VAL-8) was selected because it showed promising mechanical properties comparable with the other copolymers, the lowest amount of water uptake, had the greatest cell spreading area, and showed significantly different cell attachment when compared to glass. 30% PHE6

P(1-VAL-8), SIS-ECM (Single-layer lyophilized ECM), and polypropylene were cut in to 1 cm diameter discs with thicknesses between 0.11 - 0.13 mm. A ventral incision was made, and a hernia defect was created on either side of the midline incision, film implants were fixed covering the defect, and the incision was closed (Figure 4.4 A-C). One rat showed a hernia bulge indicating that the abdominal incision model could successfully induce a hernia (Appendix A Figure 6.15).

94

Figure 4.4. A rat hernia model was carried out where a ventral incision was made (A), and a hernia defect was created on either side of the midline incision (B). Disc implants were fixed using PGA sutures to cover the hernia defect (C) on either side of the ventral incision. After two implants were placed per animal, the incision was closed with 4-5 sutures (D). Histology pictograms were obtained by staining implanted films in hematoxylin and eosin. Pictograms are oriented with the material on the right and abdominal wall (AW) on the left. Polypropylene (A), SIS-ECM (B), and 30% PHE6 P(1-VAL-8) (C) were imaged at 7 days and again at 14 days (G, H, and I respectively). Pictograms were captured at 20 × magnification with scale bars being equal to 1 mm.

95

The early inflammatory response for PEUs has not been documented as previous studies investigated inflammation at extended timepoints (several months).88 While the inflammatory responses observed were promising, it was unknown if this was attributed to the bioresorbable nature of the material or if PEUs had less inflammation early on in the healing process. Thus, cross-sections of the hernia implant with the surrounding tissue were stained with hematoxylin and eosin to assess the cellular infiltration and inflammatory response at 7 and 14 days. A fibrous capsule is easily identifiable on the polypropylene 7 day samples (Figure 4.4 D) with cells surrounding the material which is indicative of any implanted material as phagocytosis of the implant is attempted. This contrasts to the SIS-ECM material at 7 days (Figure 4.4 E) where cellular infiltration has occurred due to the porous nature of the implant which has been previously documented.141 30% PHE6 P(1-VAL-8) at 7 days (Figure 4.4 F) compares well to polypropylene with a clear encapsulation of the implanted material. All implants at 14 days had similar morphology at 20 × magnification when compared to 7 days (Figure 4.4

G-I). To quantify inflammation, slides were scored by a board-certified pathologist with a system from 0-4 based on the quantity of neutrophils, lymphocytes, plasma cells, macrophages, multinucleated giant cells, or necrosis present on each sample in a 400 × field (Table 4.3). Score totals indicate that 30% PHE6 P(1-VAL-8) (6.3) elicits the lowest amount of inflammation through 7 days when compared to polypropylene (7.4) and SIS-

ECM (10.3) (Table 4.4).

96

Table 4.3. H&E slide modified scoring scale based on ISO 10993-6 Annex E

Cell Type/Response 0 1 2 3 4

Rare, 1-5/400X 5-10/400X Neutrophils 0 field field Heavy Infiltrate Packed Rare, 1-5/400X 5-10/400X Lymphocytes 0 field field Heavy Infiltrate Packed Rare, 1-5/400X 5-10/400X Plasma Cells 0 field field Heavy Infiltrate Packed Rare, 1-5/400X 5-10/400X Macrophages 0 field field Heavy Infiltrate Packed

Multinucleated Giant Rare, 1-5/400X 0 3-5/400X field Heavy Infiltrate Sheets Cells field Moderate (50- Severe (75- Necrosis None Minimal (<25%) Mild (25-50%) 75%) 100%) Modified scoring system outlined by the International Organization for Standardization (ISO 10993-6 Annex E)

Table 4.4. Histology slide scores Polypropylene SIS-ECM 30% PHE6 P(1-VAL-8) Cell Type 7 Days 14 Days 7 Days 14 Days 7 Days 14 Days Neutrophils 1.7 3.3 1.8 1.7 1.0 3.3 Lymphocytes 2.0 3.3 2.5 2.7 2.0 2.8 Plasma Cells 2.0 3.0 2.5 2.0 1.3 2.3 Single Macrophages 1.7 3.0 2.0 2.0 2.0 3.0 Multinucleated Giant Cells 0.0 0.7 1.5 2.0 0.0 1.0 Necrosis 0.0 0.7 0.0 0.0 0.0 0.8 Score Totals 7.4 14.0 10.3 10.4 6.3 13.2 Reported scores are the average of 4-6 sections

As expected, the SIS-ECM has significant cellular integration leading to high score totals which is not always indicative of a poor remodeling process.100 However, more multinucleated giant cells are present with SIS-ECM at 7 and 14 days when compared to polypropylene and 30% PHE6 P(1-VAL-8) which is commonly thought to be indicative of a foreign body response.124 This increased amount of inflammation can correlate to more connective tissue ingrowth, however, it has been shown that this doesn’t necessarily 97 correlate to strength at the hernia repair site.124 This rapid cellular integration also undoubtedly contributes to the premature loss in mechanical properties for ECM.

Strategies to reinforce biological grafts like human acellular dermal matrix grafts with polypropylene have been used with success for ventral hernia repair.142 Unfortunately, the implant is no longer fully resorbable with the reinforcement. Ideally, a degradable coating as shown with PEUs would be preferred as long-term inflammation would be limited because of the resorbable nature of the implant. The immune response to polypropylene and 30% PHE6 P(1-VAL-8) consists mostly of mononuclear cells. This has previously been shown with cells following neutrophils to the implant site over time attempting phagocytosis and clearing debris, culminating with phagocytosis fatigue or implant degradation.100 Cellular recruitment to the implant site is why the score numbers for polypropylene and 30% PHE6 P(1-VAL-8) increase from 7 to 14 days. Although 30%

PHE6 P(1-VAL-8) and polypropylene have an increase in cell numbers from 7 to 14 days, there are less multinucleated giant cells present when compared to SIS-ECM which is the only implant to have multinucleated giant cells at the 7 day timepoint.55 No material exhibited necrosis through the 7 day timepoint however, there is an increase in single macrophages and multinucleated giant cells as well as low levels of necrosis at 14 days.

Additionally, there is heavy infiltration of neutrophils, lymphocytes, and plasma cells around the polymer implants. The total score for SIS-ECM increases slightly from 10.3 to

10.4 from 7 days to 14 days with no observable necrosis. The lack of change and mild infiltration from 7 to 14 days for SIS-ECM is attributed to the quiescence of the material

98 after infiltration as a result of inherent growth factors found on the ECM matrix which play a role in cellular signaling processes.13,17,27 It is promising that the 30% PHE6 P(1-

VAL-8) implant displays score totals less than that of polypropylene at both the 7 day and

14 day timepoints as it promotes the translatable use of the PEU copolymer free-standing films in hernia-repair applications. It has been well documented that polypropylene can elicit a chronic foreign body response over time and lead to poor clinical outcomes.55,124

One dominant strategy has been to alter the weight of the mesh implanted to limit the amount of material present thereby limiting inflammation.89 While an improvement in tissue integration is observed, a permanent implant still remains. Additionally, a great amount of effort has been made to modulate the inflammatory response on polypropylene with degradable materials including ECM, polyglycolic acid, and cellulose.142,143 While these degradable modulatory factors have found success over a short period of time to mitigate the inflammatory response, these efforts are often emblematic of a larger problem. Regardless what polypropylene is coated with, the test of time will still reveal the material as nonresorbable which will dictate the implant-tissue interaction.100 Fully degradable films like poly(ester urea)s in this study and previously explored materials like PHASIX (poly-4-hydroxybutyrate) will not experience as many inflammatory challenges as the degradative nature of the material precludes the implant from a long term foreign body response.67 Additionally, PEUs do not elicit the same amount of cellular integration throughout the entire sample as observed with ECM which may help strike a better balance between tissue integration and normal healing. These

99 results indicate that 30% PHE6 P(1-VAL-8) could be helpful in augmenting the mechanical properties of SIS-ECM as a composite material or as a standalone material with a more attenuated inflammatory response compared to polypropylene.

4.4.7. Film Degradation

Hydrolytic degradation rates for SIS-ECM, 30% PHE6 P(1-VAL-8)-ECM and 30%

PHE6 P(1-VAL-8) films were assessed in accelerated in vitro studies (Appendix A Figure

6.16 A-B). When the SIS-ECM is coated with 30% PHE6 P(1-VAL-8), the degradation rate is significantly slowed through 48 h. Additionally, the free-standing 30% PHE6 P(1-VAL-8) film degradation rates are slower than the SIS-ECM and 30% PHE6 P(1-VAL-8) despite being in a more caustic solution. Both of these results are desirable as using 30% PHE6

P(1-VAL-8) could operate as a composite film with SIS-ECM helping to enhance the shortcomings of SIS-ECM or as a free-standing film with slowed degradation compared to

SIS-ECM. Prior to implantation of SIS-ECM, polypropylene, and 30% PHE6 P(1-VAL-8), films were sterilized using ethylene oxide gas (EtO) and film morphology was noted using

SEM (Figure 4.5 A-F). Polypropylene and 30% PHE6 P(1-VAL-8) films are predominantly smooth as would be expected after fabrication and sterilization. The porous structure of

SIS-ECM can be observed which is emblematic as to why cellular integration is so high for this material as noted from the histological assessment. Through 14 days the surface topology of polypropylene remains unchanged which is expected from the nonresorbable material. ECM has lost the majority of its porosity as cells and bodily fluid have integrated in to the mesh. The 30% PHE6 P(1-VAL-8) displays little degradation, however some

100 pitting is observed across the sample which is characteristic of the hydrolytic and enzymatic degradation of PEUs. Additionally, molecular mass was obtained for the implanted 30% PHE6 P(1-VAL-8) through 14 days (Appendix A Figure 6.17). A narrowing of the dispersity and a slight increase in molecular mass is observed which is consistent with previously published results.88 This result is attributed to lower molecular mass chains aggregating towards the surface, undergoing degradation and leaving behind a more narrow distribution of higher molecular mass chains behind.

Figure 4.5. Film surface morphology for each material was characterized using scanning electron microscopy (SEM) at 200 × magnification with scale bars equal to 10 µm. Polypropylene SIS-ECM, and 30% PHE6 P(1-VAL-8) films were analyzed after ethylene oxide (EtO) sterilization (A, B, and C respectively). Polypropylene and 30% PHE6 P(1-VAL- 8) are noticeably smooth while ECM has cavities and undulations expected with its porous structure. After 14 days in vivo implantation, samples were analyzed again with polypropylene looking unchanged (D), ECM having thick tissue accumulation (E), and 30% PHE6 P(1-VAL-8) having small defects start to appear (F).

101

4.5. Conclusion

Current biologic hernia-mesh devices leave much to be desired in sustained mechanical support. A series of L-valine-co-L-phenylalanine PEU copolymers were synthesized and their mechanical properties were assessed to determine their use in composite films with ECM or as free-standing hernia films. These polymers are easily processable via roll-to-roll film casting methods rendering them scalable and industrially relevant. The mechanical properties of these PEU copolymers demonstrate improved mechanical properties when used concomitantly with SIS-ECM as composite materials or as free-standing films thus improving on the state of the art currently employed materials for hernia repair. Additionally, these polymer materials show great in vitro cell compatibility when compared to glass slides while eliciting a lower inflammatory score in a rat hernia model in vivo through 7 and 14 days compared to polypropylene. Overall,

PEU copolymers have the potential to be translatable materials as standalone films or as composite materials with SIS-ECM for hernia-repair applications.

4.6. Acknowledgement

The authors gratefully acknowledge financial support from Cook Biotech, the NSF

Research Experience for Undergraduates (DMR# 1359321) in the College of Polymer

Science and Polymer Engineering and the W. Gerald Austen Endowed Chair in Polymer

Science and Polymer Engineering via the John S. and James L. Knight Foundation. The authors would like to thank Dr. Walter Horne for training and advice with the surgical

102 procedure. Additionally, we acknowledge the Ohio Department of Development funded

Akron Functional Material Center for the Bio-Hybrid Roll-to-Roll Processing Line.

103

CHAPTER V

ZWITTERIONIC AMINO ACID POLY(ESTER UREA)S SUPPRESS ADHESION FORMATION IN A

RAT INTRA-ABDOMINAL CECAL ABRASION MODEL

In part, this work has been submitted to Biomaterials as Dreger, N. Z.; Zander, K.

Z.; Hsu, Y.; Chen, P.; Le, N.; Parsell, T.; Søndergaard, C.; Suckow, M. D.; Hiles, M.; Becker,

M. L., Zwitterionic Amino Acid Poly(ester urea)s Suppress Adhesion Formation in a Rat

Intra-abdominal Cecal Abrasion Model, 2019, Submitted.

5.1. Abstract

Surgical repair of hernia’s has improved with more robust material options for surgeons and optimized surgical techniques. Ventral hernia repairs remain challenging with an inherent risk of post-surgical adhesions in the peritoneal space which can occur regardless of interventional material or surgical placement. Herein, anti-adhesion amino acid-based poly(ester urea)s (PEUs) with varied amount of an allyl ether side chain for post polymerization modification (5% alloc-PEU or 10% alloc-PEU) with zwitterionic (3-

((3-((3-mercaptopropanoyl)oxy)propyl) dimethylammonio)propane-1-sulfonate) used to prevent adhesions. These alloc-PEUs were processed through roll-to-roll fabrication methods to afford films which were further surface functionalized with a zwitterion-thiol.

Functional group availability on the surface was confirmed via fluorescence microscopy,

104 x-ray photoelectron spectroscopy (XPS), and quartz crystal microbalance (QCM) measurements. Anti-adhesion PEUs reduce fibrinogen adsorption in vitro when compared to unfunctionalized controls. A rat intrabdominal cecal abrasion adhesion model was used to assess the extent and tenacity of adhesion formation in the presence of the PEUs. The 10% alloc-PEU zwitterion functionalized material was found to reduce the extent and tenacity of adhesions when compared to adhesion controls and the 10% alloc-PEU blank material.

5.2. Introduction

Treatments in ventral hernia mesh repair have improved over the past several decades with the use of polymeric materials to augment a defect in open or laparoscopic ventral hernia repairs. This mature innovation has resulted in recurrence rates lower than 20%.1–

3 While significant strides have been made, the risk of postoperative adhesions between the device and the surrounding tissues remain prevalent. Ideally, the implanted device would simply cover the defect in the abdominal wall and augment the required mechanical properties. However, when non-natural materials are implanted into the body, there is an immediate deposition of proteins, albumin, cell signaling factors, and ultimately extracellular matrix polysaccharides.4–6 Cells lay down on the newly formed surface with immune cells attempting to identify toxicity, biocompatibility, and wall off the implant from the rest of the body with a fibrous capsule. This is part of the natural immune response to any foreign material and is part of the healing process.

Unfortunately, there is a chance that the mesh can become adhered to other layers inside

105 the peritoneum space.7–9 Depending on the mesh selected, the outcomes and chance of problematic adhesion can vary.10–13

Standard mesh selection criteria are generally based on surgeon preference and material availability. The types of mesh available clinically are broken down in to two main categories: resorbable or nonresorbable. Nonresorbable materials are designed to be permanent and are not degraded in vivo which can lead to a chronic inflammatory response.14–16 While this response is often tolerated, permanent implants are also prone to the formation of adhesions with various organs as the fibrous capsule possesses the ability to adhere to other intraabdominal sites.14,17,18 Alterations in surgical technique

(i.e. inlay versus onlay) influence adhesion formation, however there is no full proof way to mitigate their formation.19 The presence of these adhesions can lead to complications years after a hernia surgery. Attempts to reduce adhesions to surrounding tissue have been attempted by altering the surface chemistry of the implant.13,20 Examples including coating non-resorbable meshes in silicone to reduce surface tension have had moderate success.21 In other fields, anti-adhesion strategies using polyethylene glycol (PEG) and derivatives thereof have been used to prevent adhesions however, their degradation products are not ideal.22 More exotic chemistry techniques have been employed in the attempt to create anti-fouling and anti-adhesive surfaces with some success in vitro.

However, these surface chemistries are often challenging to scale.5,23,24 While the optics of these design strategies appear to be steps in the right direction, they have not lead to a clinical solution.

106

Resorbable materials currently used in hernia-mesh repair are advantageous from a tissue remodeling standpoint. As the implant degrades, cellular integration and remodeling occurs leaving behind native tissue.25–27 One of the most widely used resorbable materials for hernia repair is decellularized extracellular matrix (ECM), generally from xenographic sources. This material has suitable mechanical properties while also being porous allowing for cellular integration and tissue remodeling.25–29 The increase in tissue remodeling does not prevent this mesh type from adhesions. Depending on the location and time frame of adhesion formation, a resorbable material has the potential to overcome an adhesion without pain or clinical failure. While adhesions can still occur during the lifetime of a resorbable implant, the material will ultimately degrade, and the adhesion has the chance to resolve without fusing the abdominal wall to an organ in the peritoneal space.30,31 Unfortunately, if the adhesion occurs to an organ, the implant may need to be removed prior to resolution.32,33 This could lead to complications including herniation, strangulation, and repeat surgeries.

It would be advantageous to combine the antiadhesion properties present on select non-resorbable meshes with resorbable materials. An evolving class of resorbable materials are amino acid-based poly(ester urea)s (PEUs). PEUs have been shown to have tunable mechanical properties, limited inflammatory response in vivo, and lend themselves to facile roll-to-roll fabrication methods.34–42 These materials have significant clinical promise as they possess mechanical properties in the range of soft tissue and of currently employed hernia meshes.43 By varying monomer composition and

107 stoichiometry, PEUs have been derivatized post-polymerization with thiol containing compounds using a thiol-ene radical addition which has enabled them to be explored as potential antiadhesion barriers for hernia repair.

5.3. Experimental

5.3.1. Materials

1,8-octanediol, 1,6-hexanediol, sodium carbonate, p-toluenesulfonic acid monohydrate, 3-allyloxy-1,2-propanediol, 1,3-diisopropyl carbodiimide, and triphosgene were all purchased from Sigma Aldrich (Milwaukee, WI). Toluene, ethyl acetate, hexane, chloroform, acetone, and N,N-dimethylformamide were purchased from Fischer Scientific

(Pittsburgh, PA). l-valine, Boc-O-benzyl-l-tyrosine, and l-phenylalanine were purchased from Acros (Pittsburgh, PA). All solvents were reagent grade and all chemicals were used without further purification unless otherwise stated.

5.3.2. Characterization

Proton (1H) NMR and Carbon (13C) spectra were collected using 300 MHz and 500

MHz Varian NMR spectrometers, respectively. Chemical shifts are reported in ppm (δ)

1 and referenced to residual solvent resonances ( H NMR DMSO-d6 2.50 ppm).

Multiplicities were explained using the following abbreviations: s = singlet, d = doublet, t

= triplet, br = broad singlet, and m = multiplet. Size exclusion chromatography (SEC) was performed using an EcoSEC HLC-8320GPC (Tosoh Bioscience) equipped with a TSKgel

SuperH-RC 6.0 mml.D. × 15 cm mixed bed column and refractive index (RI) detector. The number-average molecular mass (Mn), weight- average molecular mass (Mw), and

108 molecular mass distribution (ĐM) for each sample was determined using a calibration curve generated from poly(styrene) standards (PStQuick MP-M standards, Tosoh

Bioscience) with THF as eluent flowing at 1.0 mL/min at 50 °C. Differential scanning calorimetry (DSC) was performed using TA Q200 DSC (Thermal Analysis) with heating and cooling cycles of 20 °C/min and 10 °C respectively with temperature sweeps from 0-100

°C. The glass transition temperature (Tg) was determined from the midpoint of the second heating cycle curve. Thermogravimetric analysis (TGA) was performed using a TA

Q50 (Thermal Analysis) with heating ramps of 20 °C/min in the temperature range from

0-500 °C/min. The degradation temperature (Td) was determined from 10% mass loss. A

Rame-Hart Contact Angle Goniometer was used to determine contact angle for 5 µL of deionized water on a sample surface. Statistical analyses were performed using a post- hoc Tukey ANOVA test unless otherwise stated.

5.3.3. Synthesis of Poly(ester urea) monomers

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-valine)-Octane 1,8-Diester

Monomer. (1-VAL-8). Synthesis of di-p-toluenesulfonic acid salts of bis(l-valine)-octane

1,8-diester (1-VAL-8) was carried out following procedures published previously.44

Briefly, in a 1 L 1-neck round bottom flask, 1,8-octanediol (43.8 g, 0.3 mol, 1 eq.), l-valine

(73.8 g, 0.63 mol, 2.3 eq.), p-toluenesulfonic acid monohydrate (131.3 g, 0.69 mol, 2.4 eq.), and toluene (1300 mL) were added and equipped with a stir bar. A Dean-Stark trap was fastened to the round bottom flask and the reaction was heated to reflux for 24 h.

The reaction was cooled to ambient temperature, and the resulting white precipitate was

109 isolated by vacuum filtration using a Buchner funnel. The product was dissolved in boiling water (1 L), hot vacuum filtered, and cooled to ambient temperature to further purify the white solid precipitate. The precipitate was collected via filtration and recrystallized

1 three. times for purity (82% yield). H NMR (300 MHz, 303 K, DMSO-d6): δ = 0.94 (m, 12H),

1.28 (s, 8H), 1.59 (m, 4H), 2.07-2.18 (m, 2H), 2.27 (s, 6H), 2.50 (m, DMSO), 3.33-3.38 (s,

3 3 H2O), 3.89 (d, JH-H = 3.0 Hz, 2H), 4.07-4.23 (m, 4H), 7.07-7.23 (d, JH-H = 8.2 Hz, 4H,

3 aromatic H ), 7.45-7.48 (d, JH-H = 8.1 Hz, 4H, aromatic H), 8.25 (br, 6H) ppm.

Synthesis of Di-p-toluenesulfonic Acid Salts of Bis(L-phenylalanine)-Hexane 1,6-

Diester Monomer. (1-PHE-6). Synthesis of di-p-toluene sulfonic acid of bis(l- phenylalanine)-hexane 1,6-diester (1-PHE-6) was carried out using the method described

1 above (81% yield). H NMR (300 MHz, 303 K, DMSO-d6): δ =1.06 (s, 4H), 1.38 (m, 4H),

3 3 2.27 (s, 6H), 2.50 (m, DMSO), 2.96−3.17 (m, 4H), 4.01 (t, JH-H = 9.0 Hz , 4H), 4.28 (t, JH-

H=6.0 Hz, 2H), 7.08−7.11 (d, 4 H), 7.20−7.35 (m, 10H), 7.45−7.48 (d, 4H), 8.37 (s, 6H) ppm.

Synthesis of Di-Hydrochloric Acid Salts of Bis-O-benzyl-L-tyrosine-1,3-Allyloxy-

Diester Monomer. (1-TYR-2 Alloc). Synthesis of di-hydrochloric acid salts of bis-O-benzyl-

L-tyrosine-1,3-allyloxy-diester monomer (1-TYR-2 Alloc) was carried out following previously published procedures.87 1-TYR-2 Alloc was prepared using Boc-O-benzyl-L- tyrosine and 1,3-allyoxyl-2-propanediol in anhydrous DMF. Once dissolved, the reaction was placed in an ice bath for 10 minutes followed by syringe addition of 1,3-diisopropyl carbodiimide (DIC, 5 eq.). The reaction gradually came to ambient temperature while stirring for 24 h, as a yellow precipitate formed. DMF was removed under reduced

110 pressure and the product was purified using column chromatography (4:1 hexanes: ethyl

1 acetate) with all fractions collected for rotary evaporation. H NMR (300 MHz, DMSO-d6):

δ = 1.31 (s, 18H), 2.50 (s, DMSO), 2.77 (m, 4H), 2.93 (m, 2H), 3.35 (s, H2O), 4.01 (m, 2H),

4.06 (s, 1H), 4.14 (s, 2H), 4.26 (s, 2H), 5.04 (s, 4H), 5.21 (m, 2H), 5.85 (m, 1H), 6.89-7.38

(m, 18H, aromatic H). Boc protecting groups were removed with 4 M HCl/dioxane under nitrogen. The yellow product was dried under reduced pressure to yield the 1-TYR-2 Alloc

1 monomer. (71% yield). H NMR (300 MHz, DMSO-d6): δ = 2.50 (s, DMSO), 3.09 (m, 6H),

3.38 (s, H2O), 3.90 (s, 2H), 4.07 (m, 1H), 4.22 (m, 2H), 4.26 (m, 2H), 5.05 (s, 4H), 5.19 (m,

2H), 5.83 (m, 1H), 6.93-7.43 (m, 18H, aromatic H), 8.82 (br, 6H).

5.3.4. Lithium Phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) Photoinitiator

Synthesis.

LAP was synthesized according to previously published work.46 Briefly, dimethyl phenylphosphonite (1.00 eq) was added to an oven dried flask under argon at 23 °C.

While stirring, 2,4,6-trimethylbenzoyl chloride (1.00 eq) was added dropwise to the solution and allowed to stir for 24 hours. A four-fold excess of LiBr (6.1 g) in ca. 100 mL of 2-butanone was added to the reaction mixture and heated to 50 °C for 1 h until a white precipitate formed. The solution was cooled to room temperature, filtered under reduced pressure, and finally rinsed three times with 2-butanone to remove excess LiBr.

The solid white precipitate (82% yield) was dried under vacuum and analyzed by 1H-NMR.

1 H-NMR (300 MHz, 303K, D2O): δ = 2.01 (s, 6H), 2.23 (s, 3H), 6.88 (s, 2H), 7.41 – 7.51 (m,

2H), 7.51 – 7.61 (m, 1H), 7.70 (m, 2H) ppm.

111

5.3.5. Synthesis of 3-((3-((3-mercaptopropanoyl)oxy)propyl)dimethylammon-

io)propane-1-sulfonate (Zwitterion-SH).

The synthesis of 3-((3-((3-mercaptopropanoyl)oxy)propyl)dimethylammon- io)propane-1-sulfonate (zwitterion-SH) was carried out in several steps. First, the synthesis of 3,3’- dithiodipropionyl chloride was carried out in a 250 mL 2 neck round bottom flask. The flask was attached with a jacketed condenser and an addition funnel.

3,3’-dithiodipropionic acid (1.0 eq.) was added and purged with nitrogen at room temperature followed by the dropwise addition of thionyl chloride (6.0 eq.). The reaction was heated (90 °C) for 24 h and a color change to yellow was noted. Excess thionyl chloride was removed using a vacuum transfer and the residual product was dried under reduced pressure. The product was stored under anhydrous conditions without further

1 3 3 purification. H-NMR (300 MHz, 303 K, CDCl3): δ = 2.95 (t, JH-H = 7.0 Hz, 4H), 3.32 (t, JH-H

= 7.0 Hz, 4H) ppm. Synthesis of bis(3-dimethylamino)propyl) 3,3’- disulfanediyldipropionate was carried out in an anhydrous 100 mL, 2 neck round bottom flask. 3-dimethylamino-1-propanol (2.0 eq.) was added through a rubber septum and subsequently dissolved in 25 mL of anhydrous dichloromethane under a nitrogen atmosphere. 3,3’- dithiodipropionyl chloride was added (1.0 eq.) slowly to the flask and allowed to stir. Following the addition, a jacketed condenser was attached to the flask and the reaction was heated to reflux for 16 h under a nitrogen atmosphere. The solution was allowed to cool, and dichloromethane was removed under reduced pressure. The product was dissolved in a saturated sodium bicarbonate solution for 15 minutes. The

112 sodium bicarbonate solution was transferred to a separatory funnel with dichloromethane where the organic layer was collected (4 ×). The organic layer was then extracted in saturated sodium bicarbonate aqueous solution twice and finally distilled water once with the final organic layer stirred over magnesium sulfate for 15 minutes.

The product was then filtered, and residual solvent was removed under reduced pressure

1 3 to obtain a tan oil (77% yield). H-NMR (300 MHz, 303 K, DMSO-d6) δ = 4.08 (q, JH-H = 6.7

3 3 3 Hz, 4H), 2.92 (t, JH-H = 6.9 Hz, 4H), 2.70 (t, JH-H = 6.9 Hz, 4H), 2.26 (t, JH-H = 7.0 Hz, 1H),

3 13 2.12 (s, 12H), 1.71 (hept, JH-H = 6.3 Hz, 1H). C-NMR (75 MHz, 303 K, DMSO-d6) δ =

170.89 (s, 2C), 62.50 (s, 2C), 55.43 (s, 2C), 44.97 (s, 2C), 33.39 (s, 2C), 32.76 (s, 2C), 26.24

(s, 2C).

The zwitterion disulfide synthesis was carried out in a 100 mL round bottom flask where 1,3-propane sultone (4.0 eq.) was dissolved and stirred in acetone (5 mL) at 0 °C.

Separately, bis(3-dimethylamino)propyl) 3,3’-disulfanediyldipropionate (1.0 eq.) was dissolved in acetone (15 mL) and added slowly to the 100 mL round bottom flask. The reaction was allowed to stir and warm up to ambient temperature for 4 h where the white product was collected in the flask. Following the reaction, the product was collected in centrifuge tubes, washed with acetone, centrifuged (5 ×), and finally the white product

1 3 was placed under reduced pressure (87% yield). H-NMR (300 MHz, d2o) δ 4.33 (t, JH-H =

3 3 5.8 Hz, 4H), 3.64 – 3.49 (m, 8H), 3.21 (s, 12H), 3.07 (dt, JH-H = 12.7, 6.2 Hz, 8H), 2.94 (t, JH-

13 H = 6.4 Hz, 4H), 2.39 – 2.19 (m, 8H). C-NMR (75 MHz, d2o) δ 174.16 (s, 1C), 62.36 (s, 1C),

61.85 (s, 1C), 61.27 (s, 1C), 50.81 (s, 1C), 47.25 (s, 2C), 33.55 (s, 1C), 32.53 (s, 1C), 21.67

113

(s, 1C), 18.20 (s, 1C). Finally, the zwitterion-SH was obtained by reducing the zwitterion disulfide using 1,4-dithiothreitol (DTT). Zwitterion disulfide (1 eq.) and DTT (4 eq.) were added to a sodium bicarbonate solution (10 mM, pH = 7.4) with nitrogen bubbling through the solution for 4 h. The reaction was quenched by the addition of acetic acid to reduce the pH to 6.5. Water was removed under reduced pressure followed by a washing step with acetone to remove DTT, and centrifugation to afford the white product. 1H-NMR

3 3 (300 MHz, d2o) δ 4.33 (t, JH-H = 5.8 Hz, 4H), 3.63 – 3.51 (m, 8H), 3.21 (s, JH-H = 19.8 Hz,

3 3 3 12H), 3.07 (q, JH-H = 6.9 Hz, 8H), 2.94 (t, JH-H = 6.4 Hz, 4H), 2.29 (tt, JH-H = 15.5, 7.9 Hz,

13 8H). C-NMR (75 MHz, d2o) δ 174.29 (s, 1C), 62.31 (s, 1C), 61.70 (s, 1C), 61.33 (s, 1C),

50.76 (s, 1C), 47.23 (s, 2C), 37.91 (s, 1C), 21.69 (s, 1C), 19.02 (s, 1C), 18.18 (s, 1C).

5.3.6. Synthesis of Poly(ester urea) Terpolymers.

The synthesis was adapted from methods published previously.47 The Interfacial polymerization of p-toluenesulfonic acid salts of bis(l-valine), p-toluenesulfonic acid salts of bis(l-phenylalanine), hydrochloric acid salts of bis-O-benzyl-l-tyrosine-1,3-allyloxy- diester monomers ((1-VAL-8), (1-PHE-6), and (1-TYR-2 Alloc) respectively) were performed by dissolving the monomers in various molar feed ratios (1 eq. total) with sodium carbonate (3.4 eq.) in distilled water (0.1 M, 35 °C) in a 2 L 2-neck round bottom flask. The solutions were stirred mechanically until the monomers dissolved.

Triphosgene (0.35 eq.) was dissolved in chloroform and subsequently added to the reaction vessel through an addition funnel. The clear solution turned white upon addition and continued to stir for 4 h. The reaction was transferred to a separatory funnel where

114 the organic phase was precipitated in to boiling water to remove chloroform and starting material impurities. The off white polymer was collected, frozen in liquid nitrogen, and lyophillized under reduced pressure (90-95 % yield).

1 Poly[(1-VAL-8)0.65-co-(1-PHE-6)0.30-co-(1-TYR-2 Alloc)0.05]. (5% Alloc-PEU). H NMR

(300 MHz, 303 K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -

NHCH(CH(CH3)2)C(O)O-), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -

NHCH(CH2Ph) C(O)O-), 5.02 (s, 4H, -OCH2Ph), 5.13 (m, 2H, -CHCH2), 5.84 (m, 1H, -CHCH2),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05, 6.88,

7.38 (all remaining protons) ppm. (Mw = 35 kDa, Mn = 24 kDa, Đm = 1.5, Tg = 40 °C, Td = 289

°C).

1 Poly[(1-VAL-8)0.60-co-(1-PHE-6)0.30-co-(1-TYR-2 Alloc)0.10]. (10% Alloc-PEU). H

NMR (300 MHz, 303 K, DMSO-d6): δ = 0.83 (m, 12H, -CH(CH3)2), 1.95 (m, 2H, -

NHCH(CH(CH3)2)C(O)O-), 2.82 – 2.97 (m, 4H, -NHCH(CH2Ph)C(O)O-), 4.38 (m, 2H, -

NHCH(CH2Ph) C(O)O-), 5.02 (s, 4H, -OCH2Ph), 5.13 (m, 2H, -CHCH2), 5.84 (m, 1H, -CHCH2),

3 3 6.36 (d, JH-H = 9 Hz, 2H, -NHCH(CH(CH3)2)C(O)O-), 6.48 (d, JH-H = 9 Hz, 2H, -

C(O)NHC(CH2Ph)HC(O)-), 7.13-7.28 (m, 10H, -C6H5), 1.18, 1.24, 1.44, 1.51, 3.95-4.05, 6.87,

7.40 (all remaining protons) ppm. (Mw = 62 kDa, Mn = 28 kDa, Đm = 2.2, Tg = 49 °C, Td = 299

°C).

5.3.7. Terpolymer PEU Free-Standing Films

115

PEU free-standing films were prepared by blade coating with a few adaptations from procedures reported previously.44 PEU analogues were dissolved in acetone (35% weight) and the polymer solutions were blade coated (8 cm blade width, 1000 µm gap height) on a poly(ethylene terephthalate) (PET) carrier film and allowed to air dry for 24 h. The PEU films were further dried under reduced pressure to remove residual solvent

(film thickness = 200 µm). Films were then cut into 2 × 8 cm sheets and sterilized using ethylene oxide. Following sterilization, films were cut in to tensile bars using a custom dye cutter (ASTM D-638-V). Bars were pulled with a constant rate of extension (25.4 mm/min). The modulus, yield strain, and yield stress for each tensile bar were calculated from the linear elastic region. Statistical analyses were done using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.05 was considered significant.

5.3.8. FITC-PEG-Thiol Surface Functionalization of Terpolymer PEUs

Blade coated PEUs containing alloc functional groups were punched in to 8 mm disks using a dye punch set. Disks were placed in a 24-well plate and divided in to three groups. Discs possessing the alloc functional groups had the aqueous solution (100 µL) containing FITC-PEG-SH (2 eq.) and LAP (1.0 eq.) added to the surface. Disks were soaked in the aqueous solution for 30 minutes. The disks were treated with UV light (λ = 365 nm, I = 0.4 mW/cm2) for 30 minutes. Physically adsorbed disks had an aqueous solution containing FITC-PEG-SH (1000 MW) (2 eq.) and LAP (1.0 eq.) placed on the surface and were kept in the dark for 30 minutes. Finally, blank samples had DI water placed on the surface for 30 minutes. After treatment, all samples were rinsed with deionized water

116 three times and then submerged in 2 mL of deionized water (2 ×) for 30 minutes. Disks were removed, dried with nitrogen, and finally dissolved in DMSO (0.5 mL) for fluorescence measurements. For surface quantification, fluorescence studies were

TM carried out using a BioTek Synergy Mx Microplate Reader (BioTek, Vermont) (λex = 495 nm, emission range 530-550 nm). A FITC-PEG-SH (1000 MW) standard curve was constructed using serial dilutions in DMSO (λem max = 550 nm). Once dissolved, 300 µL of

DMSO from each group was added to a 96 quartz well plate and subjected to the plate reader (n = 3). Dye concentration was extrapolated from the calibration curve. Statistical analyses were done using a one-way ANOVA with Tukey post hoc analysis. A value of p <

0.05 was considered significant.

5.3.9. Zwitterion Surface Functionalization of Terpolymer PEUs

Functionalizing alloc-PEU disks with the thiol containing zwitterion-SH were performed as described above for FITC-PEG-SH. Disks were placed in a 24-well plate and submerged in deionized water (2 mL) to presoak for 30 minutes. Disks were divided in to three groups. Functionalized samples had the aqueous solution removed and replaced with an Argon purged aqueous solution containing zwitterion-SH (2 eq.) and LAP (1.0 eq.).

Disks were soaked in the aqueous solution for 30 minutes. Functionalized disks were treated with UV light (λ = 365 nm, I = 1.2 mW/cm2) for 30 minutes. Physically adsorbed disks had the aqueous solution removed and replaced with an Argon purged aqueous solution containing zwitterion-SH (2 eq.) and LAP (1.0 eq.). Physically adsorbed disks were kept in the dark and not exposed to UV light. Finally, control disks were kept in deionized

117 water for 30 minutes to serve as a baseline control for the alloc-PEUs. After the respective treatments, samples were rinsed with deionized water three times and then submerged in 2 mL of deionized water for 30 minutes. Disks were removed, dried with nitrogen, and examined by x-ray photoelectron spectroscopy (XPS). The XPS spectra were obtained using a VersaProbe II Scanning XPS Microprobe from Physical Electronics (PHI), under ultrahigh vacuum conditions with a pressure of 2.0 µPa. Automated dual beam charge neutralization was used during the analysis of the samples to provide accurate data. The analyzer pass energy was 117.4 eV for the survey spectra and 23.5 eV for the high-resolution scans in the N1s and S2p regions with binding energy referenced to the high resolution N1s peak (~398.9 eV). The survey scans in the range 0 – 700 eV were used to evaluate the percentage of different atoms present on the surface of the samples.

Atomic concentrations were calculated with PHI MultiPak software. The XPS high resolution spectra of N1s were decomposed into two components using the curve fitting routine in MultiPak along with assessment of the full width at half max (FWHM) to further confirm elemental distribution. Each spectrum was collected using a monochromatic (Al

Kα) x-ray beam (E = 1486.6 eV) over a 100 μm × 1400 μm probing area with a beam power of 100 W.

5.3.10. Protein Adsorption

Protein Protein adsorption profiles on functionalized, physically adsorbed, and control alloc-PEUs were obtained using a quartz crystal microbalance with dissipation

(QCM-D) (Qsense E4, Biolin Scientific AB). Terpolymer PEUs were spin coated onto QCM

118 chips (Qsense, 335 SiO2) using a 1 % (wt/wt) solution of polymer in acetone at 8000 rpm for 1 min. Chips were further annealed for 24 hours prior to being subjected to measurements. A solution of human derived fibrinogen (Fn) in 1 × PBS was prepared (1.5 mg/mL). Samples were loaded in to their respective QCM chambers and a separate solution of 1 × PBS was run (0.150 mL/min at 37 °C) to establish a baseline. Once a baseline was achieved (~ 30 minutes), the Fn solution was added and exposed at the same rate (0.150 mL/min). The solution flow was stopped when a frequency drop was observed and Fn coverage equilibrium was allowed to occur (~ 30 minutes). Once a baseline equilibrium was established, the solution of 1× PBS was again flushed over the QCM chips and any frequency change was observed and noted. Samples were flushed until a similar equilibrium baseline was achieved. Fibrinogen adsorption to the surface was calculated from the 5th overtone using the Sauerbrey equation.48 The equilibrium mass of adsorption after Fn addition and mass removed after 1 × PBS wash were thus reported.

Statistical analyses were done using a one-way ANOVA with Tukey post hoc analysis. A value of p < 0.05 was considered significant.

5.3.11. Cytotoxicity Assessment

To assess the cell compatibility of 5% and 10% alloc-PEU analogues, a cytotoxicity assay was performed. Solvent cast films (5 × 5 cm) were submerged in 20 mL centrifuge tubes and covered in extraction media (DMEM with penicillin-streptomycin, no Bovine

Calf Serum (BCS)) (8.3 mL). Three controls were used as comparisons for cytotoxicity.

Extraction media negative and positive control samples had 10 mL of extraction media

119 added to a 15 mL centrifuge tube. Latex finger cots were used as an additional positive control and placed in a 15 mL centrifuge tube with 6 mL of extraction media. All were shaken gently at 37 °C for 24 h. NIH 3T3 cells were plated and allowed to reach confluency. Cells were subsequently harvested with trypsin and combined with complete

DMEM media and BCS. An aliquot of the cell solution (1 mL) and complete media (1 mL)

(2 mL total) were added to each plate. The plates were placed in a cell culture incubator

(37°C, 5% CO2) for 24 h. Following incubation, the extracts were removed from the shaker and added to a syringe. The volume in the syringe was noted and BCS was added to the total volume extract (except for no serum negative control) (BCS accounts to 10% of extraction volume). Extraction solutions were filtered with a nylon, 0.2 μm filter. Each well of cells had the media removed and 2 mL of the extract or control was pipetted into appropriate wells. The controls were cells in DMEM media plus FBS (filtered the same way as the sample extractions), cells in media without serum (NS), and cells in the filtered latex extract plus FBS. The plates were returned to a cell culture incubator (37°C, 5% CO2) for 48 hours. After 48 hours, one picture was taken of each well with the 10 × objective on an inverted microscope. The wells were analyzed and subsequently ranked based on cell viability.

5.3.12. PEU Terpolymer Implantation

PEU free-standing films were prepared by blade coating as described above. In short, 10% PEU analogues were dissolved in acetone at 35% weight. Polymer solutions were then blade coated (8 cm blade width, 1000 µm gap height) on PET and allowed to

120 air dry for 24 h (~150-200 µm thick). The PEU films were further dried under reduced pressure to remove residual solvent. Sections of 10% alloc-PEUs were cut (2 cm × 6 cm) and divided into two groups (functionalized with zwitterion thiol or blank unfunctionalized control, n = 8 per group). All films were sterilized using ethylene oxide for further in vivo characterization. A hernia adhesion rat model was adapted from previously reported studies.49,50 All procedures and animal handling were in accordance with The University of Minnesota Institutional Animal Care and Use Committee (IACUC

Protocol Number 1805-35945A). Female Sprague-Dawley rats (32) were used to test adhesion between four groups (sham, adhesion control, 10% alloc-PEU functionalized, and 10% alloc-PEU-blank) (n = 8 per group). Buprenorphine SR was subcutaneously injected two to four hours prior to surgery and used for analgesic effect. Each animal was anesthetized with 1-4% isoflurane administered through a nose cone. Animals underwent a paramedian laparotomy followed by exposure and abrasion of the cecum

(abrasion not performed for sham). Each implant was applied, secured in place around the cecum with sutures, followed by closure of the laparotomy. Samples were left implanted for 21 days. Animals were humanely euthanized using carbon dioxide and polymer implants were examined for the extent and tenacity of adhesions present by a board-certified veterinary pathologist. Samples were scored accordingly and reported.

Statistical analyses were done using a one-way ANOVA with Tukey post hoc analysis or a

Kruskal-Wallis ANOVA with a value of p < 0.05 being considered significant.

5.3.13. Adhesion Histological Assessment

121

Tissue samples from adhesions were placed in 10% neutral buffered formalin for fixation prior to sectioning and staining with Masson’s trichrome and with hematoxylin and eosin for microscopic evaluation. The fixed samples were to be evaluated for degree of collagen deposition, necrosis, and inflammation. After staining, samples were collected and characterization by a board-certified veterinary pathologist.

5.3.14. Film Surface Topology

Surface topology images of blade coated films were obtained using scanning electron microscopy (SEM). Using a JEOL USA SEM, samples were scanned at 2.0 kV excitation at 50 × magnification. 10% alloc-PEU functionalized and 10% alloc-PEU blank films were imaged after ethylene oxide sterilization and after implantation for 21 days to observe surface morphology changes.

5.4. Results

5.4.1. Synthesis

Amino acid-based monomers were synthesized and characterized using 1H-NMR

(Appendix A Figure 6.8-9). 1,6-hexanediol and 1,8-octanediol were coupled to the carboxylic acid of L-valine or L-phenylalanine through an esterification using p- toluenesulfonic acid to prevent reactions at the amine moiety. The resulting di-p- toluenesulfonic acid salts of bis(l-valine)-octane 1,8-diester and di-p-toluenesulfonic acid salts of bis(l-phenylalanine)-hexane 1,6-diester monomers were named based on their diol chain length and amino acid; (1-VAL-8) formed from 1,8-octanediol and l-valine, (1-

PHE-6) formed from 1,6-hexanediol and l-phenylalanine. The synthesis of 1-VAL-8 was

122 confirmed based on the characteristic l-valine methyl resonance at 0.96 ppm. The synthesis of 1-TYR-2 Alloc was carried out through a DIC coupling with Boc-O-benzyl-l- tyrosine and 1,3-allyoxyl-2-propanediol. The final product was collected and confirmed following the acidic (HCl) deprotection with the disappearance of the characteristic Boc protecting group resonances at 1.31 ppm and the appearance of the broad amine peak at 8.82 ppm (Appendix A Figure 6.18). Synthesis of the zwitterion thiol occurred in several steps including a halogenation reaction was performed between 3,3’-dithiodipropionic acid and thionyl chloride to afford 3,3’- dithiodipropionyl chloride (Appendix A Figure

6.19). This product was further reacted using SN2 substitution with 3-dimethylamino-1- propanol to obtain the bis(3-dimethylamino)propyl) 3,3’-disulfanediyldipropionate

(Appendix A Figure 6.20). This was further functionalized with 1,3-propane sultone to afford the zwitterion disulfide (Appendix A Figure 6.21-22). Finally, the zwitterion-SH was obtained by cleavage of the zwitterion disulfide using 1,4-dithiothreitol (DTT) and the product was thus collected (Appendix A Figure 6.23-24). Terpolymer PEU synthesis was carried out by altering the molar feed ratio of 1-VAL-8, 1-PHE-6, and 1-TYR-2 alloc in an interfacial polymerization with triphosgene to afford the 5% alloc-PEU and 10% alloc-PEU

(Scheme 5.1) (Figure 5.1). Polymers were named according to the amount of 1-TYR-2 alloc in the monomer feed ratio. Successful polymer synthesis and monomer incorporation was confirmed using 1H NMR. The monomers were readily identified by the characteristic methyl L-valine peaks denoted ‘a’, methylene from L-phenylalanine denoted ‘l’, and methylene of O-bzn-TYR denoted ‘v’. Additionally, the alkene

123 functionality from 1-TYR-2 alloc could be observed with peaks at 5.13 ppm (d’) and 5.84 ppm (c’).

Figure 5.1. Polymer 1H-NMR overlay of previously synthesized 0% Alloc-PEU79 and 5% alloc-PEU and 10% alloc-PEU. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine, the methylene resonances from L-phenylalanine denoted ‘l’ in blue, and the methylene resonances from the benzyl protected L-tyrosine denoted ‘v’ in green.

124

Scheme 6.1. General synthetic scheme for PEU terpolymer analogues where the monomer synthesis of 1-VAL-8, 1-PHE-6, and 1-TYR-2 alloc were carried out via a Fischer esterification or a DIC coupling between varying diol chain lengths and amino acids. In total two terpolymers were synthesized by combining 1-VAL-8 and 1-PHE-6 and 1-TYR-2 alloc in two stoichiometric amounts to form 5% alloc-PEU and 10% alloc-PEU. Both 125 polymers were formed through interfacial polymerization with triphosgene to afford urea units. Polymers were further functionalized with a thiol-ene reaction between a zwitterion-SH and the alkene functionality.

5.4.2. Physical Properties

Size-exclusion chromatography (SEC) reported molecular masses are post- precipitation in acetone (Table 5.1) (Appendix A Figure 6.25). The molar mass distributions (Đm) are less than the theoretical value of 2 because the low molecular mass fractions are lost during precipitation. Molecular masses were sufficiently high to allow for mechanical testing and subsequent in vivo characterization. Differential scanning calorimetry (DSC) curves show the glass transition temperatures (Tg) are above physiological temperature indicating that the materials will be sufficiently stiff upon any implantation application (Appendix A Figure 6.26). An increase in the Tg is observed as the amount of 1-TYR-2 alloc monomer is elevated from 5% to 10%. This is expected as the rigid side chain unit from the benzyl protected tyrosine inhibits chain mobility. Ideally, all terpolymer PEUs had sufficiently high degradation temperatures (Td) enabling them to be thermally processed with sufficient stability for processing and storage in future applications (Table 5.1) (Appendix A Figure 6.27).

Table 5.1. Physical properties of alloc-PEUs

Polymer Mn (KDa) Mw (KDa) Ðm Tg (°C) Td (°C)

5% Alloc PEU 24 35 1.5 40 289

10% Alloc PEU 28 62 2.2 49 299

126

5.4.3. Mechanical Properties

Tensile-Test Mechanics on Free-Standing Films. PEU free-standing films were prepared by blade coating, sterilized with ethylene oxide, and cut in to tensile bars for mechanical assessment. Tensile bars were pulled with a constant rate of extension (25.4 mm/min) and the stress versus strain curves were recorded (Figure 5.2 A). A significant elevation in the modulus and suppression of the yield strain was observed for 10% alloc-

PEU samples when compared to 5% alloc-PEU samples (Figure 5.2 B-C). These changes in mechanical properties were expected and are attributed to the bulky side chain in the 1-

TYR-2 alloc monomer restricting interchain mobility. It is common for a stiffer material to concomitantly possess a lower yield strain. No statistical difference was observed for the yield stress between samples (Figure 5.2 D).

127

Figure 5.2. Tensile testing mechanical properties of PEU terpolymers. The stress versus strain curves (A) indicate that there is greater stiffness (B) for the 10% alloc PEU when compared to the 5% alloc analogue. There is an increase in yield strain for the 5% alloc- PEUs when compared to the 10% alloc-PEUs (C). No statistical difference for yield stress was observed for between the PEU derivatives (D). An * indicates a p value < 0.05 (n = 3 samples).

5.4.4. Surface Functionalization

1000 MW FITC-PEG-SH. Alloc-PEUs samples were functionalized with a FITC-PEG-

SH dye to quantify how much functionality could be exploited at the surface. A calibration curve was created with values ranging from nanomolar to micromolar concentrations

(Appendix A Figure 6.28). Blade coated 5% alloc-PEUs and 10% alloc-PEUs were punched 128 in to 8 mm disks and then functionalized through the alkene moiety with previously synthesized photoinitiator LAP and FITC-PEG-SH (Appendix A Figure 6.29). Samples were divided into three groups labeled functionalized, physically adsorbed, and blank where the functionalized samples were treated with LAP, FITC-PEG-SH, and UV light, physically adsorbed only were treated with LAP and FITC-PEG-SH, and blank samples were left untreated. All samples were dissolved in DMSO and fluorescence measurements were taken to observe the emission at 550 nm (Figure 5.3) (Table 5.2). The blank polymer was subtracted out from both functionalized and physically adsorbed samples. The emission intensity at 550 nm indicates that the amount of FITC-PEG-SH present on the functionalized samples is greater than that of the physically adsorbed samples for both the 5% alloc-PEU and 10% alloc-PEU analogues. Additionally, the 10% alloc-PEU shows greater dye attachment for the functionalized material when compared to the 5% alloc-

PEU functionalized analogue. This was expected as more 1-TYR-2 alloc monomer incorporated into the polymer should correspond with an increase in the amount of available alkene functionality on the surface. The increase in observable dye on the functionalized alloc-PEUs also supports that there is covalent attachment occurring under the thiol-ene reaction conditions. This is desirable as controlled attachment will be important when modulating surface properties for an antiadhesion implantable device

(Figure 5.3).

129

Figure 5.3. FITC-PEG-SH fluorescence dye attachment for surface functionalization and physically adsorbed 5% and 10% alloc-PEU analogues at 550 nm endpoint emission. An increase in dye concentration is observed for the 10% alloc-PEU functionalized polymer when compared to 10% alloc-PEU physically adsorbed. The same trend is observed for the 5% alloc-PEU analogues.

Table 5.2. FITC-PEG-SH Dye Attachment

Dye Attachment Polymer 2 (nanomole/cm )

5% Alloc-PEU Phys. Ads. 8.8 ± 2.6 5% Alloc-PEU Func. 18.1 ± 1.3

10% Alloc-PEU Phys. Ads 13.3 ± 8.2

10% Alloc-PEU Func. 134.2 ± 78.9

X-ray Photoelectron Spectroscopy (XPS). After the model reaction with FITC-PEG-

SH, quantification of the zwitterion-SH on the surface of the 5% alloc-PEU and 10% alloc-

PEU analogues was performed using XPS. Samples were divided in to three groups labeled functionalized, physically adsorbed, and blank. The functionalized samples were

130 treated with LAP, zwitterion-SH, and UV light, physically adsorbed samples were treated with LAP and zwitterion-SH without UV light exposure, and blank samples were kept separate from LAP and the zwitterion-SH. Following XPS analysis, it was shown that the

5% alloc-PEU functionalized material has an observable broadening and two distinct nitrogen peaks when compared to one major peak observed for the physically adsorbed and blank material (Figure 5.4 A). The major peak (~399.9 eV) corresponds to the nitrogen in the urea groups while the other peak (~401.5 eV) correlates to the quaternary ammonium in the zwitterion. The integration ratios for the urea and quaternary ammonium peaks are 95.2:4.8 respectively (Figure 5.4 B). The quaternary ammonium peak is shifted to higher energy as it is electron deficient, more stable and therefore requires more energy to displace an electron from its N1s orbital. Additionally, two distinct sulfur peaks are observed for the 5% alloc-PEU functionalized samples where the zwitterion-S-C sulfur is seen at lower energy than the ring opened sultone sulfur (48:52 respectively) (Figure 5.4 C). This again was expected as the stability of the ring opened sultone outer shell of electrons is greater, requiring greater energy to displace the S 2p electrons. Similar trends were observed for the 10% PEU-alloc samples. A noticeable broadening is again observed for the raw XPS curves of the 10% alloc-PEU functionalized material when compared to the physical adsorbed and blank materials (Figure 5.4 D).

Only the 10% alloc-PEU functionalized material has the secondary peak correlating to the quaternary ammonium peak attributed to the zwitterion-SH (integration urea:quaternary ammonium is 65.5:35.5) (Figure 5.4 E). The amount of detectable quaternary ammonium

131 was greater for the 10% alloc-PEU functionalized material than it was for the 5% alloc-

PEU functionalized material. This was expected as more functional monomer incorporation into the polymer allows for greater surface attachment of zwitterion-SH.

Finally, two distinct sulfur peaks are observed for the 10% alloc-PEU functionalized material (integration ratio zwitterion-S-C sulfur:ring opened sultone sulfur is 55:45)

(Figure 5.4 F). These results correlate well with what was observed in the FITC-PEG-SH model reaction where the most detectable zwitterion-SH is from the samples functionalized with UV light. These results provide further evidence that the thiol-ene attachment of the zwitterion-SH to alloc-PEUs is feasible with reasonable control based on molar feed ratio of 1-TYR-2 alloc monomer.

132

Figure 5.4. High resolution XPS of the N1s orbital for the 5% alloc-PEU functionalized, physically adsorbed, and blank materials were plotted (A). A slight broadening is observed with curve fitting indicating that there are two distinct peaks for the urea nitrogen (~399.9 eV) and quaternary ammonium nitrogen (~401.5 eV) for the 5% alloc- PEU functionalized material (integration 95.2:4.8) (B). High resolution XPS of the S2p orbital was also taken and two distinct sulfur peaks correlating to the zwitterion-S-C sulfur (~162 eV) and ring opened sultone sulfur (~168 eV) are observed (integration 48:52) (C) 133 were observed. High resolution XPS of the N1s orbital for the 10% alloc-PEU functionalized, physically adsorbed, and blank materials were plotted (D). A broadening is observed with curve fitting for the urea nitrogen (~399.9 eV) and quaternary ammonium nitrogen (~401.5 eV) showing integration values of 65.5:35.5 (E). High resolution XPS of the S2p orbital of the 10% alloc-PEU functionalized material showed two distinct sulfur peaks correlating to the zwitterion-S-C sulfur and ring opened sultone sulfur are again observed (integration 55:45) (F).

Water Contact Angle. To observe if there was a change in the macroscopic properties for the alloc-PEUs post functionalization with the zwitterion-thiol, water contact angle measurements were taken before and after functionalization (Figure 5.5 A-

D). Samples were allowed to equilibrate for ten minutes prior to imaging and measurement. After ten minutes, the contact angle of the 5% alloc-PEU blank material

(52.3 ± 0.5) is greater than the 5% alloc-PEU functionalized material (46.6 ± 3.6) (Figure

5.5 E). The same contact angle suppression was observed for the 10% alloc-PEU analogues were the blank material (61.8 ± 3.1) contact angle was greater than that of the functionalized material (56.3 ± 7.6) (Figure 5.5 E). The subtle decrease in water contact angle is indicative of the zwitterion-SH forming a hydration layer with the water and thus improving the hydrophilicity of the material. This is ideal as the formation of a tight hydration layer has been shown to prevent the adhesion of proteins, cells, and other molecules.70,165,166

134

Figure 5.5. Contact angle of functionalized and blank alloc-PEU analogues. Alloc-PEUs were coated with deionized water (5 µL), allowed to equilibrate for 10 minutes, and finally imaged with a goniometer (A-D). Angles were measured using ImageJ software to afford the contact angle for each material (E). The contact angle of the 5% alloc-PEU blank material and the 5% alloc-PEU functionalized material is 52.3 ± 0.5 and 46.6 ± 3.6 respectively (n = 4-6). The contact angle of the 10% alloc-PEU blank and 10% alloc-PEU functionalized materials were 61.8 ± 3.1 and 56.3 ± 7.6 respectively (n = 4-6).

5.4.5. Protein Adsorption

Quartz Crystal Microbalance. Fibrinogen adsorption profiles for 5% alloc-PEU blank, 5% alloc-PEU functionalized, 10% alloc-PEU blank, and 10% alloc-PEU functionalized materials were obtained using QCM (Figure 5.6 A-B). A drop-in frequency can be observed on all curves which is indicative of the adsorption of fibrinogen to the material surface. After equilibration and the 1 × PBS rinse, an amount of reversibly adsorbed material is washed from the surface indicated by an increase in the frequency. 135

The amount of fibrinogen initially attached is an important feature for an antiadhesion material as fibrinogen is likely to be one of the first substances to interact with the polymer implant.68,167 The calculated amount of attached fibrinogen for 5% alloc-PEU blank sample was statistically greater than all of the other materials under study (Figure

5.6 C). This was ideal as functionalization of the 5% alloc-PEU is able to inhibit adhesion to the surface when compared to the blank counterpart (0.024 ± 0.002 ng/cm2 and 0.123

± 0.015 ng/cm2 respectively). While there were no significant differences between the

10% alloc-PEU blank (0.031 ± 0.017 ng/cm2) and 10% alloc-PEU functionalized (0.017 ±

0.001 ng/cm2) materials, an average adhesion suppression is still observed for the functionalized material. Additionally, the material with the lowest average adhesion is the 10% alloc-PEU functionalized material. These data lend further evidence to successful surface functionalization and the promise for these materials as antiadhesion barriers. In addition to initial fibrinogen attached to the material surface, irreversibly adsorbed fibrinogen was calculated after the 1 × PBS rinse (Figure 5.6 D). As was observed with the initial fibrinogen attachment, the 5% alloc-PEU blank (0.105 ± 0.046 ng/cm2) material has statistically more irreversibly adsorbed fibrinogen when compared to the other alloc-PEU materials under these conditions. No statistical difference is observed between the 5% alloc-PEU functionalized (0.175 ± 0.009 ng/cm2), 10% alloc-PEU blank (0.007 ± 0.009 ng/cm2), and 10% alloc-PEU functionalized (0.011 ± 0.001 ng/cm2) materials (Figure 5.6

D). It is expected that the tight hydration layer formed from the functionalization of the zwitterion-SH to the alloc-PEU surfaces helps to prevent adhesion of competing

136 molecules like fibrinogen. The differences observed in fibrinogen adhesion based on polymer composition and surface functionalization are promising results for modulating the adhesion properties of an implantable material.

Figure 5.6. QCM for 5% and 10% alloc-PEU analogues with fibrinogen. QCM representative curves for fibrinogen (1.5 mg/mL) on alloc-PEU analogues from the 5th overtone are reported (A, B). The amount of initially attached fibrinogen to the surface is reported (C) for 5% alloc-PEU blank (0.123 ± 0.015 ng/cm2), 5% alloc-PEU functionalized (0.024 ± 0.002 ng/cm2), 10% alloc-PEU blank (0.031 ± 0.017 ng/cm2), and 10% alloc-PEU functionalized (0.017 ± 0.001 ng/cm2) materials (n = 3-4) (an * indicates a statistical difference between 5% alloc-PEU blank materials and any material that shares this symbol, p < 0.05). The amount of irreversibly attached fibrinogen to the surface is reported (D) for 5% alloc-PEU blank (0.105 ± 0.046 ng/cm2), 5% alloc-PEU functionalized

137

(0.175 ± 0.009 ng/cm2), 10% alloc-PEU blank (0.007 ± 0.009 ng/cm2), and 10% alloc-PEU functionalized (0.011 ± 0.001 ng/cm2) materials (n = 3-4) (an * indicates a statistical difference between 5% alloc-PEU blank materials and any material that shares this symbol, p < 0.05).

5.4.6. Cytotoxicity Assessment

Alloc-PEU Films. Any material will be exposed to a wide array of bodily fluids that will interact with the surface upon implantation. Even the most inert implantable devices are not precluded from some level of swelling which can lead to small molecule impurities and physically adsorbed or trapped compounds migrating in to the dynamic physiological system. With a solvent cast surface functionalized material like the alloc-PEU analogues, it is important to ensure that the processing methods and functionalizing conditions do not lead to unwanted cytotoxicity side effects upon prolonged exposure. A solution- based cytotoxicity assay was performed to assess cell viability for the alloc-PEU materials.

Briefly, alloc-PEU films were submerged in extraction media for 24 hours. This extraction media was then used to culture NIH-3T3 cells for 48 hours and cell culture plates were used to assess cell viability and were subsequently ranked (Figure 5.7 A-G). Scoring was based on a scale from 0-4 where the extreme scores of 0 indicates very little cell distress and no observable empty spaces in the culture plate in contrast to a score of 4 which indicates >75% of observable cell death or distress with large empty areas on the culture plate (Figure 5.7). Cells exposed to the alloc-PEU sample extracts were similar to the negative control cells and had very few empty spaces. All sample extracts and the negative controls passed (Score = 0) (Figure 5.7 A-D, F). Both of the positive controls did

138 not pass as expected (Score = 4) (Figure 5.7 E, G). Additionally, the alloc-PEU samples did not change the pH of the extraction media and the extracts filtered well. During the 24- hour incubation in extraction media, the polymers went from clear to cloudy which is attributed to a small amount of water uptake leading to a change in opacity. All alloc-PEU analogues eliciting a cytotoxic response equivalent to the negative control was ideal as it shows the compatibility of these PEU materials under the processing conditions. These results lend further evidence that alloc-PEUs would not be considered cytotoxic for an implantable adhesion barrier in a dynamic fluid environment.

Figure 5.7. In vitro cytotoxicity assessment of alloc-PEUs. 5% alloc-PEU blank (A), 5% alloc-PEU functionalized (B), 10% alloc-PEU blank (C), and 10% alloc-PEU functionalized 139

(D) materials’ extracted media was cultured with NIH-3T3 cells for 48 hours. Cell culture plates were then imaged and cell death and distress were scored from 0-4 (0 = limited cell death and equivalent to negative control, 1 = < 25% cell death and distress, 2 = < 50% cell death and distress, 3 = < 75% cell death and distress, and 4 = > 75% cell death and distress). Three controls were used where cells cultured without serum acted as a positive control (E), cells cultured with serum acted as a negative control (F), and cells cultured with media from Latex cots acted as a positive material control (G). All groups were scored with all alloc-PEU materials scoring equivalent to the negative control (score = 0) (n = 3 for alloc-PEUs, n = 2 for negative control). Both positive controls elicited a cytotoxic score (score = 4) (n = 2 for positive controls).

5.4.7. In Vivo Characterization

Adhesion Animal Model. Despite in vitro analysis, no benchtop assessment or computer model to date can successfully encompass adhesiogenesis and the complex interplay factors involved. For this reason, an intrabdominal rat adhesion model was used to assess adhesiogenesis of the alloc-PEUs understudy. 10% alloc-PEU blank and 10% alloc-PEU functionalized derivatives were selected for implantation because of successful adhesion prevention observed in QCM.

Scheme 5.2. Surgical procedures for the antiadhesion rat hernia model where a paramedium laparotomy was performed (A) to expose the cecum. Surgical sham control group animals were closed without further procedures. All other groups had the cecum abraded to ensure adhesion development (B). The adhesion control group was replaced without further modification. 10% alloc-PEU blank and functionalized films (2 × 8 cm)

140 were sutured around the abraded cecum acting as a physical barrier between the underlying tissue (C) and subsequently replaced in to the abdominal cavity (n = 8 for all surgical groups).

All procedures and animal handling were in accordance with The University of

Minnesota Institutional Animal Care and Use Committee (IACUC Protocol Number 1805-

35945A) where female Sprague-Dawley rats (32) were divided in to four groups (sham, adhesion control, 10% alloc-PEU blank, and 10% alloc-PEU functionalized) (n = 8 per group). All animals underwent a paramedian laparotomy followed by exposure and abrasion of the cecum (Scheme 5.2 A). Cecum abrasion was not performed for the sham group and consequently no adhesion was expected. Following abrasion, the 10% alloc-

PEUs were applied, secured in place around the cecum with sutures, and the incision was subsequently closed (Scheme 5.2 B-C). Following implantation (21 d) all groups were examined for extent and tenacity of adhesions present. Extent of adhesion was scored based on the relative percentage of adhesion to the abraded cecum (Table 5.3) and the tenacity of adhesions present was scored based on the resistance for separation (Table

5.4). The goal of an antiadhesion material in a ventral hernia application is to keep inner organs from interacting with the abdominal wall. Additionally, it would be preferable for limited adhesion between tissue and the device itself. The overall extent of adhesion

(includes adhesions between tissue and abraded cecum and adhesion of abraded cecum to the device) was scored with the 10% alloc-PEU functionalized material (2.0 ± 1.1) obtaining the lowest average although no significant difference was noted between the abraded cecum adhesion control (2.3 ± 1.4) and the 10% alloc-PEU blank materials (2.4 ± 141

1.0) (Appendix A Figure 6.30 A-B) (Figure 5.8 A). No extent of adhesions was observed for the sham animals which was anticipated (Figure 5.8 A). Tenacity of adhesions to the device and underlying tissue were scored with the 10% alloc-PEU functionalized material

(2.3 ± 0.7) again eliciting the lowest score (Figure 5.8 B). The observed drop in adhesion tenacity was not statistically significant from the abraded cecum control and 10% alloc-

PEU blank material. While the reduction in extent and tenacity of adhesions present is ideal for the functionalized material, a larger study may show statistical differences between the materials under study. The extent of adhesions exclusively attached to the implanted device and abraded cecum were also scored and reported (Figure 5.8 C).

Promisingly, there were minimally observed adhesions between the 10% alloc-PEU functionalized material (0.3 ± 0.5) and 10% alloc-PEU blank material (0.9 ± 0.7) to the abraded cecum (Figure 5.8 C). The tenacity between the materials and tissue were also recorded with no tenacity of adhesion being observed for the 10% alloc-PEU functionalized material (0.0 ± 0.0) and only minimal tenacity shown for the 10% alloc-PEU blank material (1.1 ± 0.9) (Figure 5.8 D). Similar to when device, tissue, and abraded cecum were used to assess extent and tenacity of adhesions, there was no significant difference observed between groups for extent and tenacity to device exclusively.

However, the functionalized material has a lower average extent and tenacity in all cases which could be an indication that the surface treatment is having a desired effect. An increase in sample size could help differentiate between the groups under study to find statistical significance.

142

Table 5.3. Extent of adhesions to abraded cecum scoring scale

Table 5.4. Tenacity of adhesions scoring scale

143

Figure 5.8. Adhesion extent and tenacity of explanted materials. The overall extent of adhesion (includes adhesions between tissue and abraded cecum and adhesion of abraded cecum to the device) with the sham, adhesion positive control, 10% alloc-PEU blank, and 10% alloc-PEU functionalized material having scores of 0 ± 0, 2.3 ± 1.4, 2.4 ± 1.0, and 2.0 ± 1.1 respectively (A). Tenacity of adhesions to the device and underlying tissue were scored with the sham, adhesion positive control, 10% alloc-PEU blank, and 10% alloc-PEU functionalized material having scores of 0.0 ± 0.0, 2.4 ± 0.5, 2.6 ± 0.5, and 2.3 ± 0.7 respectively (B). The extent of adhesions exclusively attached to the implanted device and abraded cecum were scored with minimally observed adhesions for the 10% alloc-PEU functionalized material (0.3 ± 0.5) and the 10% alloc-PEU blank material (0.9 ± 0.7) (C). No tenacity of adhesion was observed for the 10% alloc-PEU functionalized material with slight tenacity being shown for the 10% alloc-PEU blank material (1.1 ± 0.9) (D). No statistical differences were observed between groups on a p < 0.05 confidence interval (n = 7-8).

144

Adhesion Histological Assessment. Tissue samples were collected post-mortem and subjected to histological staining to identify cell type and inflammation present at the location of observed adhesions. The degree of collagen deposition, necrosis, and inflammation were all qualitatively compared between treatment groups. The sham control did not demonstrate histological evidence of adhesion formation. Primarily, the cecum was normal with no notable hyperplastic lymphoid tissue and limited fibroblasts

(Figure 5.9A-B). This was expected as no adhesion method was employed to these samples. All other groups demonstrated adhesion tissue characterized by infiltration of fibroblasts, macrophages, and occasional giant cells. The adhesion in these areas was moderate to heavy and cecal hyperplastic lymphoid tissue was observed. This is indicative of an inflammatory response attributed the tissue adhesions (Figure 5.9C-D).

Additionally, collagen deposition was observed with blue-stained collagen bundles surrounded by multiple fibroblasts and identifiable giant cells (Figure 5.9E-F). These cell types are indicative of an inflammatory response that could be expected from the observed adhesions. No implant material was observed, likely due to dissolution during alcohol processing as part of the histological preparation.

145

Figure 5.9. Tissue samples were stained with hematoxylin and eosin (H&E) or Masson’s trichrome (tri) to identify collagen deposition and cell types present surrounding the adhesion areas. Representative images are shown for the groups under study. Sham control groups displayed a normal cecum without abnormal cellular densities (A-B). All other groups displayed hyperplastic lymphoid tissue (C-D) and blue-stained bundles of collagen fibers containing numerous fibroblasts and several giant cells (E-F).

Polymer Degradation. The implanted polymer samples were further characterized to assess the extent of degradation in vivo. Scanning electron microscopy (SEM) was used to observe the surface morphology of films after sterilization and after implantation

(Figure 5.10). Both 10% alloc-PEU blank and 10% alloc-PEU functionalized films are predominantly smooth after sterilization (Figure 5.10 A). This was ideal as it indicates that the sterilization technique does not damage the implant surface topology which would alter adhesion properties. After implantation, surface roughness is noticeable for both the blank and functionalized derivatives (Figure 5.10 B-C). While there is some change, the degradation does not appear extreme, as would be expected if the material had significantly eroded.

146

Figure 5.10. Scanning electron microscopy (SEM) images of solvent cast films post-EtO sterilization (A), and of 10% alloc-PEU blank (B) and 10% alloc-PEU functionalized (C) films after 21 days of implantation.

In addition to surface morphology, polymer molecular mass was determined using

SEC-GPC to assess the extent of polymer chain scission in vivo. Differential distribution curves for the SEC-GPC traces were collected and molecular mass values were reported

(Figure 5.11) (Table 5.5). Minimal changes in molecular mass are observed with the implanted materials when compared to their initial starting values. The sustained molecular mass through implantation is ideal for this study as the extent and tenacity of adhesion can be more easily quantified. However, the observed molecular mass fidelity is an indication that a thinner and lower molecular mass film could prove useful as an adhesion barrier for hernia applications.

147

Figure 5.11. Molecular mass degradation of 10% alloc-PEU derivatives using size- exclusion chromatography-gel permeation chromatography (SEC-GPC) in THF.

Table 5.5. Molecular mass degradation values of 10% alloc-PEU derivatives

Polymer Mn (KDa) Mw (KDa) ĐM 10% Alloc-PEU Initial 14.0 29.5 2.10

10% Alloc-PEU Blank 21d 15.0 29.9 1.99

29.6 2.17 10% Alloc-PEU Func. 21d 13.7 Size-exclusion chromatography-gel permeation chromatography (SEC-GPC) molecular mass and dispersity values are calculated from a polystyrene standard in THF.

5.5. Conclusion

To date, limited materials are available to inhibit adhesion in traditional hernia repair surgeries. Regardless of mesh selected for a repair, an inherent risk for adhesion formation exists for ventral hernia repair. Surface modification of hernia repair materials could be utilized to prevent adhesion in vitro and in vivo. Two anti-adhesion poly(ester urea)s (PEUs) with varied amount of an allyl ether side chain with terminal alkene functionality for surface modification (5% alloc-PEU or 10% alloc-PEU) with a zwitterion

148 thiol were studied. Controls and two formulations with varying functional group concentrations available on the surface were confirmed through attachment of a FITC-

PEG-SH dye with both functionalized materials displaying increased dye attachment. As the molar concentration of alkene functional side chain increased from 5-10% for the alloc-PEUs, an increase in dye attachment was also observed. A zwitterion-thiol was attached through a click thiol-ene reaction to the alloc-PEU surface. Attachment was confirmed through water contact angle, XPS, and QCM. Anti-adhesion PEUs reduce fibrinogen attachment when compared to unfunctionalized controls demonstrating reasonable control over surface modification. Furthermore, these materials display limited cytotoxicity in vitro towards NIH-3T3 cells with all alloc-PEU analogues behaving similar to negative controls. 10% alloc-PEU materials were implanted in a rat intrabdominal adhesion model with the 10% alloc-PEU zwitterion functionalized material being found to reduce the average extent and tenacity of adhesions when compared to adhesion controls and the 10% alloc-PEU blank material. Although in vivo results were not found to be statistically significant, a difference could be observed with future studies with a larger sample size. As a preliminary study, functionalized anti-adhesion PEUs demonstrate their promise as materials used to prevent adhesion in ventral hernia repair.

5.6. Acknowledgement

The authors gratefully acknowledge financial support from Cook Biotech and the

W. Gerald Austen Endowed Chair in Polymer Science and Polymer Engineering from the

149

John S. and James L. Knight Foundation. The authors would like to also thank Dr. Suckow for his experience and expertise in adhesiogenesis surgical techniques.

150

CHAPTER VI

CONCLUSION

6.1. Conclusion and Outlook

6.1.1. Preclinical In Vitro and In Vivo Assessment of Linear and Branched Poly(ester urea)s for Soft tissue Applications

In this first initial study, a synthetic alternative to the currently employed polypropylene was explored. Nonresorbable materials like polypropylene have certainly served their purpose as light weight implantable meshes. They are commonly employed because of their robust mechanical properties and limited shrinkage upon implantation.

Despite this, these materials remain permanent and are often indicated with adhesion, prolonged inflammatory response, and subsequent infection from recurrent surgeries.

Current limitations might be remedied by long term resorbable polymers. We therefore explored a series of L-valine based PEUs as alternatives for soft-tissue repair applications with their synthesis, mechanical properties, and biocompatibility under study. The PEUs tested show Young’s moduli comparable to polypropylene (105 ± 30 - 269 ± 12 MPa) with the 2% branched poly(1-VAL-8) maintaining the greatest mechanical properties measured up to 3 month following in vivo implantation. This was promising as it indicated that the

PEUs in question could be tuned to match materials currently on the market for hernia- repair. Most importantly, all PEUs generated a limited inflammatory response through

151 three months relative to polypropylene. As the limiting factor for many promising materials is cytotoxicity, an improvement on an already limited inflammatory response was ideal. There were several major limitations to this first series of materials moving forward that would need to be addressed. The first was that the branched PEUs had noticeable solubility issues in biologically relevant solvents. This would preclude these materials from other processing methods. Additionally, under compression molding conditions, the branched PEUs became brittle which is less than ideal for a soft-tissue application. Finally, the degradation rates could be too rapid for successful tissue integration to ensure mechanical integrity at the surgical site. Despite these limitations, limited inflammatory response through 3 months along with tunable mechanical properties make L-valine based PEUs an attractive material to be explored for soft-tissue repair. Further investigation into polymer compositions and in a clinically relevant hernia model were required.

6.1.2. Amino Acid-Based Poly(ester urea) Copolymer Films for Hernia-Repair Applications

Alternatives to nonresorbable materials for hernia-repair are currently limited to biologic materials. Despite improved cell viability and cell signaling, current biologic hernia-mesh devices leave much to be desired in sustained mechanical support. Small intestine submucosa extracellular matrix (SIS-ECM) has been shown to degrade over a several months. While this could prove beneficial for some hernia applications, the degradation rates have been shown to be too rapid for certain complex hernia applications. To improve upon the initial linear and branched PEU series, a series of L- 152 valine-co-L-phenylalanine PEU copolymers were synthesized and their mechanical properties were assessed to determine their use in composite films with ECM or as free- standing hernia films. In these PEU copolymer systems, solubility issues were avoided as these materials were readily dissolved in biologically relevant solvents (i.e. acetone, ethanol). Furthermore, these polymers were able to be processed in to flexible films without thermal processing through roll-to-roll film casting methods. While this is still an ongoing process, these materials lend themselves to have the potential to be processed in a scalable and industrially relevant method. By adding the L-phenylalanine monomer in to the PEU copolymer system, initial mechanical properties were bolstered in the hopes of extending the length of time that these materials prevent recurrence. Free-standing films of the PEU copolymers demonstrate improved mechanical properties when compared to SIS-ECM alone. Additionally, these copolymers were fabricated into composite materials with SIS-ECM. No delamination was observed indicating that PEU copolymers could be used as a way to bolster the mechanical properties of SIS-ECM without losing cell viability. As solvent cast films these materials show great in vitro cell compatibility when compared to glass slides. When considering how a material will degrade under a given application, the site of implantation is of utmost importance.

Because of this, the 30% PHE6 P(1-VAL-8) copolymer was implanted in a rat hernia model in vivo through 7 and 14 days and compared to polypropylene and SIS-ECM where it was found to have a lower inflammatory score when compared to polypropylene at both timepoints. Overall, this follow-up study indicates PEU copolymers have the potential to

153 be translatable materials as standalone films or as composite materials with SIS-ECM for hernia-repair applications. Further mechanical degradation studies are required to identify the correct material for a given hernia-repair application.

6.1.3. Poly(ester urea) Adhesion Barriers Aid in the Treatment of Hernia-Mesh Repair

In this final installment, the topic of adhesions related to hernia-repair are taken understudy. To date, limited materials are available to inhibit adhesion and there is an inherent risk for adhesion regardless of mesh selected. PEUs have been shown to be modified which lends them to surface modification. This strategy was employed to incorporate an alkene functionality on the side chain or PEUs which could be further used to modulate the surface of these materials as solvent cast films. Surface modification of hernia repair materials could be utilized to prevent adhesion in vitro and in vivo. Two anti-adhesion poly(ester urea)s (PEUs) with varied amount of an allyl ether side chain with terminal alkene functionality for surface modification (5% alloc-PEU or 10% alloc-PEU) with a zwitterion thiol were used and proof of concept attachment was performed using a FITC-PEG-SH dye. A clear difference in dye attachment was quantified and allowed for the proper surface modification to be explored. Charged surfaces have been shown to create successful modulation of a materials surface. Depending on the size and mechanism of charge distribution, zwitterions have been able to reduce adhesion by forming a tight hydration layer as a preferential conformation. A zwitterion-thiol was attached through a click thiol-ene reaction to the alloc-PEU surfaces and attachment was confirmed through a variety of characterization techniques including water contact angle, 154

XPS, and QCM. The attachment of a zwitterion-SH to the surface was shown to reduce fibrinogen attachment on alloc-PEU films which was ideal as fibrinogen is widely known to be one of the first materials to interact with an implanted surface. Furthermore, these materials displayed limited cytotoxicity in vitro towards NIH-3T3 cells even when functionalized with the zwitterion-SH indicating their promise moving forward. The best indication of adhesion performance is in an animal model. The 10% alloc-PEU materials subjected to a rat intrabdominal adhesion model were found to have a reduction in the average extent and tenacity of adhesions when compared to adhesion controls. Future studies will seek to further explore this exciting material as an adhesion barrier for any mesh selected for hernia-repair. While promising results were observed, larger animal studies could better predict the differences between tested functionalized groups.

Additionally, more study should be given to the zwitterion selected as greater differences and antiadhesion properties could be observed. As a preliminary study, functionalized anti-adhesion PEUs are a potential candidate of interest for preventing adhesion in hernia-repair.

155

REFERENCES

1. Conn, J.; Oyasu, R.; Welsh, M.; Beal, J. Vicryl (Polyglactin 910) Synthetic

Absorbable Sutures. Am. J. Surg. 128, 19–23 (1974).

2. Makadia, H. K. & Siegel, S. J. Poly Lactic-co-Glycolic Acid (PLGA) as Biodegradable

Controlled Drug Delivery Carrier. Polymers (Basel). 3, 1377–1397 (2012).

3. Middleton, J. C. & Tipton, A. J. Synthetic biodegradable polymers as orthopedic

devices. Biomaterials 21, 2335–2346 (2000).

4. Khurana, R. N. et al. Dexamethasone Implant Anterior Chamber Migration Risk

Factors , Complications , and Management Strategies. Ophthalmology 121, 67–71

(2013).

5. 510(k) Summary of Safety and Effectiveness Modified VICRYL Plus Antibacterial

Suture. Department of Health & Human Services (2003).

6. Spooner, C. FDA Approval for Silhouette Instalift. (2015).

7. Paine, J. M. 510(k) Summary of Safety and Effectiveness MONOCRYL Plus

Antibacterial Suture. (2005).

8. Surgical, A. R. C., Nw, L. L. C., Rd, J. & Al, I. 510(k) Summary Surgical BIOTRAK Pin

System. (2007).

9. Waterhouse, A. 510(k) Summary Arthrex Bio-Pin. (2005). 156

10. 510(k) Summary Bioretec ActivaPin. (2007).

11. Drug Approval Package: Arestin (Minocycline Hydrochloride) Microspheres.

(2001).

12. Wang, Y.; Qu, W.; Choi, S. FDA’s Regulatory Science Program for Generic PLA/

PLGA-Based Drug Products. Am. Pharm. Rev. (2016).

13. Garner, J. History of PLGA in FDA approved products. ACS Akina, Inc. PolySciTech

Blog (2016). Available at:

https://communities.acs.org/people/jgakinainc/blog/2016/05/27/history-of-plga-

in-fda-approved-products. (Accessed: 24th August 2017)

14. Sung, H., Meredith, C., Johnson, C. & Galis, Z. S. The effect of scaffold degradation

rate on three-dimensional cell growth and angiogenesis. Biomaterials 25, 5735–

5742 (2004).

15. Brannigan, R. P. & Dove, A. P. Biomaterials Science aliphatic polyesters and

polycarbonates Hydrolytic degradation. Biomater. Sci. 5, 9–21 (2017).

16. Brannigan, R. P. & Dove, A. P. Synthesis, properties and biomedical applications of

hydrolytically degradable materials based on aliphatic polyesters and

polycarbonates. Biomater. Sci. 5, 9–21 (2017).

17. Kohane, D. S. & Langer, R. Polymeric Biomaterials in Tissue Engineering. Pediatr.

Res. 63, 487–491 (2008).

18. Alexis, F. Factors affecting the degradation and drug-release mechanism of poly (

lactic acid ) and poly [( lactic acid ) - co - ( glycolic acid )]. 46, 36–46 (2005).

157

19. Pham, Q. P., Sharma, U. & Mikos, A. G. Electrospinning of Polymeric Nanofibers

for Tissue Engineering Applications : A Review. Tissue Eng. 12, 1197–1212 (2006).

20. Wilson, J. A. et al. Magnesium Catalyzed Polymerization of End Functionalized

Poly(propylene maleate) and Poly(propylene fumarate) for 3D Printing of

Bioactive Sca ff olds. J. Am. Chem. Soc. 140, 277–284 (2018).

21. Ulery, B., Nair, L. & Laurencin, C. Biomedical Applications of Biodegradable

Polymers. J Polym Sci B Polym Phys. 49, 832–864 (2012).

22. Sell, S. A. et al. Electrospun polydioxanone-elastin blends: potentials for

bioresorbable vascular grafts. Biomed. Mater. 1, 72–80 (2006).

23. Ma, L., Sabino, M. A. & Gonza, S. Study of the hydrolytic degradation of

polydioxanone PPDX. Polym. Degrad. Stab. 69, 209–216 (2000).

24. Chem, P., Allcock, H. R. & Morozowich, N. L. Polymer Chemistry Bioerodible

polyphosphazenes and their medical potential. Polym. Chem. 3, 578–590 (2012).

25. Peterson, G. I., Childers, E. P., Li, H., Dobrynin, A. V & Becker, M. L. Tunable Shape

Memory Polymers from α ‑ Amino Acid-Based Poly(ester urea)s. Macromolecules

50, 4300–4308 (2017).

26. Li, S., Xy, Y., Yu, J. & Becker, M. L. Enhanced osteogenic activity of poly(ester urea)

scaffolds using facile post-3D printing peptide functionalization strategies.

Biomaterials 141, 176–187 (2017).

27. Yang, L. et al. The degradation of poly(trimethylene carbonate) implants: The role

of molecular weight and enzymes. Polym. Degrad. Stab. 122, 77–87 (2015).

158

28. Zhang, Z., Kuijer, R., Bulstra, S. K., Grijpma, D. W. & Feijen, J. The in vivo and in

vitro degradation behavior of poly(trimethylene carbonate). Biomaterials 27,

1741–1748 (2006).

29. Heller, J. & Barr, J. Poly(ortho esters)-From Concept to Reality.

Biomacromolecules 5, 1625–1632 (2004).

30. Heller, J., Barr, J., Ng, S. Y., Abdellauoi, K. S. & Gurny, R. Poly(ortho esters):

synthesis, characterization, properties and uses. Adv. Drug Deliv Rev. 54, 1015–

1039 (2002).

31. Poets, K. L. & Shipp, D. A. Polyanhydrides: Synthesis, Properties, and Applications.

Aust. J. Chem. 69, 1223–1239 (2016).

32. Ulery, B., Nair, L. & Laurencin, C. Biomedical Applications of Biodegradable

Polymers. J Polym Sci B Polym Phys. 49, 832–864 (2012).

33. Tschan, M. J., Ieong, N. S., Todd, R., Everson, J. & Dove, A. P. Unlocking the

Potential of Poly(Ortho Ester)s: A General Catalytic Approach to the Synthesis of

Surface-Erodible Materials. Angew Chem Int Ed Engl 56, 16664–16668 (2017).

34. Rawla, P., Sunkara, T. & Raj, J. P. Role of biologics and biosimilars in inflammatory

bowel disease : current trends and future perspectives. J. Inflamm. Res. 11, 215–

226 (2018).

35. Oerline, M. et al. Surgical Skill and Complication Rates after Bariatric Surgery. N.

Engl. J. Med. 369, 1434–42 (2013).

36. Anand, A., Sinha, P. A., Kittappa, K. & Brookstein, R. Review of Inguinal Hernia

159

Repairs by Various Surgical Techniques in a District General Hospital in the UK.

Indian J. Surg. 73, 13–18 (2011).

37. Diego, S. & Binion, D. G. Biologic Therapies for Crohn ’ s Disease : Update from

the 2009 ACG Meeting. Gastroenterol. Hepatol. (N. Y). 6, (2010).

38. Krupski, W. C. et al. Coronary-Artery Revascularization before Elective Major

Vascular Surgery. N. Engl. J. Med. 351, 2795–2804 (2004).

39. Legutko, J., Pach, R., Solecki, R., Matyja, A. & Kulig, J. The history of treatment of

groin hernia. Folia Med Cracov 49, 57–74 (2008).

40. Holihan, J. L. et al. Adverse Events after Ventral Hernia Repair : The Vicious Cycle

of Complications. J. Am. Coll. Surg. 221, 478–485 (2015).

41. Trujillo, C. N. et al. Complex Ventral Hernias : A Review of Past to Present. Perm. J.

22, 1–6 (2018).

42. Vasileff, W. K., Nekhline, M. & Kolowich, P. A. Inguinal Hernia in Athletes: Role of

Dynamic Ultrasound. Sports Health 9, 414–21 (2017).

43. Donkor, C. et al. Current perspectives in robotic hernia repair. Robot. Surg. Res.

Rev. 4, 57–67 (2017).

44. Lati, R. et al. Risk-adjusted adverse outcomes in complex abdominal wall hernia

repair with biologic mesh : A case series of 140 patients. Int. Jounal Surg. 43, 26–

32 (2017).

45. Taylor, C. et al. Laparoscopic inguinal hernia repair without mesh fixation , early

results of a large randomised clinical trial. Surg Endosc 22, 757–762 (2008).

160

46. Katri, K. M. Open preperitoneal mesh repair of recurrent inguinal hernia. Hernia

13, 585–589 (2009).

47. Sakorafas, G. H. et al. Open tension free repair of inguinal hernias; the

Lichtenstein technique. BMC Surg. 1, (2001).

48. Luijendijk, R. W. et al. A COMPARISON OF SUTURE REPAIR WITH MESH REPAIR

FOR INCISIONAL HERNIA. N. Engl. J. Med. 343, 392–398 (2000).

49. Toro, A., Mannino, M., Reale, G. & Di Carlo, I. TachoSil use in abdominal surgery :

a review. J. Blood Med. 2, 31–36 (2011).

50. Dunlop, D. et al. Open Mesh versus Laparoscopic Mesh Repair of Inguinal Hernia.

N. Engl. J. Med. 350, 1819–1827 (2004).

51. Misiakos, E. P., Machairas, A., Patapis, P. & Liakakos, T. Laparoscopic Ventral

Hernia Repair : Pros and Cons Compared With Open Hernia Repair. JSLS 12, 117–

125 (2008).

52. Slater, N. J. et al. Criteria for definition of a complex abdominal wall hernia.

Hernia 18, (2014).

53. Demir, U. et al. Comparison of prosthetic materials in incisional hernia repair.

Surg. Today 35, 223–227 (2005).

54. Cobb, W. S. et al. Textile Analysis of Heavy Weight, Mid-Weight, and Light Weight

Polypropylene Mesh in a Porcine Ventral Hernia Model. J. Surg. Res. 136, 1–7

(2006).

55. Lopresti, S. et al. Macrophage polarization in response to ECM coated

161

polypropylene mesh. Biomaterials 35, 6838–6849 (2015).

56. Rosen, M. J. Polyester-based mesh for ventral hernia repair: is it safe? Am. J. Surg.

197, 353–359 (2009).

57. Klinge, U., Klosterhalfen, B., Ottinger, A. P., Junge, K. & Schumpelick, V. PVDF as a

new polymer for the construction of surgical meshes. Biomaterials 23, 3487–3493

(2002).

58. Novitsky, Y. W. et al. Comparative Evaluation of Adhesion Formation, Strength of

Ingrowth, and Textile Properties of Prosthetic Meshes After Long-Term Intra-

Abdominal Implantation in a Rabbit. J. Surg. Res. 140, 6–11 (2007).

59. Shah, B. C. et al. Does type of mesh used have an impact on outcomes in

laparoscopic inguinal hernia? Am. J. Surg. 198, 759–764 (2009).

60. Lee, H., Gu, L., Mooney, D. J., Levenston, M. E. & Chaudhuri, O. Mechanical

confinement regulates cartilage matrix formation by chondrocytes. Nat. Mater.

16, 1243–1255 (2017).

61. Vining, K. H. & Mooney, D. J. Mechanical forces direct stem cell behaviour in

development and regeneration. Nat. Rev. Mol. Cell Biol. 18, 728–742 (2017).

62. Mizuno, K., Tolmachova, T., Ushakov, D. S., Romao, M. & Ferenczi, M. A. Rab27b

Regulates Mast Cell Granule Dynamics and Secretion. Traffic 8, 883–892 (2007).

63. Rock, K. L. & Kono, H. The inflammatory response to cell death. Annu Rev Pathol

3, 99–126 (2011).

64. Conze, J. et al. Polypropylene in the intra-abdominal position: Influence of pore

162

size and surface area. Hernia 8, 365–372 (2004).

65. Lambrecht, J. R. et al. Prophylactic mesh at end-colostomy construction reduces

parastomal hernia rate: A randomized trial. Color. Dis. 17, 191–197 (2015).

66. Pomahac, B. & Aflaki, P. Use of a non-cross-linked porcine dermal scaffold in

abdominal wall reconstruction. Am. J. Surg. 199, 22–27 (2010).

67. Deeken, C. R. & Matthews, B. D. Characterization of the Mechanical Strength,

Resorption Properties, and Histologic Characteristics of a Fully Absorbable

Material (Poly-4-hydroxybutyrate-PHASIX Mesh) in a Porcine Model of Hernia

Repair. ISRN Surg 2013, 1–12 (2013).

68. Slack, S. M. & Horbett, T. A. Changes in the Strength of Fibrinogen Attachment to

Solid Surfaces : An Explanation of the Influence of Surface Chemistry on the

Vroman Effect. J. Colloid Interface Sci. 133, (1989).

69. Von Eiff, C., Jansen, B., Kohnen, W. & Becker, K. Infections associated with

medical devices: Pathogenesis, management and prophylaxis. Drugs 65, 179–214

(2005).

70. Zander, Z. K. & Becker, M. L. Antimicrobial and Antifouling Strategies for

Polymeric Medical Devices. ACS Macro Lett. 7, 16–25 (2018).

71. Boersema, G. S. A., Grotenhuis, N., Bayon, Y., Lange, J. F. & Bastiaansen-

jenniskens, Y. M. The Effect of Biomaterials Used for Tissue Regeneration

Purposes on Polarization of Macrophages. BioResearch 5, 6–14 (2016).

72. Sridharan, R., Cameron, A. R., Kelly, D. J., Kearney, C. J. & Brien, F. J. O.

163

Biomaterial based modulation of macrophage polarization : a review and

suggested design principles. Mater. Today 18, 313–325 (2015).

73. Bertani, F. R. et al. Classification of M1 / M2-polarized human macrophages by

label-free hyperspectral reflectance confocal microscopy and multivariate

analysis. Sci. Rep. 7, 1–9 (2017).

74. Köckerling, F. & Schug-Pass, C. What do we know about titanized polypropylene

meshes? An evidence-based review of the literature. Hernia 18, (2014).

75. Kulig, K. M. et al. Biologic properties of surgical scaffold materials derived from

dermal ECM. Biomaterials 1–9 (2013). doi:10.1016/j.biomaterials.2013.02.055

76. Wolf, M. T. et al. Polypropylene Surgical Mesh Coated with Extracellular Matrix

Mitigates the Host Foreign Body Response. J Biomed Mater Res A 102, 234–246

(2014).

77. Choi, Y. C. et al. Decellularized Extracellular Matrix Derived from Porcine Adipose

Tissue as a Xenogeneic Biomaterial for Tissue Engineering. Tissue Eng. 18, 1–11

(2012).

78. Badylak, S. F., Freytes, D. O. & Gilbert, T. W. Extracellular matrix as a biological

scaffold material: Structure and function. Acta Biomater. 5, 1–13 (2009).

79. Dreger, N. Z. et al. Amino acid-based Poly ( ester urea ) copolymer films for

hernia-repair applications. Biomaterials 182, 44–57 (2018).

80. Engelberg, I. & Kohn, J. Physico-mechanical properties of degradable polymers

used in medical applications: A comparative study. Biomaterials 12, 292–304

164

(1991).

81. Kohn, J. & Langer, R. Polymerization Reactions Involving the Side Chains of α-L-

Amino Acids. J. Am. Chem. Soc. 109, 817–820 (1987).

82. Hooper, K. A., Macon, N. D. & Kohn, J. Comparative histological evaluation of new

tyrosine-derived polymers and poly (L-lactic acid) as a function of polymer

degradation. J. Biomed. Mater. Res. 41, 443–454 (1998).

83. Bourke, S. L. & Kohn, J. Polymers derived from the amino acid L-tyrosine:

Polycarbonates, polyarylates and copolymers with poly(ethylene glycol). Adv.

Drug Deliv. Rev. 55, 447–466 (2003).

84. Yu, J., Lin, F., Lin, P., Gao, Y. & Becker, M. L. Phenylalanine-based poly(ester urea):

Synthesis, characterization, and in vitro degradation. Macromolecules 47, 121–

129 (2014).

85. Gao, Y., Childers, E. P. & Becker, M. L. L-Leucine-Based Poly ( ester urea ) s for

Vascular Tissue Engineering. ACS Biomater. Sci. Eng. 1, 795–804 (2015).

86. Yu, J., Lin, F., Lin, P., Gao, Y. & Becker, M. L. Phenylalanine-Based Poly(ester urea):

Synthesis, Characterization, and in vitro Degradation. Macromolecules 47, 121–

129 (2014).

87. Yu, J., Lin, F. & Becker, M. L. Branched amino acid based poly(ester urea)s with

tunable thermal and water uptake properties. Macromolecules 48, 2916–2924

(2015).

88. Dreger, N. Z. et al. Preclinical in vitro and in vivo Assessment of Linear and

165

Branched L -Valine Based Poly ( ester urea ) s for Soft-tissue Applications. ACS

Biomater. Sci. Eng. 4, 1346–1356 (2018).

89. Agarwal, B. B., Agarwal, K. A., Sahu, T. & Mahajan, K. C. Traditional polypropylene

and lightweight meshes in totally extraperitoneal inguinal herniorrhaphy. Int. J.

Surg. 8, 44–47 (2010).

90. Shah, B. C. et al. Does type of mesh used have an impact on outcomes in

laparoscopic inguinal hernia? Am. J. Surg. 198, 759–764 (2009).

91. Dabbas, N. Frequency of abdominal wall hernias : is classical teaching out of

date ? Clin. Rev. 336, 1–6 (2011).

92. Jenkins, J. T. & Dwyer, P. J. O. Inguinal hernias. J R Soc Med Sh Rep 336, 269–272

(2008).

93. Hee, R. Van. History of Hernia. Jurnalul Chir. 7, (2011).

94. Peter, L., Andersen, H., Klein, M., Gögenur, I. & Rosenberg, J. Long-term

recurrence and complication rates after incisional hernia repair with the open

onlay technique. BMC Surg. 5, 1–5 (2009).

95. Mcgreevy, J. M. et al. A prospective study comparing the complication rates

between laparoscopic and open ventral hernia repairs. 1778–1780 (2003).

doi:10.1007/s00464-002-8851-5

96. Goodney, P. P., Birkmeyer, C. M. & Birkmeyer, J. D. Short-term Outcomes of

Laparoscopic and Open Ventral Hernia Repair. 137, (2017).

97. Breuing, K. et al. Incisional ventral hernias: Review of the literature and

166

recommendations regarding the grading and technique of repair. Surgery 148,

544–558 (2010).

98. Dedemadi, G. et al. Laparoscopic versus open mesh repair for recurrent inguinal

hernia: A meta-analysis of outcomes. Am. J. Surg. 200, 291–297 (2010).

99. Hwang, C. S., Wichterman, K. A. & Alfrey, E. J. Laparoscopic ventral hernia repair

is safer than open repair: analysis of the NSQIP data. J Surg Res 156, 213–6 LID–

10.1016/j.js (2009).

100. Sadava, E. E., Krpata, D. M., Gao, Y., Rosen, M. J. & Novitsky, Y. W. Wound healing

process and mediators: Implications for modulations for hernia repair and mesh

integration. J. Biomed. Mater. Res. - Part A 295–302 (2014).

doi:10.1002/jbm.a.34676

101. Boland, G. M. & Weigel, R. J. Formation and Prevention of Postoperative

Abdominal Adhesions. J. Surg. Res. 132, 3–12 (2006).

102. Fitzgerald, J. F. & Kumar, A. S. Biologic versus synthetic mesh reinforcement:

What are the pros and cons? Clin. Colon Rectal Surg. 27, 140–148 (2014).

103. Berger, D. & Bientzle, M. Polyvinylidene fluoride: A suitable mesh material for

laparoscopic incisional and parastomal hernia repair! Hernia 13, 167–172 (2009).

104. Brown, C. N. & Finch, J. G. Which mesh for hernia repair? Ann. R. Coll. Surg. Engl.

92, 272–278 (2010).

105. Ebersole, G. C. et al. Development of novel electrospun absorbable

polycaprolactone (PCL) scaffolds for hernia repair applications. Surg. Endosc.

167

Other Interv. Tech. 26, 2717–2728 (2012).

106. Hashizume, R. et al. Morphological and mechanical characteristics of the

reconstructed rat abdominal wall following use of a wet electrospun

biodegradable polyurethane elastomer scaffold. Biomaterials 31, 3253–3265

(2010).

107. Pascual, G., Pérez-Köhler, B., Rodríguez, M., Sotomayor, S. & Bellón, J. M.

Postimplantation host tissue response and biodegradation of biologic versus

polymer meshes implanted in an intraperitoneal position. Surg. Endosc. Other

Interv. Tech. 28, 559–569 (2014).

108. Klink, C. D. et al. Comparison of long-term biocompability of PVDF and PP meshes.

J. Invest. Surg. 24, 292–299 (2011).

109. You, H. et al. Biocompatibility and resorption pattern of newly developed

hyaluronic acid hydrogel reinforced histologic observation in rabbit calvarial

defect model. Biomater. Res. 18, 1–7 (2014).

110. Maurice, S. M. & Skeete, D. A. Use of human acellular dermal matrix for

abdominal wall reconstructions. Am. J. Surg. 197, 35–42 (2009).

111. Blatnik, J., Jin, J. & Rosen, M. Abdominal hernia repair with bridging acellular

dermal matrix-an expensive hernia sac. Am. J. Surg. 196, 47–50 (2008).

112. Kim, H., Bruen, K. & Vargo, D. Acellular dermal matrix in the management of high-

risk abdominal wall defects. Am. J. Surg. 192, 705–709 (2006).

113. Candage, R. et al. Use of human acellular dermal matrix for hernia repair: Friend

168

or foe? Surgery 144, 703–711 (2008).

114. Suwandi, J. S., Toes, R. E. M., Nikolic, T. & Roep, B. O. Inducing tissue specific

tolerance in autoimmune disease with tolerogenic dendritic cells. Clin. Exp.

Rheumatol. 33, 97–103 (2015).

115. Badylak, S. F. & Gilbert, T. W. Immune response to biologic scaffold materials.

Semin. Immunol. 20, 109–116 (2008).

116. Wade, M. B., Rodenberg, E., Patel, U., Shah, B. & Becker, M. L. Influence of

Sterilization Technologies on Electrospun Poly(ester urea)s for Soft Tissue Repair.

Biomacromolecules 17, 3363–3374 (2016).

117. Gao, Y. et al. Pilot Mouse Study of 1 mm Inner Diameter (ID) Vascular Graft Using

Electrospun Poly(ester urea) Nanofibers. Adv. Healthc. Mater. 5, 2427–2436

(2016).

118. Childers, E. P. et al. Adhesion of Blood Plasma Proteins and Platelet-rich Plasma

on L-Valine-Based Poly(ester urea). Biomacromolecules 3396–3403 (2016).

doi:10.1021/acs.biomac.6b01195

119. Peterson, G. I., Dobrynin, A. V & Becker, M. L. α ‑ Amino Acid-Based Poly(Ester

urea)s as Multishape Memory Polymers for Biomedical Applications. 10–13

(2016). doi:10.1021/acsmacrolett.6b00648

120. Franz, S., Rammelt, S., Scharnweber, D. & Simon, J. C. Biomaterials Immune

responses to implants e A review of the implications for the design of

immunomodulatory biomaterials. Biomaterials 32, 6692–6709 (2011).

169

121. Lee, T. T. et al. Light-triggered in vivo activation of adhesive peptides regulates

cell adhesion, inflammation and vascularization of biomaterials. Nat. Mater. 14,

352–360 (2015).

122. Kalaba, S., Gerhard, E., Winder, J. S., Pauli, E. M. & Haluck, R. S. Bioactive

Materials Design strategies and applications of biomaterials and devices for

Hernia repair. Bioact. Mater. 1, 2–17 (2016).

123. Z, J., Paul, M., O, T., M, F. & M, J. M. Polyurethane-Covered Dacron Mesh Versus

Polytetrafluoroethylene DualMesh for Intraperitoneal Hernia Repair in Rats. Surg.

Today 32, 884–886 (2002).

124. Garcia-Urena, M. A. et al. Differences in polypropylene shrinkage depending on

mesh position in an experimental study. Am. J. Surg. 193, 538–542 (2007).

125. Zhu, L., Schuster, P. & Klinge, U. Mesh implants : An overview of crucial mesh

parameters. World J. Gastrointest Surg. 7, 226–236 (2015).

126. Sadowski, B. et al. Comparison of polypropylene versus polyester mesh in the

Lichtenstein hernia repair with respect to chronic pain and discomfort. Hernia 15,

643–654 (2011).

127. Hiles, M., Ritchie, R. D. R. & Altizer, A. M. Are Biologic Grafts Effective for Hernia

Repair? A Systematic Review of the Literature. Surg. Innov. 16, 26–37 (2009).

128. Novitsky, Y. W. et al. Comparative Evaluation of Adhesion Formation, Strength of

Ingrowth, and Textile Properties of Prosthetic Meshes After Long-Term Intra-

Abdominal Implantation in a Rabbit. J. Surg. Res. 140, 6–11 (2007).

170

129. Sbitany, H. et al. Outcomes Analysis of Biologic Mesh Use for Abdominal Wall

Reconstruction in Clean-Contaminated and Contaminated Ventral Hernia Repair.

Reconstr. Surg. 75, 201–204 (2015).

130. Bittner, J. G. et al. First human use of hybrid synthetic/biologic mesh in ventral

hernia repair: a multicenter trial. Surg. Endosc. 32, 1123–1130 (2017).

131. Kartvelishvili, T., Tsitlanadze, G., Edilashvili, L., Japaridze, N. & Katsarava, R. Amino

acid based bioanalogous polymers. Novel regular poly(ester urethane)s and

poly(ester urea)s based on bis(L-phenylalanine) α,ω-alkylene diesters. Macromol.

Chem. Phys. 198, 1921–1932 (1997).

132. Gomurashvili, Z. D., Katsarava, R. & Tugushi, D. Poly(ester urea) Polymers and

Methods of Use.

133. Murase, S. K. et al. Amino acid-based poly(ester amide) nanofibers for tailored

enzymatic degradation prepared by miniemulsion-electrospinning. RSC Adv.2 5,

55006–55014 (2015).

134. Gomurashvili, Z., Kricheldorf, H. R. & Katsarava, R. Amino Acid Based

Bioanalogous Polymers. Synthesis and Study of New Poly(ester amide)s

Composed of Hydrophobic α-Amino Acids and Dianhydrohexitoles. J. Macromol.

Sci. Part A 3, 215–227 (2000).

135. Childers, E. P. et al. Enhanced Rotator-Cuff Repair Using Platelet-Rich Plasma

Adsorbed on Branched Poly(ester urea)s. Biomacromolecules 19, 3129–3139

(2018).

171

136. Policastro, G. M. et al. OGP Functionalized Phenylalanine-Based Poly(ester urea)

for Enhancing Osteoinductive Potential of Human Mesenchymal Stem Cells.

Biomacromolecules 16, 1358–1371 (2015).

137. Bastos, Eduardo Lemos de Souza; Fagundes, D. J., Taha, M. O., Novo, N. F. &

Juliano, Y. The role of bovine preserved peritoneum in rats ventral hernia. A

histological evaluation. Acta Cir Bras 21, 328–331 (2006).

138. Ulrich, D. et al. A preclinical evaluation of polypropylene / polylacticacid hybrid

meshes for fascial defect repair using a rat abdominal hernia model. PLoS One 1–

11 (2017). doi:https://doi.org/10.1371/journal.pone.0179246

139. Sahoo, S., Delozier, K. R., Erdemir, A. & Derwin, K. A. Clinically relevant

mechanical testing of hernia graft constructs. J. Mech. Behav. Biomed. Mater. 41,

177–188 (2015).

140. Kalaba, S., Gerhard, E., Winder, J. S., Pauli, E. M. & Haluck, R. S. Bioactive

Materials Design strategies and applications of biomaterials and devices for

Hernia repair. Bioact. Mater. 1, 2–17 (2016).

141. Chang, C. W. et al. Mesenchymal stem cell seeding of porcine small intestinal

submucosal extracellular matrix for cardiovascular applications. PLoS One 11, 1–

19 (2016).

142. Sahoo, S., DeLozier, K. R., Dumm, R. A., Rosen, M. J. & Derwin, K. A. Fiber-

reinforced dermis graft for ventral hernia repair. J. Mech. Behav. Biomed. Mater.

34, 320–329 (2014).

172

143. Konerding, M. A., Chantereau, P., Delventhal, V., Holste, J. L. & Ackermann, M.

Biomechanical and histological evaluation of abdominal wall compliance with

intraperitoneal onlay mesh implants in rabbits: A comparison of six different

state-of-the-art meshes. Med. Eng. Phys. 34, 806–816 (2012).

144. Holihan, J. L. et al. Mesh Location in Open Ventral Hernia Repair: A Systematic

Review and Network Meta-analysis. World J. Surg. 40, 89–99 (2016).

145. Xia, X. et al. Two cases about mesh adhesion to intra-abdominal cavity tissue after

using mesh to repair an incisional hernia. Pak J Med Sci 33, 1018–1021 (2017).

146. Van ’T Riet, M. et al. Prevention of adhesion to prosthetic mesh: Comparison of

different barriers using an incisional hernia model. Ann. Surg. 237, 123–128

(2003).

147. Muller, S., Langø, T., Brekken, R. & Ystgaard, B. Degree of Adhesions After Repair

of Incisional Hernia. JSLS J. Soc. Laparoendosc. Surg. 14, 399–404 (2010).

148. Baptista, M. L., Bonsack, M. E. & Delaney, J. P. Seprafilm reduces adhesions to

polypropylene mesh. Surgery 128, 86–92 (2000).

149. Yu, Q., Zhang, Y., Wang, H., Brash, J. & Chen, H. Anti-fouling bioactive surfaces.

Acta Biomater. 7, 1550–1557 (2011).

150. Lambertz, A. et al. Analysis of adhesion formation of a new elastic thermoplastic

polyurethane (TPU) mesh in comparison to polypropylene (PP) meshes in IPOM

position. J. Mech. Behav. Biomed. Mater. 53, (2016).

151. Campanelli, G., Canziani, M., Frattini, F., Cavalli, M. & Agrusti, S. Inguinal hernia:

173

State of the art. Int. J. Surg. 6, 26–28 (2008).

152. Langbach, O., Holmedal, S. H., Grandal, O. J. & Røkke, O. Adhesions to Mesh after

Ventral Hernia Mesh Repair Are Detected by MRI but Are Not a Cause of Long

Term Chronic Abdominal Pain. Gastroenterol. Res. Pract. 2016, (2016).

153. Takács, I. et al. Tissular incorporation of different silicone covered polypropylene

meshes is distinct. Magy. Sebészet 63, 340–346 (2010).

154. Lin, L. X. et al. Evaluation of surgical anti-Adhesion products to reduce

postsurgical intra-Abdominal adhesion formation in a rat model. PLoS One 12, 1–

9 (2017).

155. Banerjee, I., Pangule, R. C. & Kane, R. S. Antifouling coatings: Recent

developments in the design of surfaces that prevent fouling by proteins, bacteria,

and marine organisms. Adv. Mater. 23, 690–718 (2011).

156. Yan, S. et al. Hierarchical Polymer Brushes with Dominant Antibacterial

Mechanisms Switching from Bactericidal to Bacteria Repellent.

Biomacromolecules 17, 1696–1704 (2016).

157. Candage, R. et al. Use of human acellular dermal matrix for hernia repair: Friend

or foe? Surgery (2008). doi:10.1016/j.surg.2008.06.018

158. Kim, H., Bruen, K. & Vargo, D. Acellular dermal matrix in the management of high-

risk abdominal wall defects. Am. J. Surg. (2006).

doi:10.1016/j.amjsurg.2006.09.003

159. Basta, M. N., Fischer, J. P. & Kovach, S. J. Assessing complications and cost-

174

utilization in ventral hernia repair utilizing biologic mesh in a bridged underlay

technique. Am. J. Surg. 209, 695–702 (2015).

160. Primus, F. E. & Harris, H. W. A critical review of biologic mesh use in ventral

hernia repairs under contaminated conditions. Hernia (2013).

doi:10.1007/s10029-012-1037-8

161. Yu, J., Lin, F. & Becker, M. L. Branched amino acid based poly(ester urea)s with

tunable thermal and water uptake properties. Macromolecules 48, 2916–2924

(2015).

162. Zander, Z. K. et al. Biomaterials Post-fabrication QAC-functionalized thermoplastic

polyurethane for contact-killing catheter applications. Biomaterials 178, 339–350

(2018).

163. Hodde, J. P., Suckow, M. D., Johnson, C., Rodenberg, E. & Ritchie, R. D.

Histological and adhesiogenic characterization of the Zenapro Hybrid Hernia

Repair Device. Int. J. Surg. Open 5, 27–32 (2016).

164. Suckow, M. A. et al. Addition of nimesulide to small intestinal submucosa

biomaterial inhibits postsurgical adhesiogenesis in rats. J Biomed Mat Res Pt B

Appl Biomater 93B, 18–23 (2010).

165. Chen, S., Li, L., Zhao, C. & Zheng, J. Surface hydration: Principles and applications

toward low-fouling/nonfouling biomaterials. Polymer (Guildf). 51, 5283–5293

(2010).

166. Damodaran, V. B. & Murthy, S. N. Bio-inspired strategies for designing antifouling

175

biomaterials. Biomater. Res. 20, 1–11 (2016).

167. Shen, M., Garcia, I., Maier, R. V. & Horbett, T. A. Effects of adsorbed proteins and

surface chemistry on foreign body giant cell formation, tumor necrosis factor

alpha release and procoagulant activity of monocytes. J. Biomed. Mater. Res. 70A,

533–541 (2004).

176

APPENDICES

177

APPENDIX A-SUPPORTING FIGURES

Figure 6.1. A) 1H-NMR overlay shows the spectra of linear monomers 1-VAL-8, 1-VAL-10, and 1-VAL-12. The resonances “b” indicated in the light blue highlighted regions denote the methylene resonances for each of the three monomers. B) 1H-NMR of the Triol-TYR is shown (top spectrum), and successful Boc-deprotection (bottom spectrum) was identified by the disappearance of the methylene resonance at 1.28 ppm and the appearance of the broad amine proton resonance between 8.72-8.78 ppm, both of which are highlighted in blue.

178

179

Figure 6.2. A) 1H-NMR overlay spectra of linear poly(ester urea)s poly(1-VAL-8), poly(1- VAL-10), and poly(1-VAL-12). The highlighted resonances “b” indicate the methylene resonances from the diol chain lengths.for each of the three polymers. B) 1H-NMR spectra of branched poly[(1-VAL-8)0.98-co-(Triol-TYR)0.02] and C) poly[(1-VAL-10)0.98-co-(Triol- TYR)0.02] are shown. The degree of branching was determined by integration of the six methylene protons denoted “e” from the Triol-TYR monomer and comparing them to the twelve methyl l-valine protons denoted “n” from the linear monomers.

180

Figure 6.3. Thermogravimetric analysis (TGA) of the linear (A) and branched polymers (B) shows that the degradation temperatures for these materials is well above the temperature required for compression molding.

181

. Figure 6.4. Differential scanning calorimetry (DSC) of the linear and branched polymers shows that the glass transition temperatures of these materials are near physiological conditions. The glass transition temperatures are significantly lower than the degradation temperature which allowed for these materials to be processed.

182

Figure 6.5. Size-exclusion chromatographs for each PEU synthesized in this study comparing the initial molecular mass, post EtO sterilization, and the 2 and 3 month time points in vivo for (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) 2% branched P(1- VAL-8), and (E) 2% branched P(1-VAL-10).

183

Figure 6.6. Stress-strain curves for (A) P(1-VAL-8), (B) P(1-VAL-10), (C) P(1-VAL-12), (D) PP, (E) 2% branched P(1-VAL-8), and (F) 2% branched P(1-VAL-10) were obtained from tensile tests performed at 25 °C with an extension rate of 25.4 mm/min. All mechanical data were extrapolated from the curves which represent an average of 4-6 samples.

184

Figure 6.7. 1H-NMR monomers 1-PHE-8 monomer shows successful synthesis based on the characteristic L-phenylalanine aromatic resonances denoted ‘d, e, f’ and the p- toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties.

185

Figure 6.8. 1H-NMR monomers 1-PHE-6 monomer shows successful synthesis based on the characteristic L-phenylalanine aromatic resonances denoted ‘d, e, f’ and the p- toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties.

186

Figure 6.9. 1H-NMR of 1-VAL-8 monomer shows successful synthesis based on the characteristic L-valine methyl resonance denoted ‘k’ and p-toluenesulfonic acid aromatic resonances. Integration confirms that this monomer is a bifunctional monomer with two protonated amine moieties.

187

Figure 6.10. 1H-NMR overlay of P(1-VAL-8) and PHE6 poly(1-VAL-8) PEU polymers show successful synthesis. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine and the methylene resonances from L-phenylalanine denoted ‘l’ in blue. As the molar composition rises from 10-30% more of the L-phenylalanine resonances can be observed.

188

Figure 6.11. 1H-NMR overlay of P(1-VAL-8) and PHE8 poly(1-VAL-8) PEU polymers show successful synthesis. The monomer molar composition in the afforded polymers were calculated from the characteristic ‘a’ resonances in pink from L-valine and the methylene resonances from L-phenylalanine denoted ‘l’ in blue. As the molar composition rises from 10-30% more of the L-phenylalanine resonances can be observed.

189

Figure 6.12. 13C-NMR of PEU copolymers shows successful synthesis. The characteristic L-valine methyl resonances are observed between 18-20 ppm while the L-phenylalanine ring resonances can be seen with between 125-136 ppm.

190

Figure 6.13. Water contact angle was used to determine surface properties of the six copolymers and how they compared to p(1-VAL-8) (n = 3) (D). An increase in water contact was observed for all copolymers with incorporation of L-phenylalanine content.

191

Figure 6.14. Relative stiffness of PEU-ECM composite films was recorded by dividing the force at break by the extension at break (A). There was no significant difference among any samples which indicates that the films force and extension at break is proportional to that of free standing ECM. Force at break for ECM and PEU-ECM composite films were recorded from the force versus extension curves (* indicates p value < 0.05 between ECM and 10% PHE8 P(1-VAL-8) and between ECM and 30% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8) and between 10% PHE6 P(1-VAL-8) and 30% PHE8 P(1-VAL-8). *** indicated p value < 0.05 between 10% PHE8 P(1-VAL-8) and 20% PHE8 P(1-VAL-8). **** indicates p value < 0.05 between 20% PHE8 P(1-VAL-8) and 30% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (B). Extension at break for ECM and PEU-ECM composite films were recorded from the force versus extension curves (* indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (C). Relative stiffness of free-standing films was recorded calculated the same way as previously described (* indicates p value < 0.05 between ECM and all six copolymers. ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (D). Force at break for free-standing films were recorded from the force versus extension curves (* 192 indicates p value < 0.05 between ECM and 10% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 10% PHE6 P(1-VAL-8) and 10% PHE8 P(1-VAL-8) and between 10% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL-8), n = 3 samples). No other samples were significantly different (E). Extension at break for free-standing films were recorded from the force versus extension curves (* indicates p value < 0.05 between ECM and 10% PHE6 P(1-VAL- 8), between ECM and 20% PHE6 P(1-VAL-8), between 30% PHE6 P(1-VAL-8), and between ECM and 30% PHE8 P(1-VAL-8). ** indicates p value < 0.05 between 20% PHE6 P(1-VAL- 8) and 10% PHE8 P(1-VAL-8) and between 20% PHE6 P(1-VAL-8) and 20% PHE8 P(1-VAL- 8), n = 3 samples). No other samples were significantly different (F).

193

Figure 6.15. Sprague-Dawley rat showing abdominal bulge indicating that the abdominal incisional model can induce a hernia.

194

Figure 6.16. Hydrolytic degradation rates for the fabricated films, SIS-ECM, 30% PHE6 P(1- VAL-8)-ECM in 0.025 M NaOH (A) and 30% PHE6 P(1-VAL-8) films in 0.1 M NaOH (n = 3).

195

Figure 6.17. Size-exclusion chromatography was used to obtain the molecular mass of the 30% PHE6 P(1-VAL-8) copolymer prior to implantation (red) and after 14 day implantation (black).

196

Figure 6.18. 1H-NMR of 1-TYR-2 Alloc monomer shows successful synthesis based on the successful deprotection noted by the disappearance of the boc peak and appearance of the broad amine.

197

Figure 6.19. 1H-NMR of 3,3’-dithiodipropionyl chloride.

198

Figure 6.20. 1H-NMR of bis(3-(dimethylamino)propyl) 3,3'-disulfanediyldipropionate.

199

Figure 6.21. 1H-NMR of zwitterion disulfide.

200

Figure 6.22. 13C-NMR of zwitterion disulfide.

201

Figure 6.23. 1H-NMR of 3-((3-((3-mercaptopropanoyl)oxy)propyl)di-methylammon- io)propane-1-sulfonate (zwitterion-SH).

202

Figure 6.24. 13C-NMR of 3-((3-((3-mercaptopropanoyl)oxy)propyl)di-methylammon- io)propane-1-sulfonate (zwitterion-SH).

203

Figure 6.25. Size-exclusion chromatography of 5 and 10% alloc-PEUs in THF.

204

Figure 6.26. Differential scanning calorimetry of 5 and 10% alloc-PEUs.

205

Figure 6.27. Thermogravimetric analysis of 5 and 10% alloc-PEUs.

206

Figure 6.28. 1000 MW FITC-PEG-SH calibration curve with concentration values reported in triplicate and mean intensity of emission at 550 nm recorded.

207

Figure 6.29. 1H-NMR of lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP).

208

Figure 6.30. Extent of adhesion images for implanted 10% alloc-PEU materials after three weeks of implantation. Highlighted areas show adhesions that circumvented the device and adhesions that formed to the sutures used to retain the 10% alloc-PEU blank device (A). The highlighted area shows adhesions circumventing the device to form on the abraded cecum underneath for the 10% alloc-PEU functionalized device (B).

209

APPENDIX B-SUPPORTING SCHEMES AND TABLES

Table 6.1. Fibrous capsule histology measurements

Fibrous Capsule 2 Month (μm) 3 Month (μm) Thickness

P(1-VAL-8) 117.7 ± 48.6 87.1 ± 16.4

P(1-VAL-10) 105.7 ± 27.0 80.3 ± 37.8

P(1-VAL-12) 67.3 ± 33.4 87.6 ± 21.5

2% Branched P(1-VAL-8) 77.0 ± 21.6 84.0 ± 25.9

2% Branched P(1-VAL-10) 97.0 ± 26.9 103.0 ± 32.7

Poly(propylene) 123.5 ± 29.8 126.2 ± 34.1

210

Table 6.2. Burst-test PEU-ECM mechanical properties comparison

Relative Stiffness Force at Break Extension at Break Polymer (N/cm) (N) (cm) ECM 82.4 ± 8.6 101.3 ± 1.6 1.2 ± 0.1

10% PHE8 149.4 ± 7.9 1.9 ± 0.4 P(1-VAL-8)-ECM 83.0 ± 16.2

20% PHE8 75.5 ± 9.2 103.7 ± 9.9 1.4 ± < 0.1 P(1-VAL-8)-ECM

30% PHE8 93.2 ± 2.4 151.4 ± 11.3 1.6 ± 0.1 P(1-VAL-8)-ECM

10% PHE6 89.1 ± 3.0 102.6 ± 6.5 1.2 ± 0.1 P(1-VAL-8)-ECM

20% PHE6 75.0 ± 12.7 122.5 ± 30.0 1.7 ± 0.4 P(1-VAL-8)-ECM

30% PHE6 84.4 ± 6.1 123.9 ± 8.5 1.5 ± < 0.1 P(1-VAL-8)-ECM

211

Table 6.3. Free-standing film burst-test mechanical properties

Relative Stiffness Force at Break Extension at Break Polymer (N/cm) (N) (cm) ECM 82.4 ± 8.6 101.3 ± 1.6 1.2 ± 0.1

10% PHE8 17..5 ± 4.9 54.2 ± 9.8 3.2 ± 0.8 P(1-VAL-8)

20% PHE8 20.8 ± 1.6 64.1 ± 27.0 3.1 ± 1.4 P(1-VAL-8)

30% PHE8 23.2 ± 4.4 88.1 ± 6.4 3.9 ± 0.7 P(1-VAL-8)

10% PHE6 29.1 ± 1.6 118.0 ± 25.3 4.0 ± 0.7 P(1-VAL-8)

20% PHE6 16.1 ± 3.0 90.9 ± 10.0 5.8 ± 1.0 P(1-VAL-8)

30% PHE6 24.2 ± 2.0 110.3 ± 5.0 4.6 ± 0.7 P(1-VAL-8)

212