<%BANNER%>

Engineered Microtopographies and Surface Chemistries Direct Cell Attachment and Function

Permanent Link: http://ufdc.ufl.edu/UFE0042414/00001

Material Information

Title: Engineered Microtopographies and Surface Chemistries Direct Cell Attachment and Function
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Magin, Chelsea
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: adhesion, algae, antifouling, attachment, bacteria, biofouling, biomaterial, cell, contact, elongation, endothelial, flow, fouling, graft, guidance, hydrogel, marine, microtopography, orientation, polydimethylsiloxane, silicone, smooth, topography, vascular
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Harrison, in 1914, first recognized that cells respond to physicochemical cues such as substratum topography when he observed that fibroblasts elongated while cultured on spider silk. Recently, techniques developed in the micro-electronics industry have been used to create molds for producing microscaled topographies with various shapes and spatial arrangements. Although these patterning techniques are well-established, very little is known about the mechanisms underlying cell sensing and response to microtopographies. In this work cellular micro-environments with varying surface topographies and chemistries were evaluated with marine organisms and mammalian cells to investigate cellular sensing and response. Biofouling ? the accumulation of micro-organisms, plants, and animals on submerged surfaces ? is an environmental and economic concern. Engineered topographies, replicated in polydimethylsiloxane elastomer (PDMSe) and functionalized poly(ethylene glycol)-dimethacrylate (PEGDMA) hydrogels, were evaluated for inhibition of marine fouling organism attachment. Microtopographies replicated in PDMSe inhibited attachment of the marine bacterium, Cobetia marina up to 99% versus smooth. The average normalized attachment densities of cells of C. marina and zoospores of the green algae Ulva on PDMSe topographies scaled inversely with the Engineered Roughness Index (ERI II), a representation of surface energy. Attachment densities of Ulva from four assays and C. marina from two growth phases to PDMSe surfaces scaled inversely with one equation: ERIII multiplied by the Reynolds number of the organism (Re) (R2 = 0.77). The same microtopographies created in PDMSe reduced the initial attachment density and attachment strength of cells of the diatoms Navicula incerta and Seminavis robusta compared to smooth PDMSe. The average normalized attachment density of Navicula after exposure to shear stress (48 Pa) was correlated with the contact area between the diatom and a topographically modified surface (R2=0.82). Functionalized PEGDMA hydrogels significantly reduced attachment and attachment strength of Navicula and C. marina. These hydrogels also reduced attachment of zoospores of Ulva compared to PDMSe. Attachment of Ulva to microtopographies in PDMSe and PEGDMA-co-HEMA negatively correlated with ERIII*Re (R2 = 0.94 and R2 = 0.99, respectively). Incorporating a surface energy term into this equation created a correlation between the attachment densities of cells from two evolutionarily diverse groups on substrates of two surface chemistries with an equation that describes the various microtopographies and surface chemistries in terms of surface energy (R2 = 0.80). The current Attachment Model can now be used to design engineered antifouling surface microtopographies and chemistries that inhibit the attachment of organisms from three evoluntionarily diverse groups. Hydrogels based on PEGDMA were also chosen as a substratum material for mammalian cell culture. Capturing endothelial progenitor cells (EPCs) and inducing differentiation into the endothelial cell (EC) phenotype is the ideal way to re-endothelialize a small-diameter vascular graft. Substratum elasticity has been reported to direct stem cell differentiation into specific lineages. Functionalized PEGDMA hydrogels provided good compliance, high fidelity of topographic features and sites for surface modification with biomolecules. Fibronectin grafting and topography both increased EC attachment. This combination of adjustable elasticity, surface chemistry and topography has the potential to promote the capture and differentiation of EPCs into a confluent EC monolayer. Engineered microtopographies replicated in PDMSe directed elongation and alignment of human coronary artery endothelial cells (HCAECs) and human coronary artery smooth muscle cells (HCASMCs) compared to smooth surfaces. Engineered cellular micro-environments were created with specific surface energies defined by chemistry and topography to successfully direct cell attachment and function.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Chelsea Magin.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Brennan, Anthony B.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042414:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042414/00001

Material Information

Title: Engineered Microtopographies and Surface Chemistries Direct Cell Attachment and Function
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Magin, Chelsea
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: adhesion, algae, antifouling, attachment, bacteria, biofouling, biomaterial, cell, contact, elongation, endothelial, flow, fouling, graft, guidance, hydrogel, marine, microtopography, orientation, polydimethylsiloxane, silicone, smooth, topography, vascular
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Harrison, in 1914, first recognized that cells respond to physicochemical cues such as substratum topography when he observed that fibroblasts elongated while cultured on spider silk. Recently, techniques developed in the micro-electronics industry have been used to create molds for producing microscaled topographies with various shapes and spatial arrangements. Although these patterning techniques are well-established, very little is known about the mechanisms underlying cell sensing and response to microtopographies. In this work cellular micro-environments with varying surface topographies and chemistries were evaluated with marine organisms and mammalian cells to investigate cellular sensing and response. Biofouling ? the accumulation of micro-organisms, plants, and animals on submerged surfaces ? is an environmental and economic concern. Engineered topographies, replicated in polydimethylsiloxane elastomer (PDMSe) and functionalized poly(ethylene glycol)-dimethacrylate (PEGDMA) hydrogels, were evaluated for inhibition of marine fouling organism attachment. Microtopographies replicated in PDMSe inhibited attachment of the marine bacterium, Cobetia marina up to 99% versus smooth. The average normalized attachment densities of cells of C. marina and zoospores of the green algae Ulva on PDMSe topographies scaled inversely with the Engineered Roughness Index (ERI II), a representation of surface energy. Attachment densities of Ulva from four assays and C. marina from two growth phases to PDMSe surfaces scaled inversely with one equation: ERIII multiplied by the Reynolds number of the organism (Re) (R2 = 0.77). The same microtopographies created in PDMSe reduced the initial attachment density and attachment strength of cells of the diatoms Navicula incerta and Seminavis robusta compared to smooth PDMSe. The average normalized attachment density of Navicula after exposure to shear stress (48 Pa) was correlated with the contact area between the diatom and a topographically modified surface (R2=0.82). Functionalized PEGDMA hydrogels significantly reduced attachment and attachment strength of Navicula and C. marina. These hydrogels also reduced attachment of zoospores of Ulva compared to PDMSe. Attachment of Ulva to microtopographies in PDMSe and PEGDMA-co-HEMA negatively correlated with ERIII*Re (R2 = 0.94 and R2 = 0.99, respectively). Incorporating a surface energy term into this equation created a correlation between the attachment densities of cells from two evolutionarily diverse groups on substrates of two surface chemistries with an equation that describes the various microtopographies and surface chemistries in terms of surface energy (R2 = 0.80). The current Attachment Model can now be used to design engineered antifouling surface microtopographies and chemistries that inhibit the attachment of organisms from three evoluntionarily diverse groups. Hydrogels based on PEGDMA were also chosen as a substratum material for mammalian cell culture. Capturing endothelial progenitor cells (EPCs) and inducing differentiation into the endothelial cell (EC) phenotype is the ideal way to re-endothelialize a small-diameter vascular graft. Substratum elasticity has been reported to direct stem cell differentiation into specific lineages. Functionalized PEGDMA hydrogels provided good compliance, high fidelity of topographic features and sites for surface modification with biomolecules. Fibronectin grafting and topography both increased EC attachment. This combination of adjustable elasticity, surface chemistry and topography has the potential to promote the capture and differentiation of EPCs into a confluent EC monolayer. Engineered microtopographies replicated in PDMSe directed elongation and alignment of human coronary artery endothelial cells (HCAECs) and human coronary artery smooth muscle cells (HCASMCs) compared to smooth surfaces. Engineered cellular micro-environments were created with specific surface energies defined by chemistry and topography to successfully direct cell attachment and function.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Chelsea Magin.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Brennan, Anthony B.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042414:00001


This item has the following downloads:


Full Text

PAGE 1

1 ENGINEERED MICROTO POGRAPHIES AND SURFACE CHEMISTRIES DIRECT CELL ATTACHMENT AND FUNCTION By CHELSEA MARIE MAGIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Chelsea Marie Magin

PAGE 3

3 To my fianc, Steven J. Kirschner

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my sincere gratitude to my advisor Dr. Anthony Brennan for his guidance and support. I also appreciate the advice and guidance provided by my doctoral committee: Dr. Christopher Batich, Dr. Benjamin Keselowsky, and Dr. Daniel Purich. I must also acknowledge and thank my research collaborators: Dr. Mark Segal, Ms. Larysa (Laura) Sautina, Dr. Maureen Callow, Dr. James Callow, Dr. John Finlay, Dr. Gabriel Lopez Ms. Linnea Ista, and Dr. Michael Schultz I extend special thanks to Jennifer Wrighton her administrative assistance and friendship throughout my graduate studies. Graduate and undergraduate students, past and present, have been essential to my progress. I must first thank Dr. Michelle Carman, Drs. James and Iri s Schumacher, Dr. Leslie Wilson, Dr. Christopher Long, and Mr. Kenneth Chung for their guidanc e and support as mentors. I would also like to thank Dr. Sheema Freeman for teaching me how to properly dispose of biohazardous waste. I must acknowledge the friendship and collaboration provided by m y peers and current members of the Brennan Research Group: Scott Cooper Angel Eijiasi, Dave Jackson, Jack (JiunJeng) Chen, and Julian Sheats. Undergraduate students that worked with me to make my research possible include Matthew Blackburn, Cristina Fernandez, Sara Mendelson, Kern Hast and Michael Showalt er. I also express my great appreciation for the loving support from my mother, Tina Capizzi, and my father, Greg Magin who inspired me to grow up to be an engineer I am truly grateful for the constant encouragement from my dearest friend, Johannah Mahfood, and my loving fianc, Steve Kirschner.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 9 LIST OF FIGURES ........................................................................................................ 10 LIST OF ABBREVIATIONS ........................................................................................... 14 ABSTRACT ................................................................................................................... 15 CHAPTER 1 INTRODUCTION .................................................................................................... 18 Scope of Research ................................................................................................. 18 Specific Aims .......................................................................................................... 18 Specific Aim 1: Determine the Contribution of Organism Size and Shape to the Slope of the Attachment Model using Cobetia marina ............................ 18 Specific Aim 2: Predict Attachment Densities and Attachment Strength of Cells of the Diatoms Navicula incerta and Seminavis robusta to Engineered Microtopographies ..................................................................... 18 Specific Aim 3: Determine the Contribution of Surface Chemistry to the Slope of the Attachment Model using Functionalized Poly(ethylene glycol) Dimethacrylate Hydrogels ............................................................................. 19 Specific Aim 4: Identify a Combination of Surface Chemistry and Topography that Enhances the Attachment of Human Coronary Artery Endothelial Cells or Human Coronary Artery Smooth Muscle Cells .............. 20 2 BACKGROUND ...................................................................................................... 21 Non toxic Antifouling Strategi es .............................................................................. 25 Introduction ....................................................................................................... 25 Natural Antifouling Surfaces ............................................................................. 31 Physicoche mical Antifouling Strategies ............................................................ 33 Physical Antifouling Strategies ......................................................................... 38 Conclusion ........................................................................................................ 44 Influence of Surface Chemistry and Topography on Cell Culture ........................... 44 Contact Guidance ............................................................................................. 44 Medical and Economic Impac t of Controlling Cell Adhesion ............................. 45 Summary ................................................................................................................ 46 3 ENGINEERED ANTIFOULING MICROTOPOGRAPHIES: THE ROLE OF REYNOLDS NUMBER IN A MODEL T HAT PREDICTS ATTACHMENT OF ZOOSPORES OF ULVA AND CELLS OF COBETIA MARINA .............................. 48

PAGE 6

6 Introduction ............................................................................................................. 48 Materials and Methods ............................................................................................ 52 Materials ........................................................................................................... 52 Pattern Designs ................................................................................................ 52 Pattern Fabrication ........................................................................................... 53 Topographical Replication ................................................................................ 53 Sample Preparation .......................................................................................... 53 Bacteria ............................................................................................................ 54 Chemostat culture ...................................................................................... 54 Stationary phase culture ............................................................................ 54 C. marina Attachment Assay ............................................................................ 54 Statistical Methods ........................................................................................... 55 Results and Discussion ........................................................................................... 56 4 ENGINEERED ANTIFOULING TOPOGRAPHIES: INHIBITION OF ATTACHMENT, MOVEMENT AND BIOFILM ADHESION OF NAVICULA INCERTA and SEMINAVIS ROBUSTA .................................................................. 67 Background ............................................................................................................. 67 Materials and Methods ............................................................................................ 71 Materials ........................................................................................................... 71 Pattern Designs ................................................................................................ 71 Sam ple Preparation .......................................................................................... 71 Initial Attachment of Cells ................................................................................. 72 Attachment Strength ......................................................................................... 73 Assessment of Movement ................................................................................ 73 Biofilm Growth and Attachment Strength ......................................................... 74 Statistical Methods ........................................................................................... 74 Results .................................................................................................................... 74 Seminavis Initial Attachment and Attachment Strength .................................... 74 Assessment of Movement of Seminavis and Navicula ..................................... 76 Seminavis Biofilm Growth and Attachment Strength ........................................ 7 7 Navicula Biofilm Growth and Attachment Strength ........................................... 78 Discussion .............................................................................................................. 79 5 ENGINEERED ANTIFOULING MICROTOPOGRAPHIES: THE ROLE OF SURFACE ENERGY OF CROSSLINKED HYDROGELS IN A MODEL THAT P REDICTS ATTACHMENT OF ZOOSPORES OF ULVA AND CELLS OF COBETIA MARINA AND NAVICULA INCERTA ..................................................... 84 Introduction ............................................................................................................. 84 Experimental Section .............................................................................................. 88 Materials ........................................................................................................... 88 Sample Preparation .......................................................................................... 88 Microtopography Character ization .................................................................... 90 Chemical Composition ...................................................................................... 91 Surface Energy Measurements ........................................................................ 91

PAGE 7

7 Biological Attachment Assays .......................................................................... 92 Ulva ............................................................................................................ 93 Navicula incerta ......................................................................................... 93 C. marina ................................................................................................... 94 Results .................................................................................................................... 95 Microtopography Characterization .................................................................... 95 Chemic al Composition ...................................................................................... 96 Surface Energy Measurements ........................................................................ 96 Biological Attachment Assays .......................................................................... 97 Ulva attachment ......................................................................................... 97 Attachment of cells of Navicula .................................................................. 98 Attachment of cells of C. marina ................................................................ 99 Discussion ............................................................................................................ 100 Hydrogel Characterization .............................................................................. 100 Biological Attachment Assays ........................................................................ 100 6 ENGINEERED MICROTOPOGRAPHIES AND CHEMISTRIES INFLUENCE ATTACHMENT AND FUNCTION OF ENDOTHELIAL AND SMOOTH MUSCLE CELLS .................................................................................................................. 106 Background ........................................................................................................... 106 Biology of the Vascular Wall ........................................................................... 109 Topography .................................................................................................... 110 Approach ........................................................................................................ 112 Materials and Methods .......................................................................................... 114 Materials ......................................................................................................... 114 Sample Preparation ........................................................................................ 115 Attaching Hydrogels to Glass Slides .............................................................. 116 Grafting Fibronectin to Hydrogel Surfaces ...................................................... 117 Hydrogel Characterization .............................................................................. 117 Cell Culture Assays ........................................................................................ 122 Sample preparation .................................................................................. 122 Porcine vascular endothelial cell culture .................................................. 123 Human cell culture ................................................................................... 123 Image analysis ......................................................................................... 124 Statistical methods ................................................................................... 124 Results and Discussion ......................................................................................... 125 Hydrogel Characteriz ation Results ................................................................. 125 Solubility parameter ................................................................................. 125 Average molecular weight between crosslinks and modulus ................... 126 Hydrogel composition .............................................................................. 127 Fibronectin grafting .................................................................................. 128 Cell Culture Assay Results ............................................................................. 130 PVEC assay ............................................................................................. 130 Human cell assays ................................................................................... 132 Discussion ...................................................................................................... 146

PAGE 8

8 7 CONCLUSIONS AND FUTURE WORK ............................................................... 150 Conclusions .......................................................................................................... 150 Non Toxic Antifouling Strategies .................................................................... 150 Cell Culture Substrates .................................................................................. 151 Future Work .......................................................................................................... 152 Non Toxic Antifouli ng Strategies .................................................................... 152 Cell Culture Substrates .................................................................................. 153 APPENDIX A REYNOLDS NUMBER CALCULATIONS ............................................................. 156 B SUMMARY OF RECENT SMALL DIAMETER VASCULAR GRAFT LITERATURE ....................................................................................................... 158 LIST OF REFERENCES ............................................................................................. 166 BIOGRAPHICAL SKETCH .......................................................................................... 182

PAGE 9

9 LIST OF TABLES Table page 3 1 Feature geometry and engineered roughness index calculations for topographies exposed to the chemostat culture ................................................. 59 3 2 Feature geometry and engineered roughness index calculations for topographies exposed to the overnight culture ................................................... 59 5 1 Average heights plus or minus standard deviation of hydrogel and PDMSe replicates of wafers measured with white lig ht optical profilometry and SEM ..... 95 5 2 Contact angle measurements and calculated surfac e energies for smooth hydrogels ........................................................................................................... 96 5 3 Contact angle measurements and calculated surface energies ( s) for smooth PDMSe .................................................................................................. 97 6 1 Properties of solvents used to calculate solubility parameters adapted from (Brandup et al. 1999) ........................................................................................ 119 6 2 Hydrogel characterization results: average molecular weight between crosslinks and modulus values ......................................................................... 127

PAGE 10

10 LIST OF FIGURES Figure page 2 1 Sche matic of v arious wetting models ................................................................. 23 2 2 Schematic of various surface topographies and the associated dimensions ...... 24 2 3 Macrofouling on the hull of a ship increases drag and fuel consumption. Image courtesy of North Florida Shipyards, Jacksonville FL .............................. 26 2 4 Schematic demonstrating the hierarchy of fouling organisms ............................ 28 2 5 Schematic of the dynamic biofouling process which takes pl ace over numerous length scales ...................................................................................... 29 2 6 Scanning electron micrographs of natural textured surfaces .............................. 32 2 7 The Baier curve demonstrates the relative amount of biofouling versus critical surface tension of the substrate .......................................................................... 34 2 8 White light optical profilometry image of Sharklet AF ...................................... 39 2 9 Correlation of Ulva spore settlement density and Engineered Roughness Index (ERI) ......................................................................................................... 41 3 1 Scan ning electron micrographs of (a) pillars, (b) ridges, (c) triangle/pillars, (d) Sharklet AF, (e) Recessed Sharklet AFTM surfaces in PDMSe ....................... 51 3 2 C. marina attachment vs time ............................................................................. 56 3 3 C. marina attachment data on PDMSe surfaces represented as mean cell density (cells mm2) +9 5% confidence interval (n = 30) ...................................... 58 3 4 The attachment model shows the correlation between C. marina attachment and ERIII at 120 min for the stationar y and logarithmic growth phases .............. 60 3 5 The attachment model shows the correlation between attachm ent of zoospores of Ulva and cells of C. marina and ERIII*Re ...................................... 65 4 1 Topographies created in PDMSe A) Sharklet AF (+2.8SK2x2) and B) Re cessed Sharklet AF ( 3.1SK2x2) ................................................................ 72 4 2 Initial attachment and attachment after exposure to shear stress (26 Pa) of Seminavis ........................................................................................................... 75 4 3 Seminavis and Navicula on +2.8SK2x2 replicated in PDMSe ............................ 77

PAGE 11

11 4 4 Density of attached diatoms of Seminavis after 6 d, before and after expo sure to a shear stress of 26 Pa ................................................................... 78 4 5 Density of attac hed diatoms of Navicula after 6 d, before and after exposure to a shear stress of 45 Pa ................................................................................... 79 4 6 Attachment density of Navicula after exposure to shear st ress (45 Pa) on each topography ................................................................................................. 80 4 7 Navicula attachment after gentle washing on A) +2.8SK2x2, B)+2.8SK2x5 and C) +2.8SK10x2 t opographies replicated in PDMSe ..................................... 81 4 8 The nor malized, transformed attachment density of Navicula after exposure to a shear stress of 45 Pa versus the maximum area of contact ........................ 82 5 1 Chemical structures of monomers used to produce functional ized hydrogels .... 89 5 2 Schematic of mold for hydrogel production ........................................................ 90 5 3 Schematic of captive air and oil bubble measurements for calc ulating surface energy ................................................................................................................ 92 5 4 Scanning electron micrographs of A)+0.6CH2x2 and B)+1SK2x2 topographies in PEGDMA co GMA hydrogel ...................................................... 95 5 5 White light optical profilometry image of +1SK2x2 t opography cast in PEGDMAco GMA .............................................................................................. 95 5 6 Functionalized poly(ethylene glycol) based hydrogels reduce the attachment density of zoospores of Ulva ............................................................................. 97 5 7 Functionalized poly(ethylene glycol) based hydrogels reduce attachment density and attachment strength of Navicula ...................................................... 98 5 8 Functionalized poly(ethylene glycol) based hydrogels reduce attachment density and attachment strength of C. marina .................................................... 99 5 9 Normalized, transformed Ulva attachment density on PDMSe and PEGDMA co HEMA topographies plotted versus ERIII* Re* 102. ....................................... 103 5 10 Normalized, transformed spore attachment density on PDMSe and PEGDMAco HEMA topographies plotted versus ERIII* Re* 0 102 ................ 104 5 11 Normalized, transformed attachment densities of spores of Ulva and cells of C. marina on PDMSe and PEGDMA co HEMA versus ERIII* 0 102 ............. 104 6 1 Chemical structures of monomers used to produce functionalized hydrogels .. 115

PAGE 12

12 6 2 Sharklet surfaces varying in the distinct number of features (n) and Channels replicated in PDMSe ......................................................................................... 122 6 3 Plot of volumetric swelling ratio versus solvent solubility parameter for PEGDMA, PEGDMAco GMA and PEGDMAco HEMA h ydrogels .................. 126 6 4 Spectral subtraction of ATR FTIR spectra of PEGDMA co GMA minus PEGDMA .......................................................................................................... 127 6 5 Spectral subtraction of ATR FTIR spectra of PEGDMA co HEMA minus PEGDMA .......................................................................................................... 128 6 6 Spectral subtraction of ATR FTIR spectra of PEGDMA graft Fn minus PEGDMA and PEGDMAco G MA graft Fn minus PEGDMAco GMA ............. 128 6 7 Epifluorescent micrographs (magnific ation 500x) of immunofluorescently labeled Fn ......................................................................................................... 129 6 8 Fluorescence intensity of immunofluorescently labeled Fn .............................. 130 6 9 PVEC cell culture assay results: average number of cells mm2 on smooth surfaces ........................................................................................................... 131 6 10 PVEC cell culture assay results: average number of cells mm2 on smooth and topographically modified surfaces ............................................................. 131 6 11 Arrows indicate the direction of the channels in each topography on phase contrast micrographs ........................................................................................ 132 6 12 HCAECs seeded at 5x104 cells/well on PDMSe after 24 h ............................... 133 6 13 HCAECs seeded at 5x104 cells/well on PEGDMAco GMA graft Fn after 24 h. ................................................................................................................. 134 6 14 HCAECs seed ed at 2.5x104 cells/well on PDMSe after 24 h ............................ 135 6 15 HCAECs seeded at 2.5x104 cells/well on PDMSe after 7 d .............................. 136 6 16 HASMCs seeded at 2.5x104 cells/well on PDMSe after 24 h ........................... 137 6 17 HASMCs seeded at 2.5x104 cells/well on PDMSe after 7 d ............................. 138 6 18 HASMCs seeded at 2.5x104 cells/well on PEGDMAco GMA graft Fn after 24 h .................................................................................................................. 139 6 19 Average CSI for HCAECs cultured on topographi es replicated in PDMSe A) after 24 h and B) after 7 d ................................................................................. 140

PAGE 13

13 6 20 Average CSI for HC A SMCs cultured on topographies replicated in PDMSe A) after 24 h and B) after 7 d ................................................................................. 140 6 21 Distribution of the percentage of HCAECs at angles in each 10 degree range after 24 h .......................................................................................................... 141 6 22 Distribution of the percentage of HCAECs at angles in each 10 degree range after 7 d ............................................................................................................ 142 6 23 Distribution of the percentage of HCASMCs at angles in each 10 degree range after 24 h. ............................................................................................... 143 6 24 Distribution of the percentage of HCASMCs at angles in each 10 degree range after 7 d .................................................................................................. 144 6 25 Average angle of cell orientation relative to the channels in the topography is plotted for A) HCAEC after 24 h in culture and B) HCAEC after 7 d in culture 145 6 26 Average angle of cell orientation relative to the channels in the topography is plotted for A) HCASMC after 24 h in culture and B) HCASMC after 7 d ........... 145

PAGE 14

14 LIST OF ABBREVIATIONS describes the risk of making a Type I error in statistical analysis ANOVA Analysis of Variance ATRFTIR Attenuated Total Reflectance Fourier Transform Infrared Spectrometry C. marina Cobetia marina CSI Cell Shape Index EC Endothelial Cell ERI Engineered Roughness Index Fn Fibronectin GMA Glycidyl methacrylate HCAEC Human Coronary Artery Endothelial Cell HCASMC Human Coronary Artery Smooth Muscle Cell HEMA 2 Hydroxyethylmethacrylate Navicula Navicula incerta p p value or observed significance level PDMSe P olydimethylsiloxane elastomer Silastic T2 Dow Corning Corporation PEG Poly(ethylene glycol) PEGDMA Poly(ethylene glycol) dimethacrylate PVEC Porcine Vascular Endothelial Cell SMC Smooth Muscle Cell Re Reynolds Number Seminavis Seminavis robusta TCPS Tissu e culture polystyrene Ulva Ulva linza

PAGE 15

15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ENGINEERED MICROTOPOGRAPHIES AND SURFACE CHEMISTRIES DIRECT CELL ATTACHMENT AND FUNCTION By Chelsea Marie Magin December 2010 Chair: Anthony B. Brennan Major: Biomedical Engineering Harrison, in 1914, first recognized that cells respond to physicochemical cues such as substratum topography when he observed that fibroblasts elongated while cultured on spider silk. Recently, techniques developed in the microelectronics industry have been used to c reate molds for producing microscaled topographies with various shapes and spatial arrangements. Although these patterning techniques are well established, very little is known about the mechanisms underlying cel l sensing and response to micro topographies. In this work cellular micro environments with varying surface topographies and chemis tries were evaluated with marine organisms and mammalian cells to investigate cellular sensing and response. Biofouling the accumulation of microorganisms, plants, and animals on submerged surfaces is an environmental and economic concern. E ngineered topographies, replicated in polydimethyl siloxane elastomer (PDMSe) and functionalized poly(ethylene glycol) dimethacrylate (PEGDMA) hydrogels, were evaluated for inhibition of marine fouling organism attachment Microtopographies replicated in PDMSe in hibited attachment of the marine bacterium, Cobetia marina up to 99% versus smooth.

PAGE 16

16 The average normalized attachment densities of cells of C. marina and zoospores of the green algae Ulva on PDMSe topographies scaled inversely with the Engineered Roughnes s Index (ERIII), a representation of surface energy. Attachment densities of Ulva from four assays and C. marina from two growth phases to PDMSe surfaces scaled inversely with one equation: ERIII multiplied by the Reynolds number of the organism (Re) (R2 = 0.77). The same microtopographies created in PDMSe reduced the initial attachment density and attachment strength of cells of the diatoms Navicula incerta and Seminavis robusta compared to smooth PDMSe. The average normalized attachment density of Na vicula after exposure to shear stress (48 Pa) was correlated with the contact area between the diatom and a topographically modified surface (R2=0.82). Functionalized PEGDMA hydrogels significantly reduced attachment and attachment strength of Navicula a nd C. marina. These hydrogels also reduced attachment of zoospores of Ulva compared to PDMSe. Attachment of Ulva to micro topographies in PDMSe and PEGDMA co HEMA negatively correlated with ERIII*Re (R2 = 0.94 and R2 = 0.99, respectively). Incorporating a surface energy term into this equation created a correlat ion between the attachment densities of cells from two evolutionarily diverse groups on substrates of two surface chemistries with an equation that describes the various microtopographies and surf ace chemistries in terms of surface energy (R2 = 0.80). The current Attachment Model can now be used to design engineered antifouling surface microtopographies and chemistries that inhibit the attachment of organisms from three evoluntionarily diverse groups.

PAGE 17

17 H ydrogels based on PEGDMA were also chosen as a substratum material for mammalian cell culture. Capturing endothelial progenitor cells (EPCs) and inducing differentiation into the endothelial cell (EC) phenotype is the ideal way to reendothelialize a small diameter vascular graft. Substratum elasticity has been reported to direct stem cell differentiation into specific lineages. Functionalized PEGDMA hydrogels provided good compliance, high fidelity of topographic features and sites for surface mo dification with biomolecules. Fibronectin grafting and topography both increased EC attachment. This combination of adjustable elasticity, surface chemistry and topography has the potential to promote the capture and differentiation of EPCs into a confluent EC monolayer. Engineered microtopographies replicated in PDMSe directed elongation and alignment of human coronary artery endothelial cells (HCAECs) and human coronary artery smooth muscle cells (H C ASMCs) compared to smooth surfaces. Engineered c ellu lar micro environments were created with specific surface energies defined by chemistry and topography to successfully direct cell attachment and function.

PAGE 18

18 CHAPTER 1 INTRODUCTION Scope of Research It is widely recognized that cells and organisms respond to physicochemical cues presented by their microenvironment (Harrison 1914, Callow et al. 2002, Dalby 2005, Engler et al. 2006, Schumacher et al. 2007b, Schilp et al. 2009) In this work the influence of engineered microtopographies and surface chemistries on cell attachment and function is investigated for two different applications : marine antifouling and substrates for mammalian cell culture. Combinations of surface topography and chemistry were evaluated f or the deterrence of fouling cells and organisms in the marine environment and as substrates for mammalian cell culture. Specific Aims Specific Aim 1: Determine the Contribution of Organism Size and Shape to the Slope of the Attachment Model using Cobet ia marina A model that relates the normalized, transformed attachment density of the green algae Ulva linza to engineered microtopographies was previously developed. The attachment densities from four separate Ulva attachment assays performed over three years correlated with the ERIII (R2 = 0.88) (Long et al. 2010) It was hypothesized that the normalized, transformed attachment density of cells of C. marina would also have a negative correlation with ERIII ( =0.05). Specific Aim 2: Predict Attachment Densities and Attachment Strength of Cells of the Diatoms Navicula incerta and Seminavis robusta to Engineered Microtopographies The attachment model was hypothesized to predict the initial attachment densities of cells of the diatoms Navicula and Seminavis to engineered microtopographies (Long,

PAGE 19

19 et al. 2010) The normalized, transformed attachment densities were hypothesized to have a negative correlation with ERIII ( ). Substrates for antifouling applications were also exposed to shear stress to test foulingrelease after initial attachment. Attachment Point Theory states that the retention of cells by surface microtopographies provides protection from shear s tresses in the surrounding environment. O rganisms that are provided with larger numbers of points of contact to the surface will therefore have increased attachment strength (Verran and Boyd 2001, Scardino et al. 2006) The attachment strength of cells of both diatoms to engineered microtopographies was hypothesized to correlate with the number of attachment points available to the cell on each surface. The attachment densities of cells after exposure to shear stress were hypothesize d to have a negative correlation with the number of attachment points ( ). Specific Aim 3: Determine the Contribution of Surface Chemistry to the Slope of the Attachment Model using Functionalized Poly(ethylene glycol) Dimethacrylate Hydrogels Surfaces coated with poly(ethylene glycol) (PEG) and its oligomers exhibit resistance to protein adsorption and biofouling (Ostuni et al. 2001, Balamurugan et al. 2005, Ekblad et al. 2008, Schilp, et al. 2009) It was hypothesized that initial attachment of Ulva, Navicula and C. marina would be lower on functionalized PEG dimethacrylate (PEGDMA) hydrogels than on PDMSe and that cell and spore densities remaining on hydrogels after exposure to shear stress would be lower than those on PDMSe. It was also hypothesized that t he negative correlation of the normalized, transformed attachment density of spores of Ulva on functionalized hydrogels with the attachment model (Long, et al. 2010) ( ) would have a slope and interc ept lower than those for the correlation with attachment densities on PDMSe.

PAGE 20

20 Specific Aim 4 : Identify a Combination of Surface Chemistry and Topography that E nhances the Attachment of Human Coronary Artery Endothelial Cells or Human Coronary Artery Smooth Muscle Cells The clinical application of synthetic small diameter (d<6mm) vascular grafts has been limited due to high rates of occlusion from thrombosis and intimal hyperplasia. Intimal hyperplasia can be caused by compliance mismatch between the graft and the vessel wall and poor reendothelialization of the luminal surface. C apturing endothelial progenitor cells (EPCs) and inducing their differentiation into the endothelial cell (EC) phenotype could be the ideal way to re endothelialize a small diamet er vascular graft (Asahara et al. 1997) S ubstratum elasticity can direct stem cell differe ntiation into specific lineages (Engler, et al. 2006) Therefore, P EGDMA hydrogels which have a highly adjustable shear modulus (Pfister et al. 2007) (G=10k Pa to 1 Mpa) were chosen as a substratum material. This combination of adjustable elasticity, surface chemistry and topography has the potential to promote the capture and differentiati on of EPCs into a confluent EC monolayer that is nonthrombogenic and stable to shear stress Such capture and differentiation would make small diameter artificial vascular grafts feasible. It was hypothesized that PEGDMA co glycidyl methacrylate (GMA) g raft fibronectin (Fn) hydrogels would increase EC attachment 2fold versus PEGDMA hydrogels (Markway et al. 2008) The effects of t opography, i.e., cell elongation or cell phenotype were hypothesized to increase with increasing ERIII value (Schumacher, et al. 2007b, Long, et al. 2010) ( ).

PAGE 21

21 CHAPTER 2 BACKGROUND The chemistry and physical properties of a substratum are critical in directing cell attachment and function. Control of bioadhesion is vital in both the marine environment and in the creation of biom aterials for cell culture substrates. Several reviews have been written describing the effects of surface chemistry (Marmur 2006) topography (Verran and Boyd 2001) and combinations of the two (Abarzua and Jakubowski 1995, Yebra et al. 2004, Genzer and Efimenko 2006, Ralston and Swain 2009, Magin et al. 2010a) on the attachment of marine fouling cells and organisms. Likewise, biomaterials are designed with a focus on the influence of chemistry (Hoffman 2002, Stegemann et al. 2007, de Mel et al. 2008) and topography (Dalby et al. 2002, Lim and Donahue 2007, Moon et al. 2009, Liliensiek et al. 2010) on cell attachment and function. The energy required to wet (or dewet) a surface c ontrol s bioadhesion. Topographical modification of a surface is a parameter that controls the wetting (or dewetting) of a surface (Bico et al. 1999, Bico et al. 2002, Quere 2008) The contact angle of a liquid dr op on a flat, homogeneous solid is given by Youngs equation: = (2 1) The different surface tensions, solid/vapor, solid/liquid and liquid/vapor are given by SV, SL and respectively. Wettability of a topographically modified surface can be described by two different models. Wenzel proposed the first model in 1936 (Wenzel 1936) = (2 2)

PAGE 22

22 where 1 and r is the solid roughness The solid roughness (r) is defined as the ratio between the actual surface area and the projected surface area (Bico, et al. 1999) A drop will spread on a rough surface until it reaches *, which is not the same as the Youngs angle. Wenzels model assumes that the topography becomes fully wetted and that the change in contact angle is due to the increased wetted surface area. Wenzels Equation (22) shows that the effect of surface topography is to amplify wetting, since r > 1 in all cases i.e., a hydrophilic surface is more easily wetted and hydrophobic surface is less easily wetted. Cassie and Baxter proposed a second model for the wetting of rough surfaces while investigating waxy surfaces which were not only rough, but also porous (Cassie and Baxter 1944) Under these conditions water drops did not simply follow the contours of the topography as descri bed by Wenzel. I nstead, two wetting regimes were described: air entrapment and wicking. When air was entrapped within the features of the topography, drops rested on a composite of the surface and entrapped air In the case of wicking, the liquid was drawn into the topography at the advancing edge so that the drop rested on a combination of the surface and liquid. The CassieBaxter wetting model is described by the following equations: Air entrapment: = 1 + ( + 1 ) (2 3) Wickin g: = 1 + ( ) (2 4)

PAGE 23

23 Figure 21. Schematic of various wetting models: A) Youngs Model, B) Wenzels Model, C) Cassie Baxter Model Air Entrapment and D) CassieBaxter Model Wicking. Topographically modified surfaces have also been tested in the biomedical field. The adhesion and spreading of mammalian cells (Hatcher et al. 2002, Carman et al. 2006) was changed by topographical modification. The attachment and biofilm formation of the of the bacteria Staphyloccocus aureus (Chung et al. 2007b) has been also been inhibited using topography Hatch er and Seegert (Hatcher, et al. 2002) demonstrated that scaffolds of various porosities made from polyvinylpyrroli done modified bioactive glass fibers could increase proliferation of rat mesenchymal stem cells preceding differentiation. More recently, a pattern of 3 m diameter circles of the ECM protein fibronectin on PDMSe was used to direct formation of focal adhesions and grow an EC monolayer with density and morphology similar to that of the native artery (Feinberg et al. 2009) Additionally, topographical cues were used to induce EC morphologies in vitro that were stable to the shear stresses that would be experienced in vivo (Carman 2007) Several of the

PAGE 24

24 topographies investigated in this work are shown in the schematic below (Figure 22) Surfaces are designated by a nomenclature that includes feature orientation, feature height, surface design, feature width, feature spacing and the number of distinct features. All feature dimensions are reported in m. The Sharklet AF surface in Figure 22 Part A would be designated as +3SK2x2_n4. The channels surface in Figure 2 2 Part D would be designated +3CH2x2. Figure 22. Schematic of various surface topographies and the associated dimensions. A) Sharklet AF (+3SK2x2), B) Triangle/Pillars (+3T10P2x2), C) Pillars (+3P2x2) and D) Channels (+3SK2x2) Adapted from an image created by Christopher J. Long. In the current work, a model that relates surface energy and topography to cell response is used to control bi oadhesion and cell function in both marine antifouling and biomedical applications

PAGE 25

25 Non toxic Antifouling Strategies1 Introduction Biofou ling of marine vessels continues to plague sailors as it has for thousands of years (Wahl 1989, Callow and Callow 2002a, Yebra et al. 2004b, Chambers et al. 2006, Genzer and Efimenko 2006b, Grozea and Walker 2009a, Ralston and Swain 2009). The ancient Phoenicians, inventors of the earliest recorded anti fouling coatings, covered ships with lead sheets (Yebra, et al. 2004b). Later in the 17th century metals containing copper were also shown to be effective biofouling deterrents. Metals, such as lead, are effective antifouling agents, but have a negative impact on the environment. Ships are still slowed today by the growth of algae, barnacles, and slime on their hulls due to the absence of a universal, green, antifouling system (Figure 2 1). The United S tates (US) Naval Sea Systems Command estimates that biofouling on ship hulls results in a speed loss of approximately 2 percent and increases fuel costs 6 to 45 percent depending on the size of the ship (Ingle 2008). One source cites total costs associated with biofouling of nearly $1 billion annually (Callow and Callow 2002b). Antifouling, in this review, refers to all systems that prevent an organism from attaching to a surface. Historically, the term antifouling was associated only with biocidal compounds. Current antifouling strategies focus on green, nontoxic technologies. Foulingrelease describes the force required to remove an organism that is already attached to a surface. These two terms have been used interchangeably in the literature; however they are truly different phenomena. 1 Reprinted with permission from Elsevier from Magin CM, Cooper SP and Brennan AB. 2010a. Non Toxic Antifouling Strategie s. Ma terials Today. 13(4):36 44.

PAGE 26

26 Figure 23 Macrofouling on the hull of a ship increases drag and fuel consumption. Image courtesy of North Florida Shipyards, Jacksonville FL. [Reprinted with permission from Magin CM, Cooper SP and Brennan AB 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 37, Figure 1)] Antifouling paints have been and remain the primary strategy for combating biofouling in the marine industry. Biocides such as tributyltin (TBT) were developed in the middle of the 20th century and were the active components of antifouling paints until recently (Yebra, et al. 2004b). Biocidal paints based on TBT have been effective at reducing biofouling (Yebra et al. 2004a, Chambers, et al. 2006, Howell and Behrends 2006). However, the use of TBTbased paints has been prohibited because they are detrimental to nontarget organisms and the surrounding environment (Sonak et al. 2008). The response to this ban has been the use of copper, zinc, and a variety of or ganic compounds as the active, antifouling components. The ideal replacement for TBT is an environmentally neutral coating with both antifouling and foulingrelease

PAGE 27

27 properties (Goupil et al. 1973, Yebra, et al. 2004a, Genzer and Efimenko 2006a, Marmur 2006). Biofouling is a major challenge for the biomedical industry as well. Healthcare associatedinfections are attributed to biofilms on surfaces such as countertops, doors, beds, surgical tools, or medical devices such as catheters. The Centers for Disease Control and Prevention have reported that these healthcareassociated infections account for an estimated 1.7 million infections and 99,000 deaths annually in the US (Klevens et al. 2002). Furthermore, these infections accounted for nearly $45 billion of patient costs in 2007 (Scott 2009). The formation of an atherosclerotic plaque within the arterial wall can be broadly described as a biofouling process (Libby and Theroux 2005). The American Heart Association reported that 16.8 million people in the U S were diagnosed with coronary heart disease in 2006. Coronary heart disease is the leading cause of death in the US. The estimated direct and indirect costs of treating this disease total approximately $165.4 billion per year (He art Disease & Stroke Statistics 2010). Biofouling is a very dynamic process, which spans numerous length and time scales (Figures 2 2 and 2 3). Fouling of a new surface is typically described as a four phase process: formation of a conditioning layer of organic molecules, primary colonization by microorganisms such as bacteria and diatoms, unicellular colonization by algal spores, and attachment of multicellular macrofoulers (Wahl 1989, Abarzua and Jakubowski 1995, Chambers, et al. 2006).

PAGE 28

28 Figure 24 Schematic demonstrating the hierarchy of fouling organisms. Cells and compounds relevant to biomedical applications are shown above the scale axis. Marine organisms are shown below the scale. [Reprinted with permission from Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 37, Figure 2)] Since fouling occurs in an aqueous solution, the properties of the fluid mediate the interaction of the fouling organism and material. Ions and water molecules adsorb to a biomateri al surface to form an electric double layer immediately upon immersion. This electric double layer effectively establishes the charge associated with surface. This electrostatic charge affects the nature of the interaction of proteins and cells with the surface. Antifouling performance scales with both density and sign of the charge (Ostuni et al. 2001a).

PAGE 29

29 Figure 25 Schematic of the dynamic biofouling process which takes place over numerous length scales. (a) An electric double layer is establishe d at the surface of a solid such as a linear polymer in less than a second. This electric double layer mediates the adsorption and conformation of proteins. (b) The typeII subunits of fibronectin are shown adsorbed to the surface. These subunits are res ponsible for binding to gelatin (Pickford et al. 2001) (c) Fibronectin mediates the binding of a cell to the surface via integrins The type II domains of fibronectin are shown in yellow. (d) If a bact erial cell is bound to the surface, it undergoes a phenotypic change and excretes an EPS coating. (e) Over time, the cells replicate and continue to build the EPS. The biofilm creates swarmer cells, which leave the biofilm to inoculate another surface. Larger cells such as Ulva (in the marine environment) or phagocytes (in the human body) may subsequently interact with the initial biofilm. Figure created by coauthor Scott P. Cooper. [Reprinted with permission from Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 38, Figure 3)]

PAGE 30

30 A layer of proteins adsorbs to a pristine surface within seconds to minutes following immersion (Vroman et al. 1982). The protein conformation is strongly influenc ed by both the physical and chemical structure, including electrostatic charge, of the surface. Protein conformation defines functionality with respect to cell adhesion (Williams et al. 1982, Bergkvist et al. 2003). This protein layer acts as a conditioning film for the settlement of micro organism s such as diatoms and bacteria. A biofilm can be defined as a community of attached microorganisms connected by an extracellular polysaccharide (EPS) coating. Bacteria undergo multiple developmental stages from planktonic to attached cells. This transformation from the planktonic to attached state induces a phenotypic change that facilitates increased secretion of an EPS coating (Stewart and Franklin 2008). The EPS coating is both an adhesive and protective l ayer that modulates the diffusion of molecules in the biofilm. Consequently, cells in biofilms are more resistant to antibiotics and antibacterial agents (Costerton et al. 1987). Natural biofilms are composed of several microbial species and their EPS coa tings. These cells along with protein and enzyme structures form complex, functional microcolonies. It was first observed by Zobell and Allen in 1935 that biofilms could stimulate the settlement of secondary macro organisms (Zobell and Allen 1935) such as algal spores (Joint et al. 2002, Patel et al. 2003, Marshall et al. 2006) and larvae of barnacles and tubeworms (Hadfield and Paul 2001, Zardus et al. 2008). Reviews on the subject indicate that marine biofilms can also inhibit or have no effect at all on settlement of macroorganisms (Qian et al. 2007, Dobretsov 2008). The interaction between a marine biofilm and secondary colonizers is a complex interplay of surface chemistry, micro topography, and microbial products i.e., low molecular weight

PAGE 31

31 metabolites involved in quorum sensing (Qian, et al. 2007).The diversity of species resulting from various geographic locations creates a broad spectrum of physical, chemical and biological attributes. We have investigated natural structures that are able to re sist the adhesion of these complex fouling communities. This review discusses natural surfaces as well as physicochemical and physical antifouling strategies. Natural Antifouling Surfaces There are natural surfaces that resist biofouling in the marine and the biomedical environments. These natural antifouling surfaces appear to use a combination of chemical and physical structures to inhibit biofouling. Marine organisms such as sharks, mussels, and crabs have natural antifouling defenses. The endothelium of a healthy artery is another example of a natural antifouling system (Figure 2 4). However, it is also recognized that these surfaces will lose their antifouling characteristics due to age, injury or disease. The skin of the approximately 900 spec ies of Elasmobranchii, which include sharks, skates, and rays is embedded with placoid scales (Bone and Moore 1995). These scales have a vascular core of dentine surrounded by an acellular enamel layer similar to human teeth. For this reason, placoid sc ales are commonly referred to as dermal denticles. Denticles serve several functions including reduction of mechanical abrasion, reduced hydrodynamic drag (Lang et al. 2008) and most interestingly protection from ectoparasites (Raschi and Tabit 1992). The skin of two members of the porpoise family, i.e., the bottlenose dolphin Tursiops truncatus and the killer whale Orcinus orca, forms a system of ridges and grooves oriented transversel y to the direction of flow. The natural wavelength of the ridges and grooves is 0.3 to 0.4 mm with a trough to crest wave height of about 10 m (Gucinski et al. 1984). These

PAGE 32

32 topographic features and a mucosal coating secreted by epidermal cells contribute to the antifouling properties of these marine animals. The microtopographically structured periostraca on shells of the blue mussels Mytilus galloprovincialis (Scardino et al. 2002) and Mytilus edulis (Bers and Wahl 2004) are also effective antifouling surfaces. The grooves and ridges of the periostraca are 1 to 2m wide with an average depth of 1.5 m. The shells of M. galloprovincialis significantly reduced settlement of barnacle larvae during a 14 week field exposure trial (Scardino, et al. 2002). Microtopography replicates cast in epoxy resin from the blue mussel M. edulis edible crabs Cancer pagurus the eggcase of the lesser spotted dogfish Scyliorhinus canicula, and the brittle star Ophiura texturata reduced fouling for three to four weeks (Bers and Wahl 2004). The short term performance implies that natural ant ifouling is a combination of chemistry and microtopography. Figure 2 6 Scanning electron micrographs of natural textured surfaces: a) Spinner shark skin, b) Galapagos shark skin, c) Mussel shell ( M. edulis ) and d) Crab shell ( C. pagurus ) reprinted fro m (Bers and Wahl 2004) with permission from the publisher Taylor & Francis Group (http://www.informaworld.com), e) Porcine pulmonary artery reprinted from (Feinberg, et al. 2009) with permission from Elsevier. [Reprinted with permission from Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 39, Figu re 4)]

PAGE 33

33 The inner surface of a blood vessel is another natural surface that resists the constant presence of fouling proteins and cells. The endothelium consists of a continuous monolayer of endothelial cells with a cobblestonelike morphology and a distinc t topography (Figure 2 4). Endothelial cells express a negatively charged glycoprotein coat that repels platelets and leukocytes. These cells also secrete bioactive substances that inhibit thrombosis and smooth muscle cell proliferation (Biology of the Ar terial Wall 1999, Xue and Greisler 2000). This combination of chemistry and microtopography creates an ideal anti thrombogenic, i.e., antifouling, surface. Physicochemical Antifouling Strategies Surface chemistry is a significant factor in the formation, stability and release of adhesion by fouling organisms to surfaces. The work by Baier in the late 1960s demonstrated a correlation between relative adhesion of fouling organisms and the energy of the surfac e (Baier 2006). The Baier curve (Figure 25) as this relationship is known, has been confirmed in several marine and biomedical environments (Goupil, et al. 1973, Baier 2006). A key characteristic of the Baier curve is that minimal fouling is typically achieved at a critical surface tension of 2224 mN/m. This surface tension, often referred to as surface energy, is approximately equal to the dispersive component for water. In an aqueous system, water must rewet the system when proteins and cells are removed. For solids with a surface energy of ~22 mN/m, the thermodynamic cost for water to re wet the surface is minimized. One way of systematically varying surface energy without altering the bulk material is through self assembled monolayer (SAMs). In an extensive study, Whitesides and co workers tested the ability of a wide range of SAM chemistries to resist protein adsorption.

PAGE 34

34 Figure 27 The Baier curve demonstrates the relative amount of biofouling versus critical surface tension of the substrate. Reprinted from (Baier 2006) with kind permission from Springer Science + Business Media. [Reprinted with permission from Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 40, Figure 5)] The authors c onclude that SAMs, which are hydrophilic, electrically neutral, and contain hydrogen bond acceptors, are most effective at resisting protein adhesion. Zwitterionic structures have both positive and negative domains, but remain electrically neutral overall It has been demonstrated that zwitterionic compounds similar to phosphorylcholine such as sulfobetaine resisted protein adsorption when the surface density and chain length of the SAMS were carefully controlled (Ostuni et al. 2001b, Chang et al. 2008). Even though surface energies for poly(ethylene glycol)(PEG) and its oligomers typically fall above the zone of low cell adhesion defined by Baier, it is widely recognized that these materials exhibit resistance to protein adsorption and biofouling (Ostuni, et al. 2001b, Balamurugan et al. 2005, Schilp et al. 2007, Ekblad et al. 2008). The mechanism for protein resistance for high molecular weight PEG is well explained

PAGE 35

35 by steric repulsion (Jeon et al. 1991). Andrade and de Gennes postulated that during protein adsorption water must be removed from the PEG structure. This dehydration is thermodynamically unfavorable because it leads to confinement of polymer chains which previously had high conformational entropy. Even though the model system of oligo(ethyle ne glycol) SAMs tested by Whitesides restricted conformational freedom of end groups into densely packed films, these surfaces also showed protein repellent properties. Grunze and others have proposed that the chain conformation and packing of SAMs affect the penetration of water into the SAM surface and are also important determinants of resistance to protein adsorption (Herrwerth et al. 2003, Balamurugan, et al. 2005). The surface chemistry of SAMs is strongly influenced by their physical structure. Eth yleneglycol terminated SAMs have been shown to be especially foulingresistant in numerous studies. Ulva zoospore attachment to SAMs systematically increased with decreasing wettability and correlated with adsorption of the protein fibrinogen (Schilp, et al. 2007). Experiments have shown that higher numbers of Ulva spores attach to hydrophobic SAMs versus hydrophilic ones in static assays (Callow et al. 2000). However, the attachment strength of Ulva spores is greater on hydrophilic SAMs (Finlay et al. 2002). The mechanism for delay of Ulva attachment by PEG based surfaces is not fully understood. However, like resistance to protein adsorption, infiltration of water into the SAM surfaces may create a hydration energy that prevents effective interaction o f the adhesive used by Ulva with the surface (Schilp, et al. 2007). Bowen, et al. tested the effect of SAM chain length on the settlement and release of zoospores of Ulva and cells of the diatom Navicula incerta. This study showed that

PAGE 36

36 chain length affected release more than settlement. Alkane chains greater than 12 carbons long corresponded to higher release of these organisms under flow. This foulingrelease behavior is associated with greater rigidity of the alkane chain and subsequently higher lubr icity (Bowen et al. 2007). Attachment of a medically relevant bacterium ( Staphylococcus epidermidis ) and a marine bacterium ( Cobetia marina) was reduced up to 99.7% by surfaces coated with hexa(ethylene glycol) terminated SAMs(Ista et al. 1996). The response of C. marina to surface energy was opposite of that predicted by the Baier curve, i.e., attachment density increased with decreasing surface energy. Attachment of Ulva showed the same relationship only when the cosine of the advancing water contact angle was (Ista et al. 2004). Hydrogels crosslinked polymer networks that swell in the presence of water have also been investigated for antifouling applications. Rasmussen et al. demonstrated that hydrogel surfaces of alginate, chitosan, and polyvinyl alcohol substituted with stilbazolium groups (PVA SbQ) inhibited settlement of Balanus amphitrite (Rasmussen et al. 2002). This group also showed that the PVA SbQ surface inhibited adhesion of the marine bacterium Pseudomonas sp. NCIMB2021 (Rasmussen and Ostgaard 2003). Hydrogels based on 2 hydroxyethyl methacrylate (HEMA) reduced fouling in two algal colonization bioassays and with the addition of benzalkonium chloride remained visually clean in field testing for up to 12 w eeks (Cowling et al. 2000). Crosslinked poly(ethylene glycol) diacrylate surfaces were evaluated as foulingresistant membrane coatings. Surfaces that were more hydrophilic based on contact angle measurements exhibited less protein adsorption (Ju et al.

PAGE 37

37 2 009). The antifouling character of these surfaces is representative of high surface energy regime of the Baier curve. Amphiphilic surfaces and heterogenous surfaces formed by patterning or mixing chemistries are other examples of nontoxic polymer coating designs that have shown antifouling properties (Grozea and Walker 2009b). Self assembled and nano structured polymer thin films were also reviewed in the context of antifouling (Krishnan et al. 2008). Another class of chemical deterrents to biofouling inc ludes naturally occurring biomolecules. For instance, it has been proposed that enzymes could break down the EPS of attached cells (Dobretsov et al. 2007, Olsen et al. 2007) or catalyze the production of repellent compounds (Kristensen et al. 2008, McMast er et al. 2009). However, it remains difficult to identify a single enzyme which is effective universally. Numerous chemicals have been isolated from natural sources and several reviews discuss specific strategies in detail (Pawlik 1992, Clare 1996, Ritt schof 2000, Fusetani 2004). Dalsin, et al. (Dalsin and Messersmith 2005) have provided an extensive review of bioinspired polymers. These chemistries attempt to mimic the complex biopolymers which naturally resist fouling, such as the adhesive pad of the mussel. It is clear that protein adsorption and subsequent biofouling are strongly influenced by surface chemistry. Correlations have been observed between protein adsorption and biofouling in both the marine and biomedical environments. Resistance to protein adsorption could be used as an inexpensive way to screen new materials for antifouling properties. A single chemistry has not yet emerged as a universal antifouling strategy. However, a variety of surface chemistries have shown promise as foulingrelease

PAGE 38

38 coatings. A combination of chemical and physical antifouling strategies is therefore necessary to produce an optimal coating. Physical Antifouling Strategies It has been recognized that cells respond to substratum topography since 1914 when Har rison observed that fibroblasts found in the embryonic nervous tissue of frogs elongated when cultured on spider silk (Harrison 1914). This phenomenon was later termed contact guidance by Paul Weiss after obtaining similar results when growing nerve cel ls on glass fibers (Weiss 1945). Recently, techniques developed in the microelectronics industry, such as photolithography and electron beam lithography, have been used by several research groups including our own to create molds for producing microand nanoscaled topographies with various shapes and spatial arrangements (Curtis and Wilkinson 1998, van Kooten and von Recum 1999, Carman et al. 2006). Microtopography, in the marine environment, has been shown to deter biofouling on mollusk shells (Scardin o, et al. 2002, Bers and Wahl 2004) and affect attachment of barnacles (Berntsson et al. 2000, Schumacher et al. 2007a) and bacteria (Scheuerman et al. 1998). Nearly eight years ago our group designed engineered microtopographies composed of pillars or ridges with various heights (5 or 1.5m) and spacings (5 or 20 m) using photolithographic techniques. These particular patterns were found to systematically enhance settlement of the spores of Ulva when created in poly(dimethyl siloxane) elastomer (PDMSe) ( Callow et al. 2002). The addition of silicone oils to the PDMSe reduced overall Ulva settlement, but did not decrease settlement on microtopographies compared to smooth control surfaces (Feinberg et al. 2003, Hoipemeier Wilson et al. 2004). Carman et al. demonstrated in 2006 that a bioinspir ed

PAGE 39

39 surface, Sharklet AF (Figure 2 5 ), reduced Ulva settlement by 86% compared to smooth when feature width and spacing were 2m (Carman, et al. 2006). These dimensions are smaller than the average diameter of the spore body of Ulva (~5m). These experim ents implied that the width and spacing of topographical features necessary to deter biofouling must be tailored to the size of the organism. Figure 28 White light optical profilometry image of Sharklet AF. [Reprinted with permission from Magin CM Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 41, Figure 6)] Contact guidance was observed for endothelial cells cultured on ridges, pillars, and Sharklet AF topographies (Feinberg, et al. 2003, Carman, et al. 2006). Additionally, Feinberg (Feinberg 2004) demonstrated that a pattern of 3m diameter circles of the ECM protein fibronectin on PDMSe could be used to direct formation of focal adhesions and grow an endothelial cell monolayer with density and morphology similar to that of the native artery (Feinberg et al. 2009). Hatcher and Seegert (Hatcher et al. 2002) showed that scaffolds of various porosities made from polyvinylpyrrolidone modified bioactive glass fibers could increase proliferation of rat mesenchymal stem cells preceding differentiation. Chung and others demonstrated that the Sharklet AF topography inhibited biofilm formation of Staphylococcus aureus over a period of 21 days (Chung et al. 2007).

PAGE 40

40 The change in wettability of a surface due to microtopographical roughness is also likely to be a contributing factor to antifouling properties. The topic of wetting and dewetting on rough surfaces has been thoroughly reviewed by Qur and colleagues (Bico et al. 1999, Bico et al. 2002, Quere 2008). The application of surface roughness to alter wettability for antifouling coatings especially superhydrophobic coatings has also been reviewed extensively (Genzer and Efimenko 2006a, Howell and Behrends 2006, Marmur 2006). Long and others reported recently that seven different engineered microtopographies exhibited contact angle anisotropy between contact angles measured parallel and perpendicular to the features (Long et al. 2009). This work demonstrates the importance of anisotropy in the design and study of antifouling surfaces. An engineered roughness index was developed that demonstrated a negative correlation between the settlement behavior of the zoospore of Ulva with wettability of engineered microtopographies (Figure 26 ). The original ERI em pirically ratios the product of Wenzels roughness factor (Wenzel 1936) (r) and the degrees of freedom of the pattern ( df ) to the depressed surface area fraction (1 s) (Schumacher et al. 2007b). Bico, Qur, and others (Bico, et al. 1999, Bico, et al. 2002, Quere 2008) described the surface solid fraction (1s) as the ratio of the depressed surface area between features and the projected planar surface area. The surface solid fraction is equivalent to 1f1, the solidliquid interface term of the CassieBaxter equation for wetting (Cassie and Baxter 1944). A biological attachment model based on a modified ERI was recently proposed by Long et al (Long, et al. 2010) In this model the ERI was changed by replacing the

PAGE 41

41 degrees of freedom (df ) of the pattern with the number of distinct features in the pattern (n). Figure 29 Correlation of Ulva spore settlement density and Engineered Roughness Index (ERI). The calculated ERI for the tested PDMSe surfaces is plotted against the experi mental mean spore density (spores/mm2)+SE (n=3). Reprinted from (Schumacher, et al. 2007b) with permission from the publisher Taylor & Francis Group ( http://www.informaworld.com). [Reprinted with permission from Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644 (Page 42, Figure 7)] The number of attached organisms per area was normalized to the number of organisms attached to a smooth control. The data were transformed by t aking the natural logarithm (Equation 2 1). ln A ASM = m r n 1 s b ( 2 1) This transformation unified the data from numerous experiments onto a single plot. The attachment density of spores of Ulva for all of the experiments showed a high statistical correlation (R2=0.88) to the attachment model. The attachment model also correctly predicted a further reduction of Ulva attachment on a newly designed topography with a higher ERI value (Long, et al. 2010) This relationship can be used to create new engineered microtopographies that further reduce Ulva attachment.

PAGE 42

42 The elastic modulus of a substratum is another physical factor that has been shown to influence bioadhesion. The adhesion strength of a disc to an elastomeric substrate was proposed by Kendall (Kendall 1971) as: = 2 = (8 3( 1 2) )1 / 2 (2 2 ) i n which P = critical stress for removing the disc, Fc = critical force, a = radius of the rgy between the disc and substr ate, E = elastic modulus of the substrate, correlation between elastic modulus and surface energy of a material (Kendall 1971, Brady 1999). Vascular graft research has shown that intimal hyperplasia can be caused by compliance mismatch between the graft and the vessel wall and poor reendothelialization of the luminal surface (Tai et al. 2000). It has also been reported that substratum elasticity directs stem cell differentiation into specific lineages (Engler et al. 2006). Li kewise, in the area of marine biofouling it has been proposed that the release behavior of pseudobarnacles and spores from various coatings is inversely proportional to the pull off stress and scales with elastic modulus (E1/2) (Brady 2001, Chaudhury et al 2005). The importance of hydrodynamics to the fouling process cannot be overlooked. Work by Crisp in 1955 showed that there is a critical velocity gradient at the surface for barnacle cyprids to attach (Crisp 1955). A critical observation by Purcell (P urcell 1977) states that our physical intuition of swimming does not apply to microorganisms. Bacteria and cells swim in an environment of very low Reynolds number ( E. coli, Re~105). As a result, these organisms live in a world where viscous forces domi nate over inertial forces. It has been demonstrated both empirically and experimentally that E. coli

PAGE 43

43 is attracted to the walls of a container purely by hydrodynamic interactions (Berke et al. 2008, Lauga and Powers 2009). This hydrodynamic attraction is similar to other phenomena described by Vogel (Vogel 1994). For instance, if two spheres fall next to each other in a fluid, they are attracted to each other by viscous forces. These hydrodynamic interactions may initiate the settlement process by allowi ng the organism to find the surface. One approach to create new antifouling surfaces may be to utilize the concept of fluid slip. Fluid slip is the boundary condition in which the fluid has a finite velocity at an interface (Granick et al. 2003). This is in contrast to the no slip boundary condition which is commonly assumed in fluid mechanics. The no slip boundary condition is relevant to a fluids moving over air, which occurs in the case of superhydrophobic materials in the nonwetted or CassieB axter state (Ybert et al. 2007, Voronov et al. 2008, Lee and Kim 2009). It may be possible to prevent hydrodynamic attraction of swimming organisms through the use of fluid slip. Fluid hydrodynamics also contributes to the antifouling character of biologi cal tissues. In the case of vascular implants, thrombus formation is a common problem. Thrombogenesis follows a typical biofouling cascade in which proteins that are present in blood adsorb to the surface, followed by platelets and red blood cells. Therefore, a healthy endothelium requires a constant supply of both thrombogenic and anti coagulant factors. These factors are maintained by fluid flow through the blood vessel (Edmunds 1996). Fluid shear affects platelet and red blood cell physiology and subsequent thrombus formation (Spijker et al. 2003). The disruption of native fluid flow in a vessel either by injury or placement of an implant influences the balance of

PAGE 44

44 these thrombogenic factors. Therefore, fluid flow plays an integral role in the fouling process. Conclusion Biofouling is a dynamic process which spans numerous length scales and involves a complex variety of molecules and organisms. Antifouling strategies, therefore, must include both chemical and physical concepts. Nature provides e xamples of antifouling and fouling release surfaces that emphasize the importance of these factor s. Physical cues, such as surface roughness and fluid hydrodynamics, can act singularly or in concert with surface chemistry to enhance or inhibit the attachm ent of organisms to a surface. Chemical cues, especially surface energy, influence not only the ability of an organism to initially attach to a surface, but also the degree of foulingrelease from the surface once adhesion has been established. At this p oint, no single technology has been demonstrated universally effective at either antifouling or foulingrelease. The environmental impacts of biofouling demonstrate the need to continue the development of strategies that are truly nontoxic and broadly ef fective. Confronting the complexity of biofouling requires the cooperative effort of industry and academia in all disciplines of science and engineering. Influence of Surface Chemistry and Topography on Cell Culture Contact Guidance The first evidence of contact guidance was reported nearly a century ago when it was observed that fibroblasts found in the embryonic nervous tissue of frogs elongated when cultured on spider silk ( Harrison 1914, Weiss 1945) Recently, techniques such as photolithography and electron beam lithography have been used by several research groups to create molds for producing microand nanoscaled topographies with various

PAGE 45

45 shapes and spatial arrangements (Curtis and Wilkinson 1998, van Kooten and von Recum 1999, Carman, et al. 2006) Even though these patterning techniques are now wellestablished, very little is known about the mechanisms underlying cell sensi ng and response to topographies. It is clear that by altering substratum topography, it is pos sible to change cell morphology In recent studies c ellular micro environments have been shown to direct cell function, (Dalby 2005, Bettinger et al. 2009, Moon, et al. 2009, Schulte et al. 2009, Liliensiek, et al. 2010) and topography has been used to direct differentiation of stem or progenitor cells into a specific lineage (Engl er, et al. 2006, Chai and Leong 2007, Saha et al. 2007, Reilly and Engler 2010) In 2007 it was demonstrated that topography could enhance differentiation of human mesenchymal stem cells into the neuronal lineage (Yim et al. 2007) The aim of the current research was to create a cell culture substrate for small diameter vascular graft applications that uses a combination of surface chemistry and microtopography to reendothelialize in situ through the capture and differentiation of circulating EPCs into the EC phenotype. Medical and Economic Impact of Controlling Cell Adhesion Coronary heart disease is the leading cause of death in the United States (U.S). According to the latest mortality data, nearly 2,300 Americans die from cardiovascular disease every day, an average of one death every 38 seconds (Heart Disease & Stroke Statistics 2010) Coronary heart disease is a result of atherosclerosis, the narrowing and hardening of artery walls due to a buildup of fatty substances including cholesterol i n the arteries that supply blood to the heart muscle. This buildup can disrupt or even block the blood flow and oxygen supply to the heart ultimately resulting in acute myocardial infarction (heart attack).

PAGE 46

46 Angioplasty and stenting have been used to increase lumen size and restore blood flow in smalldiameter vessels such as the coronary arteries when heart attack is imminent or has already occurred, but not without complications. Instent restenosis renarrowing of the blood vessels occurs within 6 months in 50% of patients that undergo stenting (Padera and Schoen 2004) For this reason, the most common way to treat coronary blockages is to create new passages f or blood flow by grafting. Since there are currently no FDA approved small diameter vascular grafts, autogenous veins or arteries are typically used to reroute blood flow around the blocked arteries (Heart Disease & Stroke Statistics 2010) The patients internal mammary artery or saphenous vein are the most common grafts; however, damage to the native vasculature or a previous grafting procedure may make bypass surgery impossible, especially for procedures that require multiple grafts. Even when healthy autogenous vessels are available, it is less desirable to remove them from their positions than it would be to insert a prost hetic graft (Nerem and Seliktar 2001) In 2006, approximately 448,000 inpatient bypas s procedures were performed. The American Heart Association estimates that the total cost direct and indirect o f coronary heart disease in 2010 was $177.1 bil lion (Heart Disease & Stroke Statistics 2010) As the average age and longevity of the U.S. population increases, these figures will certainly increase and the need for new small diameter grafting techniques that have patency rates comparable to those of autogenous grafts will become more urgent. Summary Physico chemical cues presented by a surface influence bioadhesion and cellular function in both marine antifouling applications and mammalian cell culture substrates. The effects of engineered microtopographies and specifically designed surface

PAGE 47

47 chemistries on cell attachment and function are investigated in this work. The findings are applied to two different applications: marine antifouling and direction of mammalian cell function. Various combinations of surface topography and chemistry were evaluated for the deterrence of fouling cells and organisms in the marine environment and as substrates for mammalian cell culture. A model is presented that relates the surface energy of a substratum and the size and motility of an organism to the attachment density of a variety of marine fouling organisms. Surface chemistries and topographies that direct the elongation and orientation of human vascular cells are also presented.

PAGE 48

48 CHAPTER 3 ENGINEERED ANTIFOULI NG MICROTOPOGRAPHIES: THE ROLE OF REYNO LDS NUMBER IN A MODEL THAT PREDICTS ATTACHME NT OF ZOOSPORES OF ULVA A ND CELLS OF COBETIA MARINA2 Introduction Biofouling, the undesired accumulation of organic molecules, living organisms, and the metabolites of these organisms on a surface (Costerton et al. 1995, Shea et al. 1995, O'Toole et al. 2000), is a significant env ironmental concern with consequences beyond barnacles and bivalves. Biofouling increases drag on vessels by increasing hull roughness, which leads to higher fuel consumption and cost (Townsin 2003, Schultz 2007). The United States (US) Naval Sea Systems Command estimates that fuel costs increase 6 to 45 percent if the hull is fouled, depending on the size of the vessel. Slime layers composed of bacteria and diatoms that develop in the early stages of biofouling (Molino, Campbell et al. 2009, Molino, Chil ds et al. 2009) have been shown to significantly increase hydrodynamic drag and increase fuel consumption (Schultz 2007). Increased fossil fuel consumption is not the only environmental concern raised by biofouling. Hull fouling has been shown to be a pr imary cause for the introduction and spread of nonindigenous marine species (Otani et al. 2007, Pettengill et al. 2007, Piola and Johnston 2008, Yamaguchi et al. 2009). The ideal solution to the detrimental effects of biofouling will be a green technology possessing both antifouling and foulingrelease properties (Genzer and Efimenko 2006, Marmur 2006) to reduce drag and fuel consumption while remaining nontoxic. 2 Reprinted with permission from the publisher Taylor & Francis Group (http://www.informaworld.com) from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifouling Microtopographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727.

PAGE 49

49 Polydimethylsiloxane elastomers (PDMSe), more commonly known as silicones, are currently m arketed as nontoxic marine coatings. These coatings are known to have foulingrelease properties (Chaudhury et al. 2005, Holm et al. 2006, Wendt et al. 2006) due to their low surface energy and low modulus (Brady and Singer 2000, Chaudhury, et al. 2005); however, they are not inherently antifouling (Molino, Campbell, et al. 2009, Molino, Childs, et al. 2009). Nontoxic antifouling technologies that focus on the manipulation of surface topography in PDMSe have been designed to deter attachment of fouling organisms (Carman et al. 2006, Schumacher, Aldred et al. 2007, Schumacher, Carman et al. 2007, Schumacher et al. 2008, Long et al. 2010). Carman et al. presented a biomimetically inspired surface topogr aphy, Sharklet AF that reduced attachment of Ulva zoospores by 86% (Carman, et al. 2006). Ulva is a common green alga commonly found in marine biofilms found on ships, submarines, and other underwater structures. The Ulva plant produces motile spores th at disperse and colonize surrounding surfaces (Callow and Callow 2000). Evidence suggests that the swimming spores are able to select a surface suitable for attachment based on topographical, biological (Joint et al. 2002), chemical, and physicochemical c ues (Ederth et al. 2009, Schilp et al. 2009). A recent report proposes that the flagellar motion of a swimming unicellular alga, Chlamydomonas is coupled to its hydrodynamic environment (Polin et al. 2009). An empirical relationship called the Engineered Roughness Index (ERI) has been proposed to quantify topographical roughness based on parameters that describe surface energy. Results have demonstrated a correlation between the attachment behavior of Ulva zoospores and ERI (Schumacher, et al. 2007). A predictive attachment model was developed based on the second-

PAGE 50

50 generation ERI (ERIII) and correctly predicted spore density on three newly designed surface topographies. Four separate Ulva attachment data sets showed excellent correlation (R2=0.88) wit h the attachment model (Long, et al. 2010). The ERI is a dimensionless ratio based on Wenzels roughness factor (r), the depressed surface area fraction (1s), and originally the degrees of freedom of the pattern ( df ) (Schumacher, et al. 2007). Recently the ERI was extended to include additional topographical designs (Long, et al. 2010). The df term was replaced with n, which is the number of unique features in each topography. The ridges topography is comprised of one distinct feature thus an nvalue of one. T he triangle/pillars topography is composed of two unique features: an equilateral triangle and a round pillar (n=2). ERIII is represented as = ( ) / ( 1 s ) (3 1) Wenzels roughness factor (r) is the ratio of the actual surface area to the projected planar surface area (Wenzel 1936). The actual surface area includes the surface area of the feature tops, sides, and depressed surf ace area between features. The depressed surface fraction (1s) described by Bico, Qur, and others (Bico et al. 1999, Bico et al. 2002, Quere 2008) is the ratio of the depressed surface area between features and the projected planar surface area. The depressed surface fraction is equivalent to 1f1, where f1 is the solidliquid interface term of the CassieBaxter equation for wetting (Cassie and Baxter 1944). The topographies studied were selected to cover a range of ERIII values (Figure 3 1). In thi s study it was hypothesized that marine bacteria attachment and biofilm formation would be inhibited by engineered microtopographies in PDMSe and this

PAGE 51

51 inhibition would scale with the ERIII value. Cobetia marina, originally isolated from a marine biofilm ( Baumann et al. 1983), was used as a model marine fouling organism. Figure 31. Scanning electron micrographs of (a) pillars, (b) ridges, (c) triangle/pillars, (d) Sharklet AF, (e) Recessed Sharklet AFTM surfaces in PDMSe. The topographies of the Sharklet AFTM and the ridges were positioned so that the features were parallel to the direction of flow when mounted in the flow cell. [Reprinted with permission from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Anti fouling Microtopographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727 (Page 720, Figure 1)]. C. marina attachment to defined solid surfaces has been studied extensively through the use of self assembled monolayers (SAMs) (Ista et al. 1996, Ista et al. 1999, Ista et al. 2004, Ista et al. 2010) and has been reported to influence secondary colonization (Shea, et al. 1995). The present report shows the correlation between ERIII value and attachment of the marine bacteri um C. marina Additionally, a single equation is presented that relates attachment densities of Ulva and C. marina to the ERI multiplied by the estimated Reynolds number for the individual organisms

PAGE 52

52 Materials and Methods Materials The base material used for engineered topographical modification was a platinum catalysed polydimethysilo xane elastomer (PDMSe), Silastic T 2 (Dow Corning Corporation). The elastomer was prepared by hand mixing ten parts resin and one part curi ng agent by weight for 5 min. The mixture was degassed under vacuum (28 30 in. Hg) for 30 min, removed from the vacuum chamber, and poured into negative topographical molds to cure for 24 h at ~22C. Pattern Designs The patterns st udied included Sharklet AF, recessed Sharklet AF, ridges, pillars, and triangle/pillars (Schumacher, et al. 2007). Sharklet AF previously described (Carman, et al. 2006, Schumacher, et al. 2007) consists of 2 m wide ribs of various lengths (4, 8, 12, and 16 m) that are combined by feature length in the following order: 4, 8, 12, 16, 12, 8, and 4 m at a feature spacing of 2 m to form a diamond. This diamond of protruding features was the repeat unit for the arrayed pattern. The spacing between each diamond unit was 2 m. This pattern was inspired by and is similar to the skin of a shark (Bechert et al. 2000) in terms of feature arrangement. Recessed Sharklet AF is the negative of Sharklet AF. It has the same arrangement of features indented int o the surface instead of protruding out from the surface. Ridges and pillars were designed with an analogous feature spacing of 2 m. Ridges were continuous in length and separated by 2 m. The pillars were hexagonally packed so the distance between any two adjacent pillars was 2 m. The triangle/pillars pattern is a multi feature pattern formed by replacing a set of six hexagonally packed 2 m pillars with a 10 m equilateral triangle. This triangle

PAGE 53

53 placement maintained a 2 m feature spacing between the edges of the triangle and the pillars. Pattern Fabrication Pattern designs were transferred to photoresist coated silicon wafers using standard photolithographic techniques described previously (Schumacher, et al. 2008). The patterned wafers were deep reactive ion etched using the Bosch process to a depth of approximately 3 m to create negative molds of the engineered topographies. Wafers were subsequently stripped of photoresist with an O2 plasma etch. Hexamethyldisilazane was vapor deposited ont o the processed wafers to methylate the surfaces and prevent adhesion of PDMSe during the replication process. Topographical Replication Engineered topographies were transferred to PDMSe by replication of the patterned and etched silicon wafer. The resulting engineered topographies contained features that protrude from the surface at a height of approximately 3 m. Pattern fidelity and feature height were evaluated with light and scanning electron microscopy. Sample Preparation Samples were provided as smooth PDMSe surroundi ng two 13 mm x 13 mm squares of engineered topographies of interest adhered to glass coverslips (60 mm x 24 mm, No. 2 thickness). Engineered topographies were attached to glass coverslips using a two step process previously describe d (Carman, et al. 2006). Topographies (smooth, Sharklet AF, recessed Sharklet AF, ridges, pillars, and triangles/pillars) were randomly assigned to one of two positions on each coverslip. Sharklet AF and ridges topographies were positioned on coversli ps so that features were parallel to the direction of flow when mounted in the flow cell. The resulting coverslip was

PAGE 54

54 approximately 0.8 mm thick and contained an adhered PDMSe film with two 13 mm x 13 mm square areas containing topography bordered on all sides by smooth (no topography) areas. Bacteria Chemostat culture A logarithmic chemostat culture of C. marina ATCC 25374 (American Type Culture Collection, Manassas, VA) strain (Baumann, et al. 1983, Arahal et al. 2002) was established in modified basal medium (200 mM NaCl, 50 mM MgSO47H2O, 10 mM KCl, 10 mM CaCl22H2O, 19 mM NH4Cl, 0.33 mM K2HPO4, 0.1 mM FeSO4 7H2O, 5 mM Tris HCl (pH 7.5), and 2 mM glycerol) as described previously (Ista, et al. 1996). The chemostat was maintained at a flow rate of 1 ml min1 with constant stirring resulting in a cellular concentration of 5 x 107 cells ml1. Stationary phase culture A stationary phase culture of C. marina was grown in 1 L of modified basal medium described above. The i noculated culture grew for 21 h at 25C while shaking at 400 RPM. C. marina A ttachment A ssay Two bacterial attachment assays were performed with C. marina one using the chemostat culture and one the stationary phase culture described above. Samples were sterilized by immersion in ethanol for 20 min, soaked in artificial seawater for 1 h and then placed into a laminar flow cell apparatus (Ista, et al. 1996) that was mounted onto the stage of an optical microscope (Zeiss Axioscope 40). The sample coverslip forms the top plate of the fl ow chamber to minimize gravitational effects on attachment. The flow cell was then connected to the culture vessel through tubing and a peristaltic

PAGE 55

55 pump (Ista, et al. 1996, Ista, et al. 2004). The cells were introduced into the flow cell at a rate of 2x105 L s1 for 2 h. Under these conditions for a flat surface, the Reynolds number was ~2 x 103, indicating laminar flow and the surface shear rate was 6.6x102 s1 (Ista, et al. 2004). Bacterial attachment was monitored through a camera attached to a phase contrast microscope. Ten random images were taken for each topography at 15 min intervals. Images were acquired with Axiovision software and processed with ImageJ software (Rasband 19972009). Each image was first converted to 8 bit format and then processed using the Fast Fourier Transform (FFT) bandpass filter associated with ImageJ to eliminate background unevenness. The cell densities were determined through direct counting using the ImageJ Cell Counter plugin. The attached cells in 10 randomly selected fields of view were counted at each time point. The average number of cel ls per square millimeter (cells mm2) was calculated for each replicate. Three replicates were analyzed for each topography. The position of each topography was varied in the flow stream. Statistical Methods Cell counts were repor ted as the mean number of cells mm2 from 10 counts on each of three replicate topographies (n=30) with 95% confidence intervals. The cell count data were transformed using a natural logarithm. S tatistical differences between surfaces were evaluated with the transformed data using oneway analysis of variance (ANOVA) and Tukeys test for multiple comparisons. The transformed mean number of cells mm2 for each engineered topography was plotted versus time for kinetics analysis. The transformed data was also plotted against the calculated ERIII to determine if any correlations existed between C. marina

PAGE 56

56 attachment and the ERIII. Regression analysis was performed to evaluate the strength of the cor relation. Results and D iscussion Only slight variations in attachment densities were measured over the entire assay period. This includes samples exposed to both logarithmic growth phase and stationary growth phase C. marina (Figure 32 ). Figure 32. C. marina attachment vs time for (A) chemostat culture (logarithmic growth phase), all surfaces; (B) chemostat culture (logarithmic growth phase), topographically modified surfaces only; (C) overnight culture (stationary growth phase), all surfaces; and (D ) overnight culture (stationary growth phase), topographically modified surfaces only. Error bars represent 95% confidence intervals. [Reprinted with permission from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifou ling Microtopographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727 (Page 723, Figure 2)]

PAGE 57

57 C. marina was calculated to attach at a mean density of 683 79 cell s mm2 and 413 51 cells mm2 on the smooth PDMSe at the 120 min time point for the logarithmic growth phase and stationary growth phase, respectively. The cell density for both growth phases was reduced significantly (Figure 3 3) on all topographies rel ative to the smooth surface at 120 min. The logarithmic growth phase cell density was lower on the triangle/pillars (1.7 2.6), Sharklet AF (0.1 0.2), and recessed Sharklet AF (1.0 1.2) topographies compared to cell density on pillars (14.2 10.2) and ridges (15.8 7). Attachment density of C. marina cells in the stationary growth phase followed the same order; i.e., lower attachment density was measured on triangle/pillars (55 0.3) and Sharklet AF (36 0.2) patterns than on pillars (126 0. 3) and ridges (166 0.1). Nearly all of the attached cells were singular or in small clusters. Cells appeared to attach in clusters around pillars on the pillar pattern The majority of cells were attached within the ~2 m wide channels of the ridge pa tterns. Nearly all cells settled next to the edges of the triangle or clustered around the pillars of the triangle/pillar patterns. Cells did not attach to the tops of the 10 m triangles. Cells were attached either between the features of the Sharklet AF pattern or around the edges of the diamond repeat units. On the recessed Sharklet AF pattern cells attached within the ~2 m wide depressions. The feature dimensions of the topographies used in each study were measured and the average heights, widths and spacings were used to calculate the ERIII value. Due to slight variations in feature height the value of the ERIII for each pattern varied between the two bacterial attachment assays (Tables 3 1 and 3 2). Recessed Sharklet AF is a newly designed t opography with a higher ERIII value (24) than Sharklet AF

PAGE 58

58 that was included in the logarithmic growth phase attachment assay to further the ERIII series. Figure 33. C. marina attachment data on PDMSe surfaces represented as mean cell density (cells mm2) +95% confidence interval (n = 30). (A) surfaces exposed to C. marina cells in the logarithmic growth phase established in a chemostat culture for 120 min; (B) surfaces exposed to C. marina cells in the stationary growth phase established through an overnight culture for 120 min. Solid horizontal bars = statistically different groups (ANOVA p = 0.05, Tukey test p = 0.05). [Reprinted with permission from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifouling Micro topographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727 (Page 723, Figure 3)]

PAGE 59

59 Table 31. Feature geometry and engineered roughness index calculations for to pographies exposed to the chemostat culture. The Table includes the calculated values for Wenzel's roughness factor (r), the depressed surface area fraction (1 Table 32. Feature geometry and engineered roughness index calculations for topographies exposed to the overnight culture. The Table includes the calculated values for Wenzel's roughness factor (r), the depressed surface area fraction (1 Feature geometry Engineered Roughness Index Height Width Spacing r n 1 -s ERIII (r n)/(1 s ) Pillars 3.2 0.1 2.0 0.1 2.0 0.1 2.5 1 0.77 3.2 Ridges 3.0 0.1 2.3 0.1 1.7 0.1 2.5 1 0.42 5.9 Triangle/Pillars 3.2 0.1 2.0 0.1, 8.1 0.1 2.1 0.1 2.2 2 0.72 6.1 Sharklet AF 1.8 0.1 1.8 0.1 2.0 0.1 1.9 4 0.57 13 Smooth n/a n/a n/a 1 0 1 0 The natural logarithm of the normalized mean cell density measured on each surface at 120 min was plotted against ERIII of each pattern (Figure 3 4). A lin ear Feature geometry Engineered Roughness Index Height Width Spacing r n 1 s ERI II (r n)/(1 s ) Pillars 3.2 0.1 2 0.1 2 0.1 2.5 1 0.77 3.2 Ridges 2.6 0.1 2.4 0.1 1.7 0.1 2.3 1 0.41 5.5 Triangles/Pillars 3.2 0.1 1.9 0.1, 8.0 0.1 2.1 0.1 2.1 2 0.74 5.8 Sharklet AF 2.9 0.1 2.3 0.1 1.7 0.1 2.5 4 0.52 19 Recessed Sharklet AF 3.1 0.2 2.1 0.1 2.0 0.1 2.5 4 0.42 24 Smooth n/a n/a n/a 1 0 1 0

PAGE 60

60 regression model was fit to each set of data. An inverse linear relationship existed between mean cell density and ERIII for bacterial attachment in both the stationary (R2=0.78, p=0.047) and logarithmic (R2=0.40, p=0.038) growth phases. Figure 34. The attachment model shows the correlation between C. marina attachment normalized to attachment on a smooth surface and the engineered roughness index (ERIII) at 120 min for the stationary and logarithmic growth phases. Plotted is the calculated ERIII for each pattern tested vs the natural logarithm of the experimental mean bacterial cell density at the 120 min time point minus the mean bacterial cell density on a smooth surface at the same time point (ln(A/A0)) (n = 30). Data for attachment of zoospores of Ulva taken from Long et al. (2010) are also plotted. [Reprinted with permission from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifouling Microtopographies: The role of Reynolds number in a model that predic ts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727 (Page 724, Figure 4)] The lowest mean cell densities for the logarithmic growth phase C. marina correlated with the highest ERIII values, e.g. triangle/pillars (5.8), Sharklet AF (19), and Recessed Sharklet AF (24), but were not different statistically. The second lowest cell densities were measured on ridges and pillars (ERIII values of 5.5 and 3.2, respectively). Cell densities measured in the stationary phase at tachment assay followed a similar trend. Triangle/pillars and Sharklet AF with the highest ERIII values

PAGE 61

61 (6.1 and 13 respectively) had the lowest cell densities. Ridges and pillars had the next highest ERIII values (5.9 and 3.2) and the second lowest att achment densities. The regression analysis for the cell attachment to the surfaces established a statistically significant trend for both the logarithmic growth phase and stationary growth phase. These trends were analy zed in terms of the attachment model as E quations ( 3 2) and ( 3 3). The variable A is the attachment on a surface with a given ERIII value and A0 is the attachment on a smooth surface (ERIII=0). ln 0 = 0 .40 (3 2) ln 0 = 0 .20 (3 3) Adhesion of marine bacteria to surfaces has been attributed to many factors: substratum composition (Ista, et al. 2004, Ekblad et al. 2008), surface chemistry (Is ta, et al. 1996, Ista, et al. 1999, Poolman et al. 2004, Cordiero et al. 2009), substratum mechanical properties (Ekblad, et al. 2008, Cordiero, et al. 2009), and surface roughness (Kerr and Cowling 2003). Substratum composition, substratum mechanical properties and surface chemistry were kept constant in this study by using PDMSe for all surfaces. Surface roughness was systematically studied by using the ERIII algorithm to describe engineered microtopographies. Plots of C. marina attachment over time i ndicated that bacteria in both the logarithmic and stationary growth phase reacted to surface topography nearly instantaneously (Figure 3 2). Bacterial attachment did not change significantly from the first time point (15 min) to the last (120 min). This response was due to physical, not chemical, changes in the substratum. These results are consistent with the work of Kerr

PAGE 62

62 and Cowling in which the effect of surface roughness on bacterial adhesion was observed to be almost instantaneous (Kerr and Cowling 2003). Cell densities measured at the last time point indicated that all topographies showed a statistically significant reduction in the density of attached cells in both growth phases relative to the smooth surface (Figure 3). A fairly strong inverse l inear relationship existed between the natural logarithm of the mean cell density and ERIII value for both bacterial attachment assays (Figure 3 4). These relationships, however, did not follow the same regression line. Attachment of C. marina establishe d in a chemostat culture (logarithmic growth phase) correlated with a line with a steeper slope (m = 0.40) than the cells in the stationary growth phase (m = 0.20). The slopes of these lines were both different from the slope of the regression line for Ulva spore attachment versus ERIII value (m = 0.071) from four separate studies reported in previous work (Long, et al. 2010). It is possible that the slope of the line in the attachment model (m) encompasses the organisms sensitivity to a surface (Equation 3 4). ln 0 = (3 4) This sensitivity could be related to a number of factors including, the size and shape (Kerr and Cowling 2003, Carman, et al. 2006), motility (Shea, et al. 1995), and surface chemistry (Shea et al. 1991, Ista, et al. 2004, Poolman, et al. 2004, Cordiero, et al. 2009, Ista, et al. 2010) of the attaching organism and/or conditioning of the surface by culture medium and secreted products (Jain and Bhosle 2009). It has been observed that in the late st ationary phase C. marina cells tend to shrink to approximately 1 m in diameter and take on a more rounded shape. The same bacteria cells i n the logarithmic

PAGE 63

63 growth phase are rodshaped, approximately 2 m in length and often joined in pairs. This size and shape difference may influence bacterial attachment. The results from this study also showed stationary phase cells adhered to hydrophobic topographic surfaces in higher numbers than logarithmic phase cells. The higher densities of bacteria attached to surfaces exposed to the stationary phase culture could be due to the fact that there were simply more bacteria cells in the overnight culture than the chemostat culture. A standard overnight culture has approximately 109 cells mL1 whereas the chemostat is maintained at approximately 107 cells mL1. The Reynolds number is the ratio of inertial forces to viscous forces in fluid flow. Viscous forces dominate inertial forces in conditions such as those experienced by C. marina and spores of Ulva while moving through water (Purcell 1977, Berg 1983, Dusenbery 2009). Therefore C. marina and Ulva operate at low Reynolds number. Alternatively, the Reynolds number can be considered to represent the scale separation in the flow. That is to say the bacteria and alga spores are not large compared to the smallest scales in the flow at which viscosity dissipates kinetic energy. Ulva is a flagellated cell capable of propelling itself through the water. The C. marina in this study is not flagellated. The Reynolds numbers for the swimming Ulva spore and the C. marina in both growth phases were calculated using the following equation: = (3 5) velocity of the organism relative to that of the fluid, and L is the characteristic length of the organism (see Appendix A for full calculation) The characteristic length is a dimension relevant to the geometry of the flow; in this case it is the diameter of the body

PAGE 64

64 moving through the fluid. The characteristic length, i.e., the diameter of the spore body at its widest point, and velocity of the Ulva spore were taken to be 5 m and 150 m s1, respectively (Callow et al. 2002, Heydt et al. 2009). The characteristic length of C. marina varies with growth phase. It is 2 m for the logarithmic growth phase and 1 m for the stationary growth phase. The velocity of the bacteria relative to the fluid near the wall was estimated to be 20% of the average fluid velocity in the flow cell. When the bacteria enter the boundary layer near the wall, the velocity will be considerably less than the average bulk velocity due to the noslip boundary condition. Flow velocity of the fluid, i.e., a key component of the Reynolds number, has been demonstrated to affect attachment density of Ulva (Granhag et al. 2007). The size and motion of the organisms were incorporated into the slope of the attachment model by multiplying the ERIII value by the Reynolds number of the organism. m = (3 6) ln 0 = ( ) (3 7) The incorporation of the Reynolds number into the attachm ent model allowed four separate Ulva attachment assays (Schumacher, et al. 2007, Schumacher, et al. 2008, Long, et al. 2010) to be combined with both C. marina attachment assays into a single data set that yields a regression with high correlation to ERIIIRe (R2 = 0.77) (Figure 3 5). Further work is planned to investigate the role of the slope of the attachment model line as an indicator of the organisms sensitivity to a surface. This work could include attachment of different species or organisms such as new or mutant strains of bacteria to elucidate which factors contribute to the sensing of topographically modified

PAGE 65

65 surfaces. It could also include altering flow conditions to change the organisms Reynolds number and investigate the role of the hydrodynamic environment on biological attachment. Figure 35. The attachment model shows the correlation between attachment of zoospores of Ulva and cells of C. marina normalized to attachment on a smooth surface and the engineered roughness index (ERIII) multiplied by the Reynolds number of the organism. [Reprinted with permission from Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifouling Microtopographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727 (Page 725, Figure 5)] Microtopographies created in non toxic materials such as PDMSe are a green alternative to biocidal methods for reducing biofouling. The attachment model has been demonstrated to correlate the ERIII value multiplied by the Reynolds number for the organism to the attachment density of cells from the kingdoms bacteria and plantae. An inverse linear relationship exists between mean cell densities for the two di fferent organisms and ERIIIRe. Plots of C. marina attachment over time indicate that both

PAGE 66

66 logarithmic phase and stationary phase cells reacted to surface topography within 15 min of exposure to the surface. At the 120 min time point all topographically m odified surfaces showed statistically significant reduction in attachment compared to smooth for both bacterial attachment assays. The incorporation of the Reynolds number to the attachment model created a regression model with high correlation of attachm ent of Ulva and C. marina in both growth phases to ERIIIRe.

PAGE 67

67 CHAPTER 4 ENGINEERED ANTIFOULING TOPOGRAPHIES: INHIBITION OF ATTACHMENT, MOVEMENT AND BIOFILM ADHESION OF NAVICULA INCERTA AND SEMINAVIS ROBUSTA Background Biofouling is an environmental and ec onomic concern. The biofouling process begins immediately upon immersion of a clean surface in seawater. Within seconds to minutes the surface becomes coated with a conditioning film. This film typically consists of glycoproteins, humic acids, proteins, carbohydrates and other macromolecules (Jain and Bhosle 2009) The nature of a conditioning film plays an important role in subsequent mic robial adhesion to the surface (Bakker et al. 2004, Jain and Bhosle 2009) Diatoms are a diverse group of unicellular algae that along with bacteria form a major component of microbial slime layers (Molino and Wetherbee 2008) Raphid diatoms, in particular, are the most common early algal colonizers (Wetherbee et al. 1998) Raphid diatoms approach a surface passively through water currents or gravity and adhere to the surface through secretion of extracellular polymeric substances (EPS) (Hoagland et al. 1993, Wetherbee, et al. 1998, Chiovitti et al. 2006) The EPS facilitates a gliding motility whereby raphid diatoms can select a position more suitable for colonization (Edgar and Pickett Heaps 1984, Wetherbee, et al. 1998, Chiovitti, et al. 2006) Diatom slimes foul surfaces in both freshwater and marine environments. Diatom biofilms of 5 10 mm in thickness developed in 6 months in the high flow conditions through canals feeding a power station increasing drag and impeding efficiency (Andrewartha et al. 2010, Perkins et al. 2010) Layers of diatoms and their b yproducts have been shown to develop on both marine antifouling and foulingrelease

PAGE 68

68 paints (Holland et al. 2004, Molino et al. 2009a, Molino et al. 2009b) and increase hydrodynamic drag leading to an increase in fuel consumption and cost (Schultz 2007) The attachment strength of diatoms has been shown to correlate with surface energy. Cells of the diatoms Navicula perminuta (Holland, et al. 2004) and Seminavis robusta (Thompson et al. 2008) adhered more firmly to hydrophobic silicone elastomers than to hydrophilic surfaces. On self assembled monolayers (SAMs) the raphid diatom Amphora also adhered more strongly to hydrophobic than hydrophilic surfaces (Finlay et al. 2002b) Recently, the adhesion strength of Navicula was strongly correlated with the wettability (35105) and surface tension of a series of nine xerogel surfaces (Finlay et al. 2010) In studies by Scardino et al. (2006), the attachment strength of four different species of diatoms was influenced by sinusoidal ripple and peak patterns created in polyimides (Scardino, et al. 2006) Cells remained attached in higher numbers on sur faces where the numbers of attachment points were the highest. These results support the Attachment Point Theory (reviewed by (Verran and Boyd 2001) which states that cells are retained on surface microtopographies due to protection from shear stresses in the surrounding environment. Therefore, organisms provided with larger numbers of contact points by the surface, will have increased at tachment strength. The Attachment Point Theory was further examined by testing attachment of the diatom Amphora, the green alga Ulva rigida the red alga Centroceras clavulatum the serpulid tubeworm Hydroides elegans and the bryozoans Bugula neritina to microtextured polycarbonate surfaces (Scardino et al. 2008) The effect of attachment points was weak for motile microfoulers, strong for large macrofouling larvae, and nonexistent for nonmotile algal spores (Scardino, et al. 2008) To further investigate the

PAGE 69

69 influence of surface topography, the surface parameters of 36 species of shelled marine mollusks were characterized and correlated with fouling resistance and foulingrelease. Foulingrelease was shown to correlate positively with the mean waviness of the surface. Shell surfaces with the highest waviness profiles showed greater foulingrelease (Scardino et al. 2009a) Silicones, a major component in antifoulingpaints, are currently marketed as a nontoxic alternative to biocidal marine coatings. Silicones exhibi t foulingrelease properties (Chaudhury et al. 2005, Holm et al. 2006, Wendt et al. 2006) due to their low surface energy and low modulus (Brady and Singer 2000, Chaudhury, et al. 2005) Since these materials are not inherently antifouling, i.e., do not prevent cells and larvae from attaching (Molino, et al. 2009a, Molino, et al. 2009b) engineered antifouling topographies have been created in polydimethylsiloxane elastomer (PDMSe) to deter the initial attachment of fouling organisms (Carman, et al. 2006, Schumacher et al. 2007a, Schumacher, et al. 2007b, Schumacher et al. 2008, Long, et al. 2010) A biomimetically inspired surface topography, Sharklet AF, reduced attachment of zoospores of Ulva by 86% (Carman, et al. 2006) and cells of Cobetia marina by up to 91% compared to smooth standards (Magin et al. 2010b) A model that predicts attachment of zoospores of the green algae Ulva to various surface microtopographies created in PDMSe was developed (Schumacher, et al. 2007b, Long, et al. 2010) The attachment model was extended to describe the attachment of cells of the bacteria C. marina to these surfaces (Magin, et al. 2010b) The current attachment model equation relates the attachment density of a spore or cell to the Engineered Roughness Index (ERIII), a representation of surface energy, (Long,

PAGE 70

70 et al. 2010) multiplied by the Reynolds number of the organism (Magin, et al. 2010b) It is important to consider the size and physical characteristics of diatoms when evaluating their responses to topographically modified surfaces All diatom cells are encased by a rigid silica wall, which determines the size of topographical features that could deter settlement. The initial attachment of cells of the diatom Navicula incerta was reduced smooth standard (Long 2009) These results are consistent with the findings of Scardino, et al (2008) that the size of settling propagules/larvae relative to the size of microtopographies was important in the selection of attachment sites (Scardin o, et al. 2008) Cells of the diatom Navicula Surfaces with feature widths or spacings greater than or equal to the size of the diatom cells had attachment results similar to that of a smooth surface (Long 2009) Initial attachment did not correlate with the current attachment model (Magin, et al. 2010b) Decreased attachment density of Navicula on topographically modified surfaces compared to smooth and other surfaces with multiple attachment points also supported attachment point theory (Long 2009) It was hypothesized in the current work that engineered antifouling topographies would inhibit the initial attachment of Seminavis compared to smooth standards. Cells of the diatom Seminavis are larger than those of Navicula (Thompson, et al. 2008) It was also hypothesized that diatom attachment strength would fol low Attachment Point Theory, i.e., engineered antifouling topographies would reduce attachment strength of cells and biofilms of diatoms. The influence of topography on diatom gliding motion across surfaces was also observed.

PAGE 71

71 Materials and M ethods Mater ials The base material used for all engineered antifouling topographies was a platinum catalysed PDMSe, Silastic T 2 (Dow Corning Corporation). The elastomer was prepared by hand mixing ten parts by weight resin and one part curing agent by weight for 5 min. The mixture was degassed under vacuum (2830 in. Hg) for 30 min to remove bubbles removed from the vacuum chamber, and poured into negative topographical molds to cure for 24 h at ambient temperature ( ~22C ) Pattern Designs The patterns tested in this report included Sharklet AF and recessed Sharklet AF (Schumacher, et al. 2007b) Sharklet AF consists of 2 m wide ribs of various lengths (4, 8, 12, and 16 m) that are combined by feature length in the following order: 4, 8, 12, 16, 12, 8, and 4 m at a feature spacing of 2 m to form a diamond (Carman, et al. 2006, Schumacher, et al. 2007b) T he spacing between each diamond repeat unit was 2 m. This pattern was inspired by the skin of a shark (Bechert et al. 2000) and is similar to it in terms of feature arrangement. Recessed Sharklet AF is the negative of Sharklet AF. The features are depressed into the surface instead of protruding from the surface Both patterns were designed with a height of approximately 3 m. Sharklet AF and Recessed Sharklet AF are referred to as +2.8SK2x2 and 3.1SK2x2, respectively (Figure 41). Sample Preparation Engineered antifouling topographies were transferred to PDMSe by replication of a patterned and etched silicon wafer (Schumacher, et al. 2008) Pattern fidelity and feature height were evaluated with light and scanning electron microscopy. Engineered

PAGE 72

72 topographies were attached to glass slides using a two step process previously described (Carman, et al. 2006) Smooth, Sharklet AF and recessed Sharklet AF topographies were evaluated. Samples were provided as smooth PDMSe surrounding one 25 mm x 25 mm square of engineered antifouling topography adhered to glass microscope slides (75 mm x 25 mm ). Figure 4 1. Topographies created in PDMSe A) Sharklet AF (+2.8SK2x2) and B) Recessed Sharklet AF ( 3.1SK2x2). Initial A ttachment of Cells Navicula incerta and Seminavis robusta cells were cultured in F/2 medium contained in 250 m L conical flasks for approximat ely 3 d until cells reached the logarithmic growth phase. Cells were washed 3 times in fresh medium before

PAGE 73

73 harvesting and diluting to give a suspension with a chlorophyll a content of approximately 0.25 mL1 (Ho lland, et al. 2004, Thompson, et al. 2008) Samples were placed in Quadriperm dishes to which 10 m L of a diatom suspension was added. Cells were allowed to attach on laboratory benches at ambient temperature ( ~20 C ) for 2 h. Samples for measuring init ial attachment were exposed to a submerged wash in seawater to remove cells which had not attached. Samples were fixed in 2.5% glutaraldehyde, air dried and cell densities were counted manually on each slide using a fluorescence microscope. Counts were m ade for 30 fields of view (0.033 mm2) on each sample. Attachment Strength Samples for attachment strength were exposed to a shear stress of 45 Pa for Navicula and 26 Pa for Seminavis in a water channel after initial attachment was established. The sampl es were subsequently fixed with 2.5% glutaraldehyde and air dried. The cell densities were determined, using the image analysis method described above, for cells remaining on each surface. Assessment of M ovement Motility was assessed by examining the cell s under a microscope (20 x magnification) The assessment took place after the allotted contact time with the surface, i.e. either 2 h or 6 d. Each slide was observed in situ with a layer of water above it, covered by a 22 x 64 mm glass coverslip. Illum ination from transmitted light stimulated movement of the diatoms. T he total number of single cells (unclumped), the number orientated with the rapheside down, and the number moving per field of view were counted after 1.5 min of illumination. Clumped d iatoms were not counted in any field of view as these are less likely to be able to orientate or move than free cells

PAGE 74

74 Counts were made for 7 fields of view on each topography and the smooth standard. T he counts were made in two rounds so that the observations for any single slide were not consecutive t o avoid any timedependence factors on movement The counts on all surfaces were completed within 1.5 h after the initial settlement period. Biofilm Growth and Attachment Strength Samples for biofilm growth were not washed after initial attachment so that initial cell density was the same on all surfaces. The samples were placed in an illuminated incubator at 18 C with a 16:8 light to dark cycle for 6 d. Biomass of the biofilms was quantified as autofluorescence of chlorophyll using an adapted multi well plate reader (Tecan Genios Plus) (430ex/670em nm). Biomass was measured before exposure to shear stress in the water channel (46 Pa) and after. Statistical Methods The cell or biofilm density per mm squared was calculated for each sample and condition (n = 90 counts). Confidence intervals were calculated at 95% using either cell counts for the attachment density data or the arcsine transformed counts for proportions. The mean densities were compared using oneway analysis of variance (ANOVA) and Tukeys test for multiple comparisons. Results Seminavis Initial Attachment and Attachment Strength Diatom cells, unlike Ulva spores, are not motile in the water column. Cells move in the water by gravity and water currents so at the end of a 2 h incubation period, the cell density will be nearly constant for all test surfaces. Therefore, differences in initial cell density were quantified following a gentle underwater washing. There was no

PAGE 75

75 significant difference in initial attachment density of Seminavis between the smooth surface and 3.1SK2x2 or between +2.8SK2x2 and 3.1SK2x2 (Figure 42). Figure 4 2. Initial attachment and attachment after exposure to shear stress (26 Pa) of Seminavis on smooth, +2.8SK 2x2 and 3.1SK2x2 patterns in PDMSe. Each bar is a mean of 90 counts from 3 replicate samples. Error bars represent 95% confidence intervals. Exposure to shear stress of 26 Pa removed more cells from +2.8SK2x2 than from the smooth surface and 3.1SK2x2 however the only significant difference was between the two Sharklet AF topographies ( F = 3.95, p = 0.02, df = 2). Removal of 5, 10 and 15% of the total cells occurred from 3.1SK2x2, smooth and +2.8SK2x2, respectively (Figure 4 2). Previous experiments with Navicula also showed lower cell density on the +2.8SK2x2 topogr aphy compared to smooth after gentle rinsing (Long 2009) In contrast to initial atta chment of Navicula, there was only a small but statistically significant decrease of 8.5% in initial attachment of Seminavis on +2.8SK2x2 compared to the smooth surface ( F = 3.17, p = 0.044, df = 2) (Figure 42). 0 50 100 150 200 250 300 Smooth +2.8SK2x2 3.1SK2x2 Seminavis Attachment (cells/mm 2 ) Initial Attachment Attachment after Exposure to Shear Stress (26 Pa)

PAGE 76

76 Assessment of M ovement of Sem i navis and N avicula Orientation of diatoms was assessed after 1 h of settlement. The majority of Seminavis cells, 88 to 96%, were in the raphedown orientation on all three topographies. T here were no differences in the number of Navicula cells that were rapheside down among topographies. However, the average proportions of Navicula cells that were rapheside down were lower, 53 to 60%. The raphedown orientation is the optimal position of adhesion and motility for diatoms. Movement of diatom cells was evaluated after 1 h of settlement. The motility of Seminavis was much lower on the engineered antifouling topographies +2.8SK2x2 and 3.1SK2x2 than on the smooth PDMSe standard. Only one cell was seen to move on 3.1SK2x2 over the 15 min observation period and no c ells were observed moving on +2.8SK2x2. Approximately 7% of rapheside down cells of Seminavis moved across the smooth PDMSe standard. The percent of rapheside down Navicula cells moving on all three surfaces was not statistically different. Only 8% of rapheside down cells were observed to be moving on the on +2.8SK2x2 and 3.1SK2x2 compared to approximately 15% on the smooth standard. Diatoms of Navicula were observed to glide across microtopographies with spacings approximately half as wide as the diatom is long. Navicula was able to glide across spacings between topographies because only a small section of the raphe needs to contact the substratum at any one time to enable movement (Edgar and Pickett Heaps 1984) Diatoms gliding on a surface, like Ulva swimming through water, operate under low Reynolds number conditions in the region of 104 (Edgar 1982) Diatoms were moving at the solid liquid interface with a thin film of water or secreted material between the cell and the substratum when movement was

PAGE 77

77 assessed. Decreased movement of Seminavis across the Sharklet AF patterns indicates that topographical modification could influence the hydrodynamic properties of this fluid layer. The large size of the Seminavis diatom compared to Navicula (Figure 4 3) could be a reason for reduced motion. Drag is proportional to the surface area in close contact with the substratum (Edgar 1982) and Seminavis cells had a higher contact area than Navicula In addition, as can be seen from Figure 4 3, if a Navicula cell orientates itself it can fit so that the entire cell is in contact with a surface on all but one of the features in the Sharklet AFTM pattern allowing the cells to move in tracks across th e surface. Seminavis cells must straddle several features which will make them unstable and perhaps less able to move as the raphe will not be in continuous contact with any surface. Figure 4 3 Seminavis (larger diatom) and Navicula (smaller diatom) on +2.8SK2x2 replicated in PDMSe. Seminavis Biofilm Growth and Attachment Strength Biofilm growth was assessed after 6 d. Initial biofilm density for Seminavis was lowest on +2.8SK2x2, higher on the smooth standard and highest on 3.1SK2x2 (Tukey

PAGE 78

78 Test = 0.5). After exposure to shear stress (26 Pa) biofilm density remained highest on 3.1SK2x2 and lower on +2. 8SK2x2 and the smooth standard (Figure 4 4). Figure 44. Density of attached diatoms of Seminavis after 6 d, before and after exposure to a shear stress of 26 Pa. Bars represent findings from single samples of each treatment. Navicula Biofilm Growth and Attachment Strength Densities of biofilms of Navicula were similar on all three surfaces after 6 d. Cells on +2.8SK2x2 and the smooth standard were clumped while those attached to the 3.1SK2x2 surface were evenly distributed. This distribution suggests the cells were more strongly bonded to the 3.1SK2x2 pattern than to +2.8SK2x2 or the smooth standard. After exposure to a shear stress of 45 Pa, biofilm density was lowest on +2.8SK2x2 and highest on 3.1SK2x2 (Figure 45). The strength of attachment of raphid diatoms to a surface is determined by the interaction energies between the substratum and EPS. The greater attachment strength of diat oms on 3.1SK2x2 could indicate that the EPS wetted the depressed areas of the surface topography causing stronger adhesion. Biofilms dominated by diatoms attach to (Molino, et al. 2009a) and 0 500 1000 1500 2000 2500 3000 3500 Smooth +2.8SK2x2 3.1SK2x2 RFU Initial Biomass Biomass after Exposure to Shear Stress (26 Pa)

PAGE 79

79 are not readily released from PDMS based fouling release coatings (Holland, et al. 2004, Holm et al. 2004) Figure 45. Density of attached diatoms of Navicula after 6 d, before and after exposure to a shear stress of 45 Pa. Bars represent findings from single samples of each treatment. Discussion The +2.8SK2x2 engineered antifouling topography reduced initial attachment and attachment strength of Seminavis and attachment strength of Navicula compared to smooth PDMSe. Historical data from four Navicula attachment strength assays was analyzed to determine the relationship between the area of the diatom in contact with the substratum and attachment strength (Long 2009) Initial attachment of Navicula on various engineered antifouling topographies was reported (Long 2009) Pillars sur faces showed no difference in initial attachment compared to smooth. The +2.7SK2x2 surface reduced initial attachment by 35% versus a smooth standard while other surfaces based on the Sharklet AF pattern had initial attachment densities similar to smooth (Long 2009) The data from one assay was first analyzed in terms of the 0 1000 2000 3000 4000 5000 6000 7000 Smooth +2.8SK2x2 3.1SK2x2 RFU Initial Biomass Biomass after Exposure to Shear Stress (45 Pa)

PAGE 80

80 Attachment P oint Theory (Verran and Boyd 2001, Scardino, et al. 2006) The average attachment density of Navicula was plotted for each topography in order of decreasing number of attachment points (Figure 46). The topographies plotted are smoot h, pillars, channels, triangle/pillars, Recessed Sharklet AF and Sharklet AF. All topographies were created with an average feature height of 3 m and average feature widths and spacings of 2 m. It is clear that these data cannot be described by the c urrent Attachment Point Theory, i.e. surfaces with the lowest number of attachment points should have the lowest diatom density after exposure to shear stress. Figure 46. Attachment density of Navicula after exposure to shear stress (45 Pa) on each topography. Arrow indicates direction of decreasing number of attachment points. The average attachment density of Navicula on a surface after exposure to shear stress (45 Pa) was then normalized by the average attachment density on a smooth standard and tr ansformed with the natural logarithm. The normalized, transformed attachment data was plotted against the calculated contact area of a diatom on each surface. The contact area was determined by first calculating the projected area of a Navicula cell. Th e projected area was estimated as an ellipse with a = 12 and b = 4 Since Navicula attachment before and after exposure to shear stress was evenly distributed across each surface, the solid surface fraction ( s) of each topography was 0 100 200 300 400 Smooth P CH TP SK +SK Cell density (cells/mm 2 )

PAGE 81

81 multiplied by the projected area of the diatom to calculate the contact area available on each surface (Figure 47). The solid surface fraction is the fraction of solid/liquid interface below a water drop or in this case diatom contacting a topographically modified surface (Bico, et al. 1999, Bico, et al. 2002) The width or spacing of topographic features on some of the surfaces tested allowed diatoms to fit either completely between or on top of features (Figure 47) (Long 2009) To calculate the contact area for these surfaces the proportion of diatoms complet ely between or on top of a feature was counted. The proportion of diatoms completely between or on top of a feature was multiplied by the projected area of the diatom and added to the proportion of diatoms spanning across features multiplied by the projec ted area of a diatom and by the solid surface fraction. Figure 4 7. Navicula attachment after gentle washing on A) +2.8SK2x2, B)+2.8SK2x5 and C) +2.8SK10x2 topographies replicated in PDMSe. The average normalized, transformed attachment density of Navicula after exposure to shear stress correlated with the contact area between the diatom and a topographically modified surface (R2=0.82) (Figure 48). Attachment Point Theory states that organisms which are provided with larger numbers of contact points by the surface, will have increased attachment strength (Verran and Boyd 2001) The analysis in the current report shows that when diatoms are provided with a larger area of contact, attachment strength increases regardless of the number of attachment points. A C B

PAGE 82

82 Figure 48. The normalized, transformed attachment density of Navicula after exposure to a shear stress of 45 Pa is plotted ver sus the maximum area of contact between the cell and the surface. Diatoms have been shown to be a significant component of microbial biofilms on PDMSbased foulingrelease surfaces (Molino, et al. 2009a) Engineered antifouling topographies were shown to decrease the attachment strength of Semin avis on PDMSe surfaces immediately after attachment and after 6 d of biofilm growth. Immediately after attachment and exposure to shear stress (26 Pa) r emoval of 5, 10 and 15% of the total cells occurred from 3.1SK2x2, smooth and +2.8SK2x2, respectively. Engineered antifouling topographies also reduced the movement of Seminavis across the surface. Only one cell was seen to move on 3.1SK2x2 over the observation period and no cells were observed moving on +2.8SK2x2 while 7% of rapheside down cells moved across the smooth PDMSe standard. The percent of rapheside down Navicula cells moving on +2.8SK2x2 and 3.1SK2x2 was 8% compared to approximately 15% on the smooth standard. Densities of Seminavis biofilms both before and after exposure to shear 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.1 0.2 0 2 4 6 8 10 12 14 16 ln (A/A 0 ) Contact Area (mm 2 ) 10 4 Pillars 2007 ERI Series 2010 n series 2009 s series 2009 Trendline y = 0.057x 0.845 R 2 = 0.82 y = 0.057x 0.845 R 2 = 0.82

PAGE 83

83 stress were lowest on the +2.8SK2x2 topography. Densities of biofilms of Navicula were also lowest on +2.8SK2x2 after exposure to shear stress. Attachment strength of Navicula on engineered antifouling topographies replicated in PDMSe was analyzed in terms of t he Attachment Point Theory. It was shown that the data collected cannot be adequately described using the current Attachment Point Theory, i.e. surfaces with the lowest number of attachment points did not have the lowest diatom density after exposure to s hear stress. The average normalized, transformed attachment density of Navicula after exposure to shear stress correlated with the contact area between the diatom and a topographically modified surface (R2=0.82). The topographical modification of foulingrelease coatings could decrease attachment strength of diatoms and biofilms on these surfaces.

PAGE 84

84 CHAPTER 5 ENGINEERED ANTIFOULI NG MICROTOPOGRAPHIES: THE ROLE OF SURFACE ENERGY OF CROSSLINKE D HYDROGELS IN A MODEL THAT PREDICTS ATTACHMENT OF ZOOSPO RES OF U LVA AND CELLS OF COBETIA MARINA AND NAVICULA INCERTA3 Introduction Biofouling the accumulation of microorganisms, plants and animals on a wetted surface is a widespread problem in the maritime industry. The biofouling process typically begins with conditioning of the surface (Jain and Bhosle 2009) and the development of microbial slime layers containing bacteria, diatoms (unicellular algae) and their extracellular products (Molino and Wetherbee 2008, Molino, et al. 2009a, Molino, et al. 2009b) As these slime layers foul a vessel, hydrodynamic drag and consequently fuel consumption significantly i ncrease (Schultz 2007) Drag and fuel consumption increase further when macrofoulers, including macroalgae and invertebrates, colonize the surface (Schultz 2007) Fouling of ship hulls is also a primary cause for the introduction and spread of nonindigenous marine species worldwide (Otani et al. 2007, Pettengill et al. 2007, Piola and Johnston 2008, Yamaguchi et al. 2009) The green macro alga (seaweed) Ulva is found all over the world and is well known for fouling submerged structures such as ship hulls (Callow et al. 1997) Ulva colonizes substrata by releasing large numbers of motile spores (zoospores) that must select a suitable surface and transition to attached nonmotile spores before germinat ing to produce new plant s. Surface selection is influenced by chemical, physicochemical (Ederth et al. 2009, Schilp, et al. 2009) biological (Joint et 3 Submitted to the American Chemical Society for publication. Magin CM, Finlay JA Clay G, Callow ME, Callow JA, and Brennan AB. 2010. Engineered Antifouling Microtopographies: The role of surface energy of crosslinked hydrogels in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina and Navicula incerta Biomacromolecules. (Submitted).

PAGE 85

85 al. 2002) and topographic (Schu macher, et al. 2007a, Schumacher, et al. 2008) cues. Zoospores of Ulv a the cells of the diatom (unicellular alga) Navicula perminuta and the marine bacterium Cobetia marina are all used in this study as model soft fouling organism s representing three diverse phylogenetic groups, viz. the eukaryotic Plantae ( Ulva ), the eukaryotic Chromista ( Navicula ) and the prokaryotic Bacteria ( Cobetia) (Cavalier Smith 2004) Surface chemistry is an important factor in the adhesion and release of a fouling organism (Magin, et al. 2010a, Rosenhahn et al. 2010) Self assembled monolayers (SAMs) have been widely used to evaluate the influence of surface energy on attachment (Finlay et al. 2002a, Bowen et al. 2007, Schilp, et al. 2009, Zhao et al. 2009, Ista et al. 2010) Experiments have shown that higher numbers of spores of Ulva attach to hydrophobic SAMs versus hydrophilic ones in static assays (Callow et al. 2000) However, when attached s pores were exposed to shear stress in a water channel, the attachment strength on hydrophilic SAMs was greater (Finlay, et al. 2002a) M aterials composed of poly(ethylene glycol)(PEG) and its oligomers exhibit resistance to protein adsorption and have recently been evaluated as coatings for antifouling applications (Ostuni, et al. 2001, Balamurugan, et al. 2005, Schilp et al. 2007, Ekblad, et al. 2008) T he number of zoospores of Ulva and cells of Navicula firmly attached to SAMs of hexa(ethylene glycol) containing alkanethiols with systematically changing endgroup termination increased with decreasing wettabi lity (Schilp, et al. 2007) This increase in attachment has also been cor related with an increase in adsorption of the protein fibrinogen (Schilp, et al. 2009) Monolayers of high molecular weight PEG (MW=2kg/mol, 5kg/mol) SAMs resisted spore attachment (Schilp, et al. 2009) The

PAGE 86

86 mechanism for inhibition of spore attachment by PEG based surfaces is not fully understood or clearly defined in the literature. Spores did not settle (attach) to the PEG monolayers, while on oligoethylene glycol SAMs, high numbers of spores settled and secreted adhesive but they could not bind to the surface so cells were easily washed off by slight disturbance (Schilp, et al. 2009) Hydrogels, i. e., crosslinked polymer networks that swell in the presence of water, have been studied for antifouling applications. Hydrogel surfaces of alginate, chitosan, and polyvinyl alcohol substituted with stilbazolium groups (PVA SbQ) inhibited attachment of cypris larvae of Balanus amphitrite (Rasmussen et al. 2002) and the marine bacterium Pseudomonas sp. NCIMB2021 (Rasm ussen and Ostgaard 2003) Hydrogels based on 2 hydroxyethyl methacrylate (HEMA) reduced fouling in two algal colonization bioassays and remained visually clean in field testing for up to 12 weeks with the addition of benzalkonium chloride, a biocidal com pound (Cowling et al. 2000) Crosslinked poly(ethylene glycol) diacrylate surfaces have been evaluated as proteinresistant coatings. Surfaces that were more hydrophilic, based on contact angle measurements, exhibited less protein adsorption (Ju et al. 2009) A variety of crosslinked hydrogel compositions including poly(HEMA) were shown to reduce adhesion of cyprids of the barnacle Balanus amphitrite (Murosaki et al. 2009) Surface topographies created in polydimethyl siloxane elastomer (PDMSe) have been proposed as a nontoxic strategy for inhibiting the settlement (attachment) of fouling organisms. A bioinspired surface topography, Sharklet AF, reduced attachment of zoos pores of Ulva by 86% compared to smooth (Carman, et al. 2006) An empirical relationship called the Engineered Roughness Index (ERI) has been used

PAGE 87

87 to quantify topographical roughness based on parameters that describe surface energy (Schumacher, et al. 2007b) A correlation between the attachment of spores of Ulva and ERI was demonstrated (Schumacher, et al. 2007b) and a predictive attachment model (Long, et al. 2010) was developed based on a revised version (ERIII). Recently, an extension of the attachment model was reported which correlated the attachment densities of zoospores of Ulva and cells of C. marina with surface roughness by incorporating the Reynolds number of the organism into the m odel (Magin, et al. 2010b) In the present study p oly(ethylene glycol) dimethacrylate (PEGDMA), poly(ethylene glycol) dimethacrylateco glycidyl methacrylate (PEGDMAco GMA), and poly(ethylene glycol) dimethacrylate co hydroxyethyl methacrylate (PEGDMA co HEMA) hydrogels were made with a thermal curing process using ammonium persulfate (APS) and ascorbic acid (AA) as radical initiators (Shin et al. 2003, Pfister, et al. 2007) The PEGDMA co GMA hydrogel composition and a UV curing process were reported previously (Pfister, et al. 2007) The GMA chemistry allows bioactive molecules such as proteins to be covalently grafted to the surface of these hydrogels or for the hydrogel coating to be covalently linked to epoxy paint Funct ionalizing the PEGDMA hydrogel with HEMA creates a material with the same average molecular weight between crosslinks as the PEGDMA co GMA and hence similar mechanical properties while retaining a surface chemistry similar to that of the PEGDMA composition. It was hypothesized that the ability to vary hydrogel composition and surface topography will allow the investigat ion of correlations among surface energy, topography and

PAGE 88

88 attachment of cells of fouling organisms The results extend the attachment model to include substrates other than PDMSe. Experimental Section Materials PEGDMA ( = 1 kg mol1) a tetrafunctional polyethylene glycol macromonomer was purchased from Polysciences Inc. (Warrington, PA). 2 hydroxyethyl methacrylate 98% stabilized was purchased from Acros Organics (Geel, Belgium). Glycidyl methacrylate >97%, ascorbic acid (AA) 99+%, and ammonium persulfate were purchased from SigmaAldrich (Milwaukee, WI). Methacryloxypropyltriethoxysilane was purchased from Gelest Inc. ( Morrisville, PA). Ultra pure water was produced by a Barnstead Nanopure Ultra Pure Water System ( Waltham, MA ). The base material for standards was a platinum catalyzed PDMSe (Silastic T2; Dow Corning Corporation). Sample P reparation PEGDMA, PEGDMAco GMA, and PEGDMA co HEMA hydrogels were produced using a thermally activated polymerization. Aqueous solutions were prepared by combining 25 wt% PEGDMA ( = 1 kg mol1) used as is, 0.5 wt % ammonium persulfate and ascorbic acid as chemical initiators and ultra pure water to balance. To create a functionalized PEGDMA hydrogel 5 wt% of GMA or HEMA was added to the aqueous solution (Figure 51) The hydrogels were either produced as free standing films or coatings attached to 76 x 22 mm silanated microscope glass sli des Glass slides were pretreated with 0.5% methacryloxypropyltriethoxysilane (MPS) in a 95% ethanol/water solution for 10 min, rinsed thoroughly with 95% ethanol, and dried at 120C for 15 min.

PAGE 89

89 Figure 5 1. Chemical structures of monomers used to produce functionalized hydrogels. All prepolymer compounds were combined in a glass beaker and stirred until a solution was achieved, i.e., the PEGDMA was dissolved. The prepolymer solution was then poured into two centrifuge tubes and centrifuged for 10 min at 3300 RPM. The centrifuged prepolymer solution was pipette into a mold (Figure 5 2). The mold contained a PDMSe gasket with an opening (2.5 cm x 7.6 cm x 2 mm) for a pretreated glass slide. The gasket was placed on top of a glass plate (12.7 cm x 12.7 cm x 0.32 cm) and the pretreated slide was fitted into the opening in the gasket. A microtopographically modified silicon wafer was placed on top of the PDMSe spacer, with the topography facing down, to create engineered microtopographies. Smooth samples were cast against a second glass plate. The mold was assembled by adding a second glass plate on the back of the silicon wafer and clamping with 32 binder clips. The entire assembly was heated to 45C for 45 min (Shin, et al. 2003) Hydrogel coated slides were removed from the assembly by peeling. Two topographies, continuous channels 2.6 m tall, 2 m wide and spaced by 2 m (+2.6CH2x2) and the Sharklet AF pattern 2.8 m tall, 2 m wide and spaced by 2 m (+2.8SK2x2) (Schumacher, et al. 2007b) were created with this process. To create PDMSe smooth st andards and topographically modified surfaces the elastomer was prepared by mixing 10 parts by weight of resin and 1 part by weight

PAGE 90

90 curing agent. The mixture was stirred by hand for 5 min and degassed under vacuum (2830 in Hg) for 30 min to remove bubbles. An allyltrimethoxysilanecoupling agent was applied to clean glass microscope slides (0.5 wt% in 95% ethanol/water solution) and heated for 10 min at 120 C. The Silastic T2 was then placed in contact with the treated slides in a mold consisting of t wo glass plates and aluminum spacers. The elastomer was polymerized at ambient for 24 h. Topographically modified PDMSe samples were prepared in a twostep casting process previously described (Carman, et al. 2006) Figure 5 2. Schematic of mold for hydrogel production. Microtopography Characterization The +2.8SK2x2 topography (mold dimensions) was replicated as free standing films in PEGDMAco GMA and PDMSe. Hydrogel samples were immersed in deionized water for 24 hr prior to characterization. Six height measurements were made at random locations on each sample using a Wyko model NT1000 white light optical profilometer. Hydrogel samples in deionized water were prepared for scanning electron microscopy (SEM) by flash freezing with liquid nitrogen and subsequently freeze drying for 5 d at 9x103 mbar and 42C (Lyph Lock 4.5 Freeze Dry System, Labc onco, Kansas PDMSe Gasket Silicon Wafer Glass Plates

PAGE 91

91 City, MO). Freeze dried hydrogel samples and PDMS e samples were mounted onto aluminum SEM stubs with double sided tape. These samples were sputter coated with goldpalladium for 60 s at 38 mA. Samples were imaged with a JEOL JSM 6400 SEM with a tungsten filament at an accelerating voltage of 15 kV. Feature dimensions were measured using Image J software. Chemical Composition Free standing films of PEGDMA, PEGDMAco GMA, and PEGDMAco HEMA were produced and air dried for 48 h. The attenuated total reflectance Fourier transform infrared (ATR FTIR) spect rum of each film was recorded on a PerkinElmer Spectrum One spectrometer. Spectra were obtained with a ZnSe crystal with an angle of incidence of 60 and resolution of 4 cm1. Twenty scans were performed for each sample. Spectral subtraction was performed to verify composition of the various hydrogel formulations. Surface Energy Measurements Captive air and oil bubble contact angles were measured to calculate the surface energy of the functionalized hydrogels (Figure 5 3) Two replicates of each hydrogel were cast onto glass slides. Both sides of five captive air bubbles and five captive noctane bubbles were measured on each surface (n = 20). Surface energies were calculated using the method previously presented by Hu (Hu and Tsai 1996) Statistical differences between surfaces were evaluated using oneway analysis of variance (ANOVA) and Tukey s test fo r multiple comparisons ( = 0.05) The surface energy of smooth PDMSe was measured using the Owens Wendt Kaelble approach (Owens and Wendt 1969) Static contact angles of water, glycerol and diiodomethane were measured. Both sides of five drops were measured on each of two

PAGE 92

92 replicates (n=20). The polar and dispersive components of the surface energy were calculated using two pairs of polar and nonpolar liquids: water and diiodomethane and glycerol and diiodomethane. The contact angles were reduced to surface energies by solving simultaneous equations and averaging the results (Owens and Wendt 1969) The mean and standard deviation for the contact angles of each liquid were determined. However, si nce the measured values are related to each other by a system of simultaneous equations, it was not trivial to account for the covariance using a small sample size of matching pairs. Standard deviations were calculated by using Minitab Statistical Softwar e to randomly generate 1500unit sample groups with the same mean and standard deviation as the measured sample groups. The systems of equations were solved multiple times with these data and the results were used for statistical analysis. Figure 5 3. Schematic of captive air and oil bubble measurements for calculating surface energy. Biological Attachment Assays Coated glass slides were shipped to the Universit y of Birmingham overnight in 50 mL conical centrifuge tubes filled with deionized water. Prior to bioassay, the slides were transferred to sterile (0.22 m filtered) artificial seawater (ASW) (Tropic MarinTM) for 2 h Oil wvsw Airsv owswso 12Water Water Substrate Substrate

PAGE 93

93 Ulva A total of four Ulva attachment assays were performed. Six replicates of two topographies, +2.6CH2x2 and +2.8SK2x2, and smoot h created in PEGDMA co HEMA and PDMSe were attached to glass slides and provided for analysis in the fourth assay. Z oospores were obtained from fertile plants of Ulva linza collected from Llantwit Major (Wales) and prepared for attachment assays as descri bed previously (Callow, et al. 1997) Briefly, each sample was immersed in 10 mL of a spore suspension containing 1.5 x 106 spores mL1 and incubated in the dark for 45 min Attached spores on hydrogel and PDMSe slides were counted using a Zeiss epifluorescence microscope with a 10x objective while the samples were still wet. Thirty counts were taken fro m each of the three replicates. Navicula incerta Cells of Navicula incerta were cultured in F/2 medium contained in 250 mL conical flasks until cells reached the logarithmic growth phase, approximately 3 d. Cells were washed 3 times in fresh medium before harvesting and diluting to give a suspension with a chlorophyll a cont ent of ap proximately 0.25 mL1 (Holland, et al. 2004) Six replicates of each hydrogel composition and PDMSe attached to glass slides were placed in Quadriperm dishes to which 10 mL of the diatom suspension were added. Cells were allowed to attach at ambient ( ~20 C ) on laboratory benches for 2 h. Samples were exposed to a submerged wash in seawater to remove cells which had not attached (the underwater imme rsion process avoided passing the samples through the air water interface). Three replicates were counted wet using an image analysis system attached to a fluorescence microscope. Counts were made for 30 fields of view (0.064 mm2) on each

PAGE 94

94 sample. The rem aining three replicate samples were exposed to a shear stress of 45 Pa in a water channel (Schultz et al. 2000) The number of cells remaining attached was counted using the image analysis method described above. C. marina C ultures of C. marina (ATCC 25374) (Baumann et al. 1983) were grown in marine broth contained in 100 mL conical flasks, at 18C on an orbital shaker at 60 rpm overnight Cells were harvested by centrifugation (8000 rpm for 1 min) and washed 2 times in sterile (0.22 m filtered ) Tropic MarinTM ASW to remove any residual marine broth. The cells were resuspended in sterile ASW and briefly sonicated to aid dispersion. The suspension was diluted to an absorbance of 0.3 at 600 nm. Six replicates of each hydrogel composition and PDM Se attached to glass slides were placed in Quadriperm dishes to which 10 mL of the suspended bacteria were added. The dishes were incubated at ambient ( ~20C ) on the laboratory bench for 2 h. After incubation, the slides were washed gently in seawater to remove unattac hed bacteria. Three replicates w ere stained with crystal violet (0.01% in seawater) and counted under a 20x objective while still wet. 2) on each sample. The remaining three replicates with at tached bacteria were exposed to a shear stress of 50 Pa in a water channel (Schultz, et al. 2000) The number of cells remaining attached was counted as described above. The cell density per mm2 was calculated for each count (n = 90). The mean cell densities were compared using oneway analysis of variance (ANOVA) and Tukeys test for multiple comparisons.

PAGE 95

95 Results Microtopography Characterization Fidelity of topographic features replicated in PEGDMA co GMA and PDMSe was evaluated with SEM and white light optical profilometry (Figures 5 4 and 5 5). All feature heights were within 0.3 m of the mold dimensions (Table 5 1). Figure 5 4 Scanning electron micrographs of A)+0.6CH2x2 and B)+1SK2x2 topographies in PEGDMA co GMA hydrogel. Figure 5 5. White light optical profilom etry image of +1SK2x2 topography cast in PEGDMAco GMA. Table 5 1. Average heights plus or minus standard deviation of hydrogel and PDMSe replicates of wafers measured with white light optical profilometry and SEM. Feature Height (m) PEGDMA co GMA P DMSe Topography Profilometer SEM Profilometer SEM +3SK2x2 3.12 0.1 2.83 0.5 -3.24 0.1 +3SK2x2 ----3.2 0.1 3.17 0.1

PAGE 96

96 Chemical Composition Spectral subtraction of PEGDMA from PEGDMA co HEMA shows characteristic peaks for HEMA, i.e., 3400 to 3200 cm1 (OH stretch), 2863 to 2843 cm1 (CH symmetric stretch), 1750 to 1735 cm1 (C=O stretch), 1485 to 1445 cm1 (CH asymmetric deformation) and 1150 to 1060 cm1 (C O C asymmetric stretch). Spectral subtraction of PEGDMA from PEGDMA co GMA sh ows characteristic bands for GMA at 1715 to 1740 cm1 (C=O stretch), 1485 to 1445 cm1 (CH deformation), 1280 to 1230 cm1 (C O C symmetric stretch) and 950 to 815 cm1 (asymmetric stretch). Surface Energy Measurements The surface energies of PEGDMA and PE GDMAco HEMA calculated from measured captive air and oil bubble contact angles were not statistically different ( =0.05, p=0.11). The surface energy of PEGDMA co GMA was statistically higher than both PEGDMA and PEGDMA co HEMA ( =0.05, p<0). The surface energies of all three hydrogels are higher than that of PDMSe (Tables 5 2 and 5 3). Table 5 2. Contact angle measurements and calculated surface energies for smooth hydrogels. Surface energies are the interfacial interaction energy (Isw), the sw), the polar component of the p sv), the dispers ion component of surface d svsv). Contact Angle () Energy (mN/m) Captive Air Bubble Captive n -Octane Bubble I sw sw sv P sv d sv PEGMDA 48 5 47 11 85 7 3 1 35 6 17 9 52 5 P EGDMA co GMA 39 2 51 5 83 3 2 1 34 3 25 5 58 2* PEGDMA-co -HEMA 46 3 48 4 85 3 2 10 35 2 17 2 52 2

PAGE 97

97 Table 5 3. Contact angle measurements and calculated surface energies ( s) for smooth PDMSe. Contact Angle () Energy (mN/m) Water Glycerol Diiodomethane S PDMSe 109 5 96 1 65 2 19 3 Biological Attachment Assays Ulva a ttachment Results of 3 separate assays showed that PEGDMA, PEGDMA co GMA and PEGDMAco HEM A consistently reduced the attachment of spores of Ulva compared to smooth PDMSe (Figure 5 6). The t otal average percent reduction for PEGDMA versus PDMSe is 55%, PEGDMAco GMA versus PDMSe is 87% and for PEGDMAco HEMA versus PDMSe it is 85% Figure 5 6. Functionalized poly(ethylene glycol) based hydrogels reduce the attachment density of zoospores of Ulva These data are representative of three separate assays. Error bars indicate 95% confidence intervals. 0 100 200 300 400 500 600 PDMSe PEGDMA PEGDMAcoGMA PEGDMAcoHEMA Cell Density (spores/mm 2 ) Ulva

PAGE 98

98 Attachment of cells of Navicula Figure 5 7. Functionalized poly(ethylene glycol) based hydrogels reduce attachment density and attachment strength of Navicula Error bars indicate 95% confidence intervals. Diatom cells, unlike spores of Ulva are not motile in the water column. The cells co me into contact with a surface by gravity and water currents so at the end of a 2 h incubation period, approximately the same number of diatoms will be in contact with all test surfaces. Differences in the density of attached cells of Navicula were quanti fied following a gentle underwater washing, which washed away cells that were not attached to the surface. The initial attachment density was lowest on PEGDMA co GMA which was significantly lower than initial attachment densities on PEGDMA, PDMSe and PEG DMAco HEMA ( = 0.05, p<0) (Figure 5 7). Initial attachment densities on PDMSe and PEGDMAco HEMA were not statistically different ( = 0.05). Exposure to a shear stress of 45 Pa in the water channel caused removal of 77% or more of diatom cells to be removed from all hydrogel surfaces (Figure 5 7). No cells were removed 0 200 400 600 800 PDMSe PEGDMA PEGDMAcoGMA PEGDMAcoHEMA Cell density (cells/mm 2 ) Initial Attachment Density Density after Exposure to 45Pa Shear Stress 89 % removal 88 % removal 77 % removal Navicula

PAGE 99

99 from the smooth PDMSe surface. The t otal percent reduction after removal for PEGDMAco GMA versus PDMSe was 95%. Attachment of cells of C. marina Figure 5 8. Functionalized poly(ethylene glycol ) based hydrogels reduce attachment density and attachment strength of C. marina Error bars indicate 95% confidence intervals. The initial attachment density of cells of C. marina was reduced on the hydrogels compared to a smooth PDMSe standard (up to 62%) with lowest densities on PEGDMAco GMA and PEGDMAco HEMA (Figure 5 8). Initial attachment densities of C. marina were not statistically different among the three hydrogel compositions but all hydrogels significantly reduced attachment versus PDMSe ( = 0.05, p < 0). Exposure to a 50 Pa shear stress in a water channel caused 44% and 45% removal from PEGDMAco GMA and PEGDMAco HEMA, respectively (Figure 5 8). There was no statistically significant removal of C. marina from PDMSe or PEGDMA. The cell density on PEGDMAco HEMA was 77% less, after exposure to a 50 Pa shear stress, than that on PDMSe. 0 1000 2000 3000 4000 5000 6000 PDMSe PEGDMA PEGDMAcoGMA PEGDMAcoHEMA Cell Density (cells/mm 2 ) Initial Attachment Density Density after Exposure to 50Pa Shear Stress C. marina45 % removal 44 % removal

PAGE 100

100 Based on the results from all biological attachment assays, PEGDMA co HEMA was selected as the substrate for further testing. The channels (+2.6CH2x2) and Sharklet AF (+2.8SK2x2) topographies were replicated in PEGDMA co HEMA and PDMSe and tested with the standard Ulva zoospore attachment assay. The initial spore a ttachment density was reduc ed on both PDMSe +2.6CH2x2 and +2.8SK2x2 versus smooth. Smooth PE GDMA co HEMA reduced spore attachment by an average of 75% compared to smooth PDMSe. Topographies produced in PEGDMA co HEMA reduced Ulva attachment by an average of 82% for +2.6CH2x2 and 93% for +2.8SK2x2 compared to smooth PDMSe. Discussion Hydrogel Characterization White light optical profilometry and SEM were used to confirm that topographies were replicated in hydrogel with high fidelity. Spectral subtraction of ATR FTIR spectra verified the presence of GMA and HEMA in functionalized PEGDMA hydrogels Contact angle measurements and surface energy calculations showed that the surface energy of PEGDMA co GMA was significantly higher than both PEGDMA and PEGDMAco HEMA (Tukey Test =0.05). The surface energies of PEGDMA and PEGDMAco HEMA were not sign ificantly statistically different. Biological Attachment Assays Initial attachment density and attachment strength of marine fouling organisms have been attributed to many factors including surface chemistry (Gudipati et al. 2005, Cordiero et al. 2009, Ederth, et al. 2009, Schilp, et al. 2009, Zhang et al. 2009) and surface topography (Callow, et al. 2002, Carman, et al. 2006, Schumacher, et al. 2008, Scardino, et al. 2009a, Scardino et al. 2009b, Aldred et al. 2010, Bers et al. 2010, Long,

PAGE 101

101 et al. 2010, Magin, et al. 2010b) The surface energy of all three PEGDMA based hydrogels was more than twice that of the PDMSe standard in this report. Hydrogels reduced the initial att achment densities of all three marine organisms compared to the PDMSe standard. These findings are consistent with results from other PEG based materials that have been evaluated (Ekblad, et al. 2008, Krishnan et al. 2008, Schilp, et al. 2009) The low number of Ulva spores removed from the hydrogel surfaces was also expected based on the observation that spores generally attach more firmly to hydrophilic than hydrophobic surfaces (Finlay et al. 2008) All hydrogel compositions reduced the attachment strength of cells of the diatom Navicula compared to smooth PDMSe. Diatoms, unlike Ulva have been found to adhere more firmly to hydrophobic silicone elasto mers (Holland, et al. 2004) and self assembled monolayers (SAMs) (Finlay, et al. 2002b) than hydrophilic surfaces. Initial attachment of cells the diatom Navicula on a series of xerogel surfaces was comparable on all surfaces, but the percentage removal of attached cells by hydrodynamic shear stress increased with critical surface tension and increased wettability as measured by the static water contact angle (Finlay, et al. 2010) The functionalized compositions of PEGDMA co GMA and PEGDMAco HEMA reduced attachment strength of C. marina compared to PDMSe. The functionalized composi tions of PEGDMAco GMA and PEGDMAco HEMA reduced attachment strength of C. marina compared to PDMSe. The results are consistent with the Baier curve, a plot which demonstrates the relationship between substratum surface tension and the degree of biological fouling retention (Baier 2006) There are two minima observed on the Baier curve, one between 20 and 30 mN/m and another between 50

PAGE 102

102 and 70 mN/m. The surface tensions of PDMSe and the hydrogel substrates were wit hin these ranges and both exhibited low retention of fouling of specific organisms. Based on these results, PEGDMA co GMA and PEGDMAco HEMA, in particular, are more effective foulingrelease coatings than PDMSe for cells of Navicula and C. marina The normalized, transformed attachment densities of spores of Ulva and cells of C. marina to various topographies created in PDMSe has been correlated with an attachment model (Magin, et al. 2010b) This equation relates the attachment density on a particular topography to the surface energy of the substratum and the Reynolds number of the fouling organism (Equations 5 1 and 5 2). ln 0 = ( 5 1) = ( 5 2) It w as shown that attachment density on a variety of topographies could be predicted with this model (Long, et al. 2010) The attachment densities of Ulva on PDMSe and PEGDMAco HEMA topographies correlated with Equations 5 3 and 5 4, re spectively when plotted in the form of the attachment model (Figure 5 9). ln 0 = 0 .46 10 2 ( 5 3) ln 0 = 0 .15 10 2 ( 5 4) The discrepancy in the slopes of the attachment model was attributed to surface energies of the materials. The effect was incorporated into the attac hment model by multiplying by the surface energy of the smooth substrate for each material normalized by the surface energy of PDMSe. The normalized, transformed attachment densities on

PAGE 103

103 both PDMSe and PEGDMA co HEMA correlated (R2 = 0.87) with the attachm ent model multiplied by the measured surface energy ratio ( 0) (Figure 510). Figure 5 9. Normalized, transformed Ulva attachment density on PDMSe and PEGDMAco HEMA topographies plotted versus ERIII* Re* 102. Attachment density on PDMSe and PEGDMA co HEMA correlated well with the equation that describes the attachment model (R2 = 0.87 and R2 = 0.62, respectively). The new slope of the attachment model and the negative linear correlation are described by Equations 5 5 and 5 6. = 0 ( 5 5) ln 0 = ( 0 .46 10 2 0 ) ( 5 6) The normalized, transformed attachment densities of eight assays with spores of Ulva and two assays with C. marina on various engineered microtopographies created in PDMSe and PEGDMAco HEMA correlated to the ERIII with a new slope that consists of the Re of the organisms multiplied by a ratio of the surface energy measured on a smooth substrate to that of a standard (PDMSe) (R2=0.80) (Figure 511). y = 0.15x R = 0.62 y = 0.46x R = 0.87 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.5 1 1.5 2 ln(A/A 0 ) ERI II Re 10 2 PEGDMA co HEMA PDMSe Trendline (PEGDMAcoHEMA) Trendline (PDMSe) Channels +3CH2x2 Sharklet AFTM+2.8SK2x2 Smooth Sharklet AFTM+2.8SK2x2

PAGE 104

1 04 Figure 5 10. Normalized, transformed spore attachment density on PDMSe and PEGDMAco HEMA topographies plotted versus ERIII* Re* 102. Attachment density on PDMSe and PEGDMA co HEMA show a negative, linear correlation with the attachment model multiplied 0 (R2=0.88). Figure 5 11. Normalized, transformed attachment densities of spores of Ulva (Schumacher, et al. 2007b, Schumacher, et al. 2008, Long, et al. 2010) and cells of C. marina (Magin, et al. 2010b) on PDMSe and PEGDMAco HEMA topographies plotted versus ERIII*102. Attachment densities on PDMS e and PEGDMAco HEMA for eight different assays show a negative, linear correlation with ERIII* Re* 0 102 (R2 = 0.80). 1.2 1 0.8 0.6 0.4 0.2 0 0 0.5 1 1.5 2 ln(A/A 0 ) ERI II Re 0 / 10 2 PDMSe PEGDMA co HEMA Trendline y = 0.46x R 2 = 0.88 12 10 8 6 4 2 0 0 2 4 6 8 10 12 14 ln(A/A 0 ) ERI II Re 0 / 10 2 y = 0.80x R 2 = 0.80 Ulva 2007 Ulva 2008 Ulva 2009 2.0SK2x2 Ulva 2009 n Series C. marina Stationary Phase C. marina Log Growth Phase Ulva PDMSe 2010 Ulva PEGDMA co HEMA 2010

PAGE 105

105 The environmental and economic costs of biofouling have led to a need for environmentally neutral antifouling technologies (Magin, et al. 2010a) The results from attachment studies performed with three fouling organisms on functionalized PEGDMA hydrogels provide insight that will lead to improvements in antifouling and fouling release technologies. The role of the slope of the attachment model as an indicator of an organisms sensitivity to a surface was extended by incorporating a ratio of the measured surface energy of a smooth substratum to that of a standard (PDMSe) into the attachment model. The attachment model has now been shown to correlate with the attachment density of cells from two evolutionarily diverse groups on two substrate materials. This equation has been successfully used to model attachment density of zoospores and cells on micropatterned materials with different surface chemistries. Functionalized, crosslinked hydrogels reduced attachment of fouling organism s from three evolutionarily diverse groups.

PAGE 106

106 CHAPTER 6 ENGINEERED MICROTOPO GRAPHIES AND CHEMIST RIES INFLUENCE ATTACHMENT AND FUNCT ION OF ENDOTHELIAL AND SMOOTH MUSCLE CEL LS Background Coronary heart disease is the leading cause of death in the United States (U.S). According to the latest mortality data, every 38 s an American will die from a coronary event (Heart Disease & Stroke Statistics 2010) Coronary heart disease is a result of atherosclerosis, the narrowing and hardening of artery walls due to a buildup of fatty substances including cholesterol in the arteries that supply blood to the heart muscle. This buildup can bloc k the blood flow and oxygen supply to the heart ultimately resulting in a heart attack Angioplasty and stenting have been used to increase lumen size and restore blood flow in smalldiameter vessels such as the coronary arteries, but not without complica tions. R enar rowing of the blood vessels referred to as in stent restenosis occurs within 6 months in 50% of patients that undergo the procedure (Padera and Schoen 2004) Clot formation or thrombosis begins immediately after circulation is restored to the grafted area; plasma proteins adsorb to the surface of the graft, platelets and leukocytes bind to these proteins, and bulk fibrin formation begins. In areas of low blood flow such as small diameter vessels, the fiber forming step of the clotting process occurs and resul ts in a macroscopic thrombus that reduces blood flow (Biomaterials Science: An Introduction to Materials in Medicine 2004) The other common mode of failure for small diameter vascular grafts is intim al hyperplasia, thickening of the lumen caused by excessive proliferation of smooth muscle cells in the media. G rafting is the most common way to treat coronary blockages Autogenous veins or arteries are typically used to reroute blood flow around the bl ocked arteries in the

PAGE 107

107 grafting procedure. The patients internal mammary artery or saphenous vein are the most common grafts; however, damage to the native vasculature or a previous grafting procedure may make bypass surgery impossible. Even when healthy autogenous vessels are available, it is less desirable to remove them from their positions than it would be to insert a prosthetic graft (Nerem and Seliktar 2001) However, there are currently no FDA approved smalldiameter vascular grafts. Over the past 20 years researchers have worked diligently to increase patency of smalldiameter vascular grafts. The first efforts were directed toward coating synthetic grafts e.g. Dacron, a woven poly(ethylene terephthalate) (PET) material and Goretex, expanded poly(tetrafluoroethylene) (ePTFE), with proteins or blood components to reduce blood/material interactions (Rumisek et al. 1986, Drury et al. 1987, Freischlag and Moore 1990) Whe n these attempts did not significantly increase patency, investigators turned to a tissue engineering approach. Endothelial cells were seeded onto synthetic graft materials to mimic the nonthrombogenic interface that already ex ists within the blood vessels (Nerem and Seliktar 2001) Since the semi nal work of Weinberg and Bell several distinct approaches to vascular tissue engineering have emerged (Weinberg and Bell 1986) (see Appendix B ): 1) solid scaffold (made of synthetic and/or natural materials) and cell seeding in vitro or in vivo, 2) embedding of cells in a matrix, 3) assembly of cell sheets, and 4) decell urarized tissue constructs. Current small diameter vascular graft designs focus on recreating the natural extracellular microenvironment. Natural extracellular matrix (ECM) is gel made up of various protein fibers woven together within a hydrated network of glycosaminoglycan chains (Lutolf and Hubbell 2005) Gels made from

PAGE 108

108 collagen, fibrin, and various combinations of these and other proteins are the closest biologically to ECM (Cummings et al. 2004, Chan et al. 2007, Stegemann, et al. 2007, Wu et al. 2007) however, these materials often lack the mechanical st ability required in an arterial graft. In vascular tissue engineering, synthetic hydrogels are used as scaffolds to mimic ECM (Nguyen and West 2002, Zhu et al. 2005, Shen et al. 2006) as matrix material in which cells are embedded (Almany and Seliktar 2005) and for microencapsulation of growth f actors or other proteins (Lee et al. 2004, Patel et al. 2007) Biodegradable materials such as polylactic acid and polycaprolactone processed by a variety of methods including gel spinning (Chung et al. 2007a) layer by layer construction (Choi et al. 2007, Feng et al. 2007, Zhang et al. 2007) and nonwoven techniques (Roh et al. 2008) are commonly used to create vascular grafts that will allow cellular ingrowth and remodeling in vivo. The technique of electrospinning is a very popular way to mimic the nanofibrillar structure of the ECM. Electrospun vascular prostheses have been made from synthetic elastomers, biodegradable materials, and/or ECM proteins such as collagen and elastin (Stankus et al. 2006, Lee et al. 2008b, Sell and Bowlin 2008, Smith et al. 2008) These structures provide topographical cues to cells along with better mechanical properti es than collagen gel matrices. To obtain ultimate biological and mechanical properties researc hers use decellularized ECM from various sources including small intestine submucosa and the vasculature of other species (Badylak 2002, Wang et al. 2007, Zhu et al. 2008) A few investigators have created complet ely tissue engineered grafts by culturing autologous cells outside of the patient (L'Heureux et al. 2006) or implanting a biodegradable scaffold into the patient and allow ing complete cellular remodeling to occur before

PAGE 109

109 grafting (Watanabe et al. 2007) Disadvantages to completely tissue engineered grafts are numerous Cells must be retrieved from the patient, sometimes surgically and then expanded for several weeks in vitro before the graft can be implanted. This process is ver y time consuming and expensive. Biology of the Vascular Wall All blood vessels are made up of three layers: intima, media, and adventitia. The intima is the innermost, subluminal layer composed of the endothelium, basement membrane, and the internal elastic lamina. The endothelium consists of a continuous monolayer of flat endothelial cells (ECs) with a cobblestonelike morphology and extracellular matrix elements. The basement membrane or basal lamina creates a flexible substrate of collagen and glycoproteins for EC attachment. The internal elastic lamina is a layer of circularly arranged elastic fibers that are interlaced with the connective tissue of the basal lamina. These structures form a semi permeable interface between circulating blood and the rest of the body. The media, the thickest layer, is a circular arrangement of elastic fi bers, connective tissue, and smooth muscle cells. It controls the caliber of the blood vessel while providing elasticity. The adventitia is made up mainly of connective tissue. It also contains fibroblasts, nerves, and nutrient capillaries (Biology of the Arterial Wall 1999) Endothelial cells line the luminal surface of all bloodcontacting vessels arteries, veins, and capillaries. This continuous monolayer of cells creates a n ideal anti thrombogenic surface. ECs possess a glycoprotein coat that is negatively charged and repels platelet and leukocyte adhesion. ECs also secrete bioactive substances that inhibit thrombosis, promote fibrinolysis, and inhibit smooth muscle cell (SMC) proliferation (Biology of the Arterial Wall 1999, Xue and Greisler 2000) Excessive

PAGE 110

110 SMC proliferation, commonly referred to as intimal hyperplasia, is a significant contributor to vascul ar graft failure. During intimal hyperplasia SMCs in the arteries dedifferentiate from the contractile to the synthetic phenotype and proliferate uncontrollably. Theoretically, coating a small diameter vascular graft with a confluent layer of ECs would r esult in a nonthrombogenic bloodmaterial interface that has the ability to reduce intimal hyperplasia (Biology of the Arterial Wall 1999, Xue and Greisler 2000, Nerem and Seliktar 2001) Controlling SMC phenotype could also reduce intimal hyperplasia. In situ endothelialization of synthetic prostheses occurs by direct migration of ECs from anastomotic edges, by transmural migration of ECs and by transformation of ECs from EPCs (Sarkar et al. 2006b) Unfortunately, ECs have a limited capacity for regeneration and complete reendothelialization has never been shown in clinical practice (Xue and Greisler 2000, Sarkar, et al. 2006b) Despite successful reendothelialization in animal models, it has been shown that transanastomotic endothelial ingrowth does not exceed more than 1 to 2 cm in humans even after years of implantation (Nerem and Seliktar 2001, Zilla et al. 2007) The regenerative nature of EPCs was first discovered when Asahara, et al (Asahara, et al. 1997) published the first description of isolation of E PCs from human peripheral blood. These cells originate in bone marrow and have the ability to proliferate and to differentiate into mature ECs. Endothelial progenitor cells could be the ideal way to endothel ialize a synthetic graft; the cells are autologous and do not require removal from the patient or extensive culture times ex vivo Topography A key contributor to the cellular microenvironment is substratum topography. For nearly a century it has been known that substratum topography influences cell

PAGE 111

111 morphology (Harrison 1914) However, to date very little has been reported about the longterm cellular response to topography such as gene expression, proliferation, and differentiation. It has been shown that mRNA expression and protein synthesis change when cell shape is controlled through growt h on micropatter ned adhesive islands. These outcomes were accompanied by a decreased proliferation time before differentiation (Thomas et al. 2001) Fibroblasts grown on nanoscale topographies showed broad gene upregulation especially in those that influence cell signaling, proliferation, cytoskeleton, and production of ECM (Dalby, et al. 2002, Dalby et al. 2003) In 2007, it was demonstrated that topography could enhance differentiation of human mesenchymal stem cells into the neuronal lineage (Yim, et al. 2007) It has also been shown that isotropic topographies such as pillars tend to control more collective cell functions such as proliferation, while anisotropic topographies like channels more noticeably alter cell morphology and cytoskeletal organization (Lim and Donahue 2007) Cell morphology is known to affect proliferation, differentiation, cytoskeletal organizat ion and gene expression (Thomas, et al. 2001, Dalby, et al. 2002, Dalby, et al. 2003, Itano et al. 2003, Lim and Donahue 2007) Itano et al (Itano, et al. 2003) demonstrated that changes in nuclear shape led to the release of nuclear Ca2+, which is known to regulate gene expression in cells. Microchannel scaffolds with discontinuous walls initia lly supported primary vascular smooth muscle cell proliferation in the synthetic phenotype. Upon reaching confluence, the cells were aligned and transformed towards the contractile phenotype (Cao et al. 2010) This evidence supports the hypothesis that mechanical forces such as those experienced by cells grown on topographies can regulate gene expression.

PAGE 112

112 Approach The ideal vasc ular graft should provide sufficient mechanical strength to withstand arterial pressures while matching the compliance of the native artery (E 600100 kPa) (Holzapfel et al. 2002) For clinical applications, grafts should be readily available in various shapes and sizes, inexpensive, and easy for a surgeon to handle and insert flexible and kink resistant with good suture retention (Lee et al. 2008a) For this particular appl ication the graft material must provide mechanical support, good nutrient transport, and support cellular ingrowth. The graft must also create a microenvironment that induces proliferation and differentiation of EPCs to ECs. For this reason, the graft ma terial should mimic the natural ECM, provide sites for grafting bioactive molecules that recruit EPCs and maintain good fidelity of topographical features. Since natural ECM is a combination of hydrated glycosaminoglycan chains and filamentous proteins, synthetic hydrogels have often been used to mimic the extracellular environment (Lutolf and Hubbell 2005) In this case it is hypothesized that a PEGDMAbased hydrogel elastomer will provide good compliance, high fidelity of topographic features, and sites for surface modified with biomolecules. Predefined, engineered surface microtopographies have been created in PEGDMA hydrogels without significant size variations due to swelling (Pfister, et al. 2007) These hydrogels have also been reported to have a highly adjustable shear modulus wit h a range of G=10kPa to 1Mpa (Pfister, et al. 2007). S ubstratum elasticity has been shown to direct stem cell differentiation into specific lineages; therefore the ability to fine tune the elastic modul us of the graft will be advantageous when inducing differentiation of EPCs (Engler, et al. 2006)

PAGE 113

113 Since the discovery of circulating EPCs and their regenerative properties, researchers have been investigating ways to increase EPC recruitment and attachment to synthetic stents and grafts. The most widely used approach is to capture EPCs with an antibody against the surface marker CD34. Several clinical trials performed with anti CD34 coated stents have concluded that these stents are a safe, feasible way to increase EPC attachment and reduce instent thrombosis (Aoki et al. 2005, Co et al. 2008, Miglionico et al. 2008) Even though cell captur e was reported, very little has been demonstrated about capture kinetics and the final phenotype of captured cells (Markway, et al. 2008) Recently, human umbilical vein endothelial cells (HUVECs) wer e selectively captured from a flowing, heterogeneous cell pop ulation using anti kinase insert domain receptor (anti KDR) at a density of approximately 55cells/mm (Markway, et al. 2008) This cell density should be sufficient to produce a monolayer of ECs; it is slightly larger than the EC seeding density (4 x 103 cells cm-) previously used to culture confluent monolayers of ECs (Carman, et al. 2006) Peptides derived from fibronectin such as connecting segment 1 (CS 1) (Rodenberg and Pavalko 2007) and arginineglycineaspartic acid (RGD) (Blindt et al. 2006) have been shown to capture circulating EPCs. P eptides that specifically bind a form of EPCs called human blood outgrowth endothelial cells (HBOECs) with high affinity were selected with phage display technology (Veleva et al. 2007) Howev er, after incorporating these ligands into a polymeric scaffold, testing showed that specific binding was greatly reduced in the presence of serum proteins (Veleva, et al. 2007) The aim of this work is to create a cell culture substrate for smalldiameter vascular graft applications that has the potential to reendothelialize in vivo and/or

PAGE 114

114 control SMC phenotype to reduce neointimal hyperplasia and thrombosis. It was hypothesized that a combination of surface chemistry and topography on a graft surface would capture circulating EPCs in the peripheral blood and promote their differentiation into ECs to create a continuous tissue layer within the lumen of the graft. A biomolecular surface chemistry such as a bioactive peptide or antibody would increase recruitment of endothelial progenitor cells and topographical modification would provide mechanical cues to induce differ entiation into the endothelial cell phenotype. The substratum material would provide sufficient mechanical strength and compliance while allowing nutrients to flow to the cells AntiCD34 capture methods have demonstrated increased EPC attachment and hav e been shown to be safe in clinical trials. S ince the goal of this research was to use topographical cues to direct differentiation, targeting EPCs with antiCD34 should provide sufficient EPC capture and retention. An ECM protein, fibronectin, was used as a proof of concept in grafting and cell attachment experiments. Similar work should be completed to show the same effects using anti CD34. Materials and Methods Materials PEGDMA ( = 1 kg/mol) was purchased from Polysciences Inc. (Warrington, PA). 2 hydroxyethyl methacrylate 98% stabilized was purchased from Acros Organics (Geel, Belgium). Glycidyl methacrylate >97%, ascorbic acid (AA) 99+%, ammonium persulfate, and albumin from bovine serum were purchased from SigmaAldrich (Milwaukee, WI). Metha cryloxypropyltriethoxysilane (MPS) was purchased from Gelest Inc. (Morrisville, PA). Ultra pure water was produced by a Barnstead Nanopure Ultra

PAGE 115

115 Pure Water System (Waltham, MA). The base material for standards was a platinum catalyzed PDMSe (Silastic T2; Dow Corning Corporation). Sample P reparation PEGDMA, PEGDMAco GMA, and PEGDMAco HEMA hydrogels were produced using a thermally activated polymerization process. Aqueous solutions were pr epared by combining 25wt% PEGDMA ( = 1 kg mol1), 0.5 wt % ammo nium persulfate and ascorbic acid as chemical initiators, and ultra pure water to balance. To create a functionalized PEGDMA hydrogel 5 wt% of GMA or HEMA was added to the aqueous solution (Figure 61). Figure 61. Chemical structures of monomers used to produce functionalized hydrogels. The hydrogels were either produced as free standing films or attached to 76 x 22 mm microscope glass slides during the curing process by way of a silane coupling agent as described in Chapter 5. To produce free standi ng films all components of the prepolymer solution were combined in a glass beaker and stirred until the PEGDMA was dissolved. The prepolymer solution was then poured into two centrifuge tubes and centrifuged for 10 min at 3300 RPM. A pipette was used to fill a mold made of two glass plates and a PDMSe gasket with the centrifuged prepolymer solution. A

PAGE 116

116 topographically modified silicon wafer was added to the mold to create patterned samples. The mold was then placed in an oven to cure for 45 min at 45C. Hydrogels were stored in deionized water after curing. S mooth standards and topographically modified PDMSe standards were also produced. The elastomer was prepared by mixing 10 parts by weight of resin and 1 part by weight curing agent. The mixture was stirred by hand for 5 min and degassed under vacuum (2830 in Hg) for 30 min to remove bubbles. An allyltrimethoxysilanecoupling agent was applied to clean glass microscope slides (0.5 wt% in 95% ethanol/water soluti on) and polymerized for 10 min at 120C. The Silastic T2 was then placed in contact with the treated slides in a mold consisting of two glass plates and aluminum spacers. The elastomer was polymerized at ambient for 24 h. Topographically modified PDMSe samples were prepared in a two step castin g process previously described (Carman, et al. 2006) Attaching Hydrogels to Glass Slides Hydrogels were attached to glass slides by way of a silane couplin g agent. To prepare a solution of MPS, 30 mL of 190 proof ethanol was pipetted into a polypropylene cup. While mixing with a stir bar 1 to 2 drops of glacial acetic acid was added to adjust the pH of the solution to approximately 4.55.5. Then 0.17 mL of MPS was added to the solution and allowed to react for 5 min. Glass slides were cleaned by holding with forceps and passing through a flame 4 to 5 times. Slides were allowed to cool on an aluminum tray covered with Kimwipes A plastic transfer pipett e was used to cover glass slides with the MPS solution. The MPS solution was allowed to react on the glass slides for 2 to 3 min The slides were then rinsed with 190 proof ethanol and placed in the oven to dry at 120C for 10 min. After the slides were treated with MPS, a

PAGE 117

117 slide was placed into a mold consisting of two glass plates and a PDMSe gasket. A topographically modified silicon wafer was placed in this mold to make slides with topographically modified hydrogel. H ydrogel solution was pipetted into the mold and the filled mold was placed into an oven at 45C for 45 min to cure and attach to the glass slide. Grafting Fibronectin to Hydrogel Surfaces Fibronectin was grafted to hydrogel surfaces by reacting t he epoxide ring of the GMA with amine nucl eophiles on the protein in an alkaline buffer (Volcker et al., 2001). To make the buffer a 0.2 M solution of anhydrous sodium carbonate was prepared by combining 2.12 g of sodium carbonate with 100 mL of deionized water. Then a 0.2 M solution of sodium bicarbonate was prepared by combining 1.68 g of sodium bicarbonate with 100 mL of deionized water. A carbonatebicarbonate buffer was made by combining 4 mL of the sodium carbonate solution with 6 mL of sodium bicarbonate solution and bringing the volume up to 200 mL with deionized water. The pH of this solution was measured to be 9.8 at 24C. This buffer was used to make a solution of 50 g mL1 Fn in buffer. The surface to be grafted was then covered with the Fn buffer solution and incubated at 37 C with 5% CO2 for one hour. After incubation the Fn solution was removed from the surface and the surface was rinsed three times with phosphate buffered saline (PBS). Hydrogel Characterization The average molecular weight between crosslinks ( ) was determi ned experimentally using the method described by Peppas and Barr Howell (Peppas & Barr Howell, 1986). Six replicates of each hydrogel formulation: PEGDMA, PEGDMAco -

PAGE 118

118 GMA and PEGDMAco HEMA were cut from free standing films. The following measurements were made for each sample: Wa,r the sample weight in air after crosslinking Wn,r the sample weight in a nonsolvent after crosslinking Wa,s the sample weight in air after swelling Wn,s the sample weight in a nonsolvent after swelling Wa,d the sample weight in air after drying These measurements were then used to calculate the volume of the hydrogel sample after crosslinking before swelling (Vg,r) and after equilibrium swelling (Vg,s). The equations used to calculate the hydrogel volumes are based on Archimede s buoyancy principle = (6 1) = (6 2) n is the density of the nonsolvent at the temperature of the experiment. The volume of the hydrogel sample after cros slinking but before swelling is described by Equation 61 while the volume of the hydrogel after equilibrium swelling is described by Equation 62. The volume of the dry polymer was calculated using Equation 63 Vp=Wa dp (6 3) p is the density of the polymer. The polymer fractions of each hydrogel sample in both the relaxed and swollen state were calculated using Equation 64 and Equation 6 5, respectively : 2 = (6 4) 2 = (6 5)

PAGE 119

119 The sol PEGDMAco GMA and PEGDMAco HEMA were determined by swelling three replicates of each hydrogel in seven solvents with a range of solubility parameters (see Table 61 ) for 24 h. Table 61. Properties of sol vents used to calculate solubility parameters adapted from (Brandup et al. 1999) Solvent Solubility Parameter (MPa 1/2 ) Density (g/c m) Hexane 14.9 0.659 Benzene 17.2 0.874 Toluene 18.2 0.865 Chloroform 19.0 0.568 Acrylonitrile 21.5 0.806 Methanol 29.7 0.791 Water 47.9 1.0 The solubility parameter of the hydrogel was estimated as the maxima of the plot of Equation 66 (where m is the mass of the hydrogel after swelling in solvent for 24 h; m0 is the mass of the hydrogel before swelling after vacuum drying for 45 mins is the density of the solvent) versus the solubility parameter of the solvent. = 00 1 (6 6) The average molec ular weight between crosslinks (< Mc > ) for each hydrogel was calculated using the Flory Rehner Equation (Equation 6 7) (Sperling, 2001) [ ln ( 1 2) + 2+ 12 2] = 1 [ 21 3 22 ] (6 7) where 2 is the polymer volume fraction in the equilibrium swollen gel; 1 is the solvent polymer interaction parameter; V1 is the molar volume of the solvent; and n is the density of active chains between crosslinks. The value of 1for the PEGDMA co GMA

PAGE 120

120 syste m was previously determined to be a function of the polymer volume fraction (Pfister et al., 2007): 1( 2) = 0. 49 + 0. 142+ 0. 592 2 (6 8) The Youngs modulus of an elastomer may be predicted by measure the equilibrium swelling behavior. Y oungs modulus can be written = ( ) (6 9) which leads to = 20 2 ( 22+1 ) 3 20 2 (6 10) if equibiaxial extension is assumed 20 2 1. The Youngs modulus of PEGDM A co GMA, PEGDMAco HEMA and PDMSe were also measured with thermomechanical analysis (TMA). Three discs with a diameter of 10 mm and thickness of 1 mm were cut from free standing films of each hydrogel and tested. Force was applied in 0.1 N increments up to 0.5 N with a TMA2940 from Thermal Instruments. The displacement was recorded in micrometers. The forcedisplacement curve was used to calculate the shear modulus (G) of the material with Equation 611. The average molecular weight between crosslinks was calculated with Equation 612. The shear modulus was then converted to the Youngs modulus (E) using Equation 613. = ( 2 1) 2 4 [ 2 1 2 ] (6 11) = < > ( 1 2 < > ) (6 12)

PAGE 121

121 = 2 ( 1 + ) (6 13) Hydrogel composition and Fn grafting w ere verified using attenuated total reflectance Fourier transform infrared spectrometry (ATR FTIR) and immunofluorescence microscopy. Samples of PEGDMA, PEGDMAco GMA, PEGMDAco HEMA, Fn treated PEG MDA and PEGDMAco GMA graft Fn were made and dried for 48 h r at ambient temperature. A Perkin Elmer One Spectrometer with a ZnSe crystal (60) and resolution of 4 cm1 was used to record 20 scans for each sample. Spectral subtraction was performed to ver ify hydrogel composition and Fn grafting. Additionally, Fn grafting was performed as described above on four samples: PEGDMAco GMA smooth, PEGDMA co GMA +0.6CH2x2, PEGDMA co GMA +1SK2x2, and a smooth PEGDMA standard. A dilution of the primary antibody, anti Fn produced in mouse, at 1:50 antibody solution to PBS was prepared. Then a dilution of the secondary antibody, anti mouse IgG FTIC, at 1:32 antibody solution to PBS was prepared. The Fngrafted surfaces were covered with the primary antibody and in cubated at 37 C with 5% CO2 for 1 h The antibody solution was removed from the surfaces and the surfaces were rinsed three times with PBS. The surfaces were then covered with the secondary antibody solution and incubated again for 1 h at 37 C with 5% CO2. After incubation the surfaces were rinsed three times with PBS. Ten epi fluorescent micrographs were taken on each sample. ImageJ software was used to determine the average pixel intensity for each image using the Color Histogram plugin. The average fluorescence intensity for smooth PEGDMA and PEGDMA co GMA was compared.

PAGE 122

122 Cell Culture Assays Sam ple preparation Freestanding films of PEGDMA, PEGDMAco GMA and PDMSe were produced using the method described in the polymer synthesis section. Smooth and t opographically modified surfaces were tested. The topographies tested included the nseries of Sharklet topographies and channels The nseries is a group of topographies designed to have an increasing number of unique features (n) arranged in the Sharklet pattern (Figure 62) Figure 62. Sharklet surfaces varying in the distinct num ber of featur es (n) and Channels replic ated in PDMSe. A) +1 SK2x 2_n1, B) +1SK2x2_n2, C) +1SK2x2_n3, D) +1SK2x 2_n4, E ) + 1 SK2x 2_n5 and F ) +1CH2x2. The samples were created with features protruding from the surface in a Sharklet pattern Therefore, using the current nomenclature the nseries surfaces were referred to as +1SK2x2_n4 where n is the number of unique features and ranges from one to five. The channels topography is represent ed as +1CH2x2. Three 14 mm discs of each film were punched out and glued

PAGE 123

123 into wells of a 24well plate with uncured polymer The samples were sterilized by im mersion in 70% v/v ethanol in water for 1 h Samples were then rinsed with PBS three times and exposed to the Fn grafting process described above. Porcine vascular endothelial cell culture Porcine vascular endothelial cells (PVECs) from a primary culture provided by Dr. Edward Blocks laboratory were seeded onto smooth and +1CH2x2 and +1SK2x2_4 topographies created in PEGDMA graft Fn and PEGDMA co GMA graft Fn at 5x104 cells mL1. The PVECs were cultur ed on the hydrogels for 24 h. Cells were fixed with 10% formalin for 5 min and stained with crystal violet. Three transmitted light micrographs were taken per sample at a magnification of 400x using a Zeiss Axioplan 2 Microscope with a digital camera. The number of cells per field of view was counted and the average number of cells mm2 was reported as an indication of PVEC attachment to each hydrogel surface. H uman cell culture Two types of human cells were provided by Dr. Mark Segals laboratory: HCAECs and HASMCs. Topographies with a height of 1 including the nseries and channels were cast in PDMSe and PEGDMAco GMA and prepared as described above. Smooth surfaces and empty tissue culture polystyrene (TCPS) wells were used as standards and controls in these assays. Each cell type was seeded on to an identical plate full of samples at 5x104 or 2.5x104 cells per well and placed into an incubat or at 37C with 5% CO2. Samples were imaged after 24 h and 7 d using an inverted phase contrast microscope. Cell morphology was observed as an indication of cell attachment and response to the surfaces. Three phase contrast images were taken at 10x magnification for each combination of chemistry and topography.

PAGE 124

124 Image analysis Phase contrast images were analyzed with ImageJ software. Images of topographically modified surfaces were first rotated until the channels between features were oriented horizontally. The boundaries of HCAECs and HCASMCs were outlined using the freeform select tool and added to the region of interest manager. The projected area (S) and perimeter (L) of each cell were measured. ImageJ was also used to fit an ellipse to each cell and the angle between the major axis of each fitted ellipse and the direction of the channels in each topography was measured. Cell morphology was quantified by calculating t he cell shape index (CSI) (Sarkar et al. 2006a, Cao, et al. 2010) : =4 2 (6 14) Elongated cells will have a CSI approaching 0 while cells with a circular shape will have a CSI closer to 1. Orientation was quantified by tallying the number of cells within ranges of 10 degrees and plotting the percentage of cells in each range (Sarkar, et al. 2006a) Statistical methods The mean angle of cell orient ation relative to the channels on each topography was calculated using cir cular statistical methods to calculate the rectangular coordinates of the mean angle (Zar 1984) The following equations were used to calculate the rectangular coordinates of the mean angle from frequence data: = cos (6 15) = sin (6 16)

PAGE 125

125 In these equations, ai is the midpoint of the measurement interval recorded and fi is the frequency of occurrence of an angle within that interval. The magnitude of the mean angle was then calculated with Equation 617: = 2+ 2 (6 17) The mean angle was then taken to be the angle having the following cosine and sine: = (6 18) = (6 19) Since the dispersion of the data was very small compared with the period of 360, linear statistics were used to calculate 95% confidence limits and run an ANOVA with Tukey Kramer Test for multiple comparisons (Batschelet 1981) A Kolmogorov Smirnov test of each group showed that the mean angle data was not normally distributed. Therefore, the data was transformed by taking the square root. The 95% confidence limits and ANOVA were performed on the transformed data. The mean angle for every topography was plotted in bar charts with error bars representing 95% confid ence limits and horizontal bars represent results of the Tukey Kramer Test ( =0.05). Results and Discussion Hydrogel Characterization Results Solubility parameter The solubility parameter s for the PEGDMA, PEGDMAco GMA and PEGDMAco HEMA hydrogels were estimated by plotting a volumetric swelling ratio (Q) (Equation 66) versus the solubility parameter of several different solvents in which the hydrogels we re swollen for 24 h (Figure 63 ) The maximum volumetric swelling ratio corresponds

PAGE 126

126 with the solubility parameter of the hydrogel. Maximum swelling of all three hydrogels occurr ed in chloroform; therefore 1/2. Figure 63 Plot of v olumetric swelling ratio versus solvent solubility parameter for PEGDMA, PEGDMAco GMA and PEGDMAco HEMA hydrogels Average molecular weight between crosslinks and modulus The average molecular weight between crossli nks and modulus of PEGDMA co HEMA and PEGDMAco GMA were measured using TMA and swelling experiments. The Youngs modulus of PDMSe was measured to be 150 kPa, which is consistent with previous measurements of Silastic T2 PDMSe ( Feinberg et al. 2003) The measured modulus of PEGDMA co GMA was not different from that of PDMSe and both were greater than that of PEGDMA co HEMA (Table 6 2). Swelling experiments yielded lower values than TMA measurements. The was measured to be 160 to 240 g mol 1 and 200 to 280 g mol 1 for PEGDMAco HEMA and PEGDMAco GMA, respectively. 1 0 1 2 3 4 5 6 7 0 10 20 30 40 50 60 Q Solubility Parameter (MPa 1/2 ) PEGDMA PEGDMA co GMA PEGDMA co HEMA

PAGE 127

127 Table 62. Hydrogel characterization results: average molecular weight between crosslinks and modulus values. G (k Pa) E (k Pa) TMA (g/mol) Swel ling (g/mol) PEGDMA-co -HEMA 30 90 240 160 PEGDMA-co -GMA 50 160 280 200 PDMSe 5 0 0 1500 ----Hydrogel composition Spectral subtraction of PEGDMA from PEGDMA co GMA shows characteristic bands for GMA at 1715 to 1740 cm1 (C=O stretch), 1485 to 1445 c m1 (CH deformation), 1280 to 1230 cm1 (C O C symmetric stretch) and 950 to 815 cm1 (asymmetric stretch) (Figure 64 ) Spectral subtraction of PEGDMA from PEGDMA co HEMA shows characteristic peaks for HEMA, i.e., 3400 to 3200 cm1 (OH stretch), 2863 to 2843 cm1 (CH symmetric stretch), 1750 to 1735 cm1 (C=O stretch), 1485 to 1445 cm1 (CH asymmetric deformation) and 1150 to 1060 cm1 (C O C asymmetric stretch) (Figure 65 ) These results ver ified that hydrogels were functionalized by the addition of GMA or HEMA monomer into the prepolymer solution before polymerization. Figure 64 Spectral subtraction of ATR FTIR spectr a of PEGDMAco GMA minus PEGDMA. PEGDMA co GMA minus PEGDMA

PAGE 128

128 Figure 65 Spectral subtraction of ATR FTIR spectr a of PEGDMAco HEMA minus PEGDMA. Fibronecti n grafting Spectral subtraction of PEGDMA co GMA from PEGDMA co GMA graft Fn show characteristic peaks for the protein Fn, i.e., 34003200 cm1 (OH stretch) and 16501590 cm1 (Amide I Band). Spectral subtraction of PEGDMA from PEGDMA graft Fn shows the s ame peaks; however, these peaks are much less pronounced indicating a smaller amount of protein present on the surface (Figure 66) Figure 66 Spectral subtraction of ATR FTIR spectra of PEGDMA graft Fn minus PEGDMA and PEGDMAco GMA graft Fn minus PE GDMA co GMA. T he epoxide ring of the GMA was reacted with the amine nucleophiles on the Fn protein in an alkaline buffer solution. PEGDMA without GMA was used as a control surface to show that Fn was grafted, not adsorbed, to the sample surfaces. PEGDMA co HEMA minus PEGDMA PEGDMA co GMA graft Fn minus PEGDMA co GMA PEGDMA graft Fn minus PEGDMA

PAGE 129

129 Representative fluorescent images (Figure 67 ) indicated there is very little fluorescent signal coming from the PEGDMA surface compared to the three PEGDMA co GMA surfaces. Figure 67 Epifluorescent micrographs (magnification 500x) of immunofluorescently labeled Fn. A) PEGDMA Smooth, B) PEGDMAco GMA Smooth, C) PEGDMAco GMA +0.6CH2x2, D) PEGDMAco GMA +1SK2x2 Average fluorescence intensity measurements confirm ed that there is a statistically significant difference between the average fluorescence intens ity on the smooth PEGDMA and PEGDMAco GMA samples (see Figure 6 8 ). These results along with those from ATR FTIR indicated that Fn was grafted, not adsorbed to PEGDMAco GMA hydrogels.

PAGE 130

130 Figure 68 Fluorescence intensity of immunofluorescently labeled Fn. The asterisk indicates a significant statistical difference between groups based on a Ttest Cell Culture Assay Results PVEC assay The average number of cells mm2 was calculated for PEGDMA and PEGDMA co GMA hydrogels that had been exposed to the Fn grafting process. The average number of cells mm2 was 190 and 310 for PEGDMA graft Fn and PEGMDAco GMA graft Fn, respectively (Figure 610 ) A Students T a statistically significant difference between the average number of cells attached to each hydrogel. These results represent a 1.6fold increase in cells attached to the Fn grafted PEGDMA co GMA surface versus the PE GDMA control. The average number of cells mm2 attached to PEGDMA co GMA graft Fn s urfaces was 160 for the +1SK2x2_n4 topography versus 50 and 60 for smooth and +1CH2x2, respectively. 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Smooth Smooth PEGDMA PEGDMA co GMA Average Pixel Intensity (a.u) T -*

PAGE 131

131 There was a 3.2fold increase in the average number of cells attached to +1SK2x2_n4 compared to the smooth and +1CH2x2 topography. Figure 69 PVEC c ell culture assay results: average number of cells mm2 on smooth surfaces The asterisk indicates a significant statistical dif ference between groups based on a TFigure 610 PVEC cell culture assay results: average number of cells mm2 on smooth and topographically modified surfaces The horizontal bar indicates no statistically significant difference based on a T0 50 100 150 200 250 300 350 400 PEGDMA PEGDMAcoGMA Cells/mm 2 T -test ( =0.05)* 0 50 100 150 200 250 Smooth +1CH2x2 +1SK2x2 Cells/mm 2 Tukey Test ( =0.05)

PAGE 132

132 Human cell assays Cell morphology was observed as an indication of cell attachment and spreading on PEGDMAco GMA graft Fn and PDMSe surfaces. Phase contrast images show that ECs, SMCs and EPCs attached to but did not spread on PEGDMA co GMA graft Fn (Figures 613 and 617) Therefore, cell culture assays on hydrogels were discontinued after 48 h. Both HCAECs and HASMCs showed attachment and spreading on PDMSe surfaces at 24 h and 7 d. Phase contrast micrographs were tak en at 10 x magnification and the orientation of the channels in each topography is indicated with an arrow as described in Figure 611. A blank box with no arrow indicates a smooth surface. Figure 611. Arrows indicate the direction of the channels in each topography on phase contrast micrographs.

PAGE 133

133 Figure 61 2 HCAECs seeded at 5x104 cells/well on PDMSe A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS after 24 h.

PAGE 134

134 Figure 61 3 HCAECs seed ed at 5x104 cells/well on PEGDMAco GMA graft Fn A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, and G) TCPS after 24 h

PAGE 135

135 Figure 61 4 HCAECs seeded at 2.5x104 cells/well on PDMSe A) smooth, B) +1SK2x2_n1, C) +1SK2x2 _n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H ) TCPS after 24 h.

PAGE 136

136 Figure 61 5 HCAECs seeded at 2.5x104 cells/well on PDMSe A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS a fter 7 d.

PAGE 137

137 Figure 61 6 HASMCs seeded at 2.5x104 cells/well on PDMSe A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS after 24 h.

PAGE 138

138 Figure 617. HASMCs seeded at 2.5x104 cells/well on PDMSe A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS after 7 d.

PAGE 139

139 Figure 61 8 HASMCs seeded at 2.5x104 cells/well on PEGDMAco GMA graft Fn A) smooth, B) +1SK2x2_n1, C) +1SK2x2_n2, D) +1SK2x2_n3, E) +1SK2x2_n4 and F) TCPS after 24 h. Cell morphology was quantified by calculating the CSI index for each cell. The CSI ranges from 0 (elongated, linear cells) to 1 (circular shaped cells). After 24 h HCAECs showed no statistically significant elongation c ompared to a smooth PDMSe standard and a TCPS control (Figure 619). The average CSI for HCASMCs after 24 h was lowest on +1SK2x2_n4, however this value was only statistically different from +1SK2x2_n2 and the TCPS control (

PAGE 140

140 Figure 61 9 Average CSI for HCAECs cultured on topographi es replicated in PDMSe A) after 24 h and B) after 7 d. Horizontal bars indicate no statistical differences. Figure 620 Average CSI for HC A SMCs cultured on topographies replicated in PDMSe A) after 24 h and B) after 7 d. Horizontal bars indicate no statistical differences. The average CSI for HCAECs cultured for 7 d was lowest on +1CH2x2, which was not statistically different from +1SK2x2_n5, +1SK2x2_n4 and +1SK2x2_n1 (Figure 6 2 0 ). Elongation was hi ghest on +1SK2x2_n3 and +1SK2x2_n4 for HCASMCs after 7 d. The average CSI for these surfaces was not statistically different from the TCPS control (Figure 6 20) ( All topographies increased HCASMC elongation after 7 d compared to smoot h PDMSe ( 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cell Shape Index Tukey Test ( =0.05) HCAEC 24 h 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cell Shape Index Tukey Test ( =0.05) HCAEC 7 d A B 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cell Shape Index Tukey Test ( =0.05)HCASMC 24 h 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cell Shape Index Tukey Test ( =0.05) HCASMC 7 d A B

PAGE 141

141 Figure 621 Distribution of the percentage of HCAECs at angles in each 10 degree range after 24 h on A) S mooth PDMSe, B)+1SK2x2_n1, C)+1SK2x2_n2, D)+1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H ) TCPS. 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM PDMSe 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n1 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n2 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n3 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n4 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n5 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1CH2x2 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM TCPS A B C D E F G H

PAGE 142

142 Figure 6 22 Distribution of the percentage of HCAECs at angles in each 10 degree range after 7 d on A) Smooth PDMSe, B)+1SK2x2_n1, C)+1SK2x2_n2, D)+1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS. 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM PDMSe 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n1 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n2 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n3 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n4 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n5 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1CH2x2 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM TCPS A B C D E F G H

PAGE 143

143 Figure 62 3 Distribution of the percentage of HCASMCs at angles in each 10 degree range after 24 h on A) Smooth PDMSe, B)+1SK2x2_n1, C)+1SK2x2_n2, D)+1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H ) TCPS. 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM PDMSe 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n1 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n2 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n3 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n4 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n5 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1CH2x2 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM TCPS A B C D E F G H

PAGE 144

144 Figure 62 4 Distribution of the percentage of HCASMCs at angles in each 10 degree range after 7 d on A) Smooth PDMSe, B)+1SK2x2_n1, C)+1SK2x2_n2, D)+1SK2x2_n3, E) +1SK2x2_n4, F) +1SK2x2_n5, G) +1CH2x2 and H) TCPS. 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM PDMSe 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n1 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n2 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n3 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n4 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1SK2x2_n5 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) +1CH2x2 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 Percentage of Cells in Range (%) SM TCPS A B C D E F G H

PAGE 145

145 Figure 625. Average angle of cell orientation relative to the channels in the topography is plotted for A) HCAEC after 24 h in culture and B) HCAEC after 7 d in culture. Figure 626. Average angle of cell orientation relative to the channels in the topography is plotted for A) HCASMC after 24 h in culture and B) HCASMC after 7 d in culture. Both HCAECs and HCASMCs showed increasing orientation along the direction of the channels in the topographies after 24 h with increasing number of features (n) when cultured on the nseries Sharklet AF topographies. The highest degree of orientation was observed on the +1CH2x2 topography for HCAECs cultured for 24 h. Cells cultured on the smooth PDMSe standard and TCPS control did not show preferential orientation, i.e., the mean angles on these topographies were approximately 45. After 7 d the highest degree of orientation was observe d on +1SK2x2_n5 for HCAECs 0 10 20 30 40 50 60 Average Angle Tukey Test ( =0.05) HCAEC 24 h 0 10 20 30 40 50 60 Average Angle HCAEC 7 d Tukey Test ( =0.05) A B 0 10 20 30 40 50 60 Average Angle HCASMC 24 h Tukey Test ( =0.05) 0 10 20 30 40 50 60 Average Angle Tukey Test ( =0.05) HCASMC 7 d A B

PAGE 146

146 (Figure 62 5 ). More HCAECs were orientated along the direction of the channels in each topography after 7 d than after 24 h. The mean angles were lower on topographically modified surfaces after 7 d. The highest degree of orient ation was observed on +1SK2x2_n4 for HCASMCs after 24 h and 7 d (Figure 62 6 ). Discussion Functionalized PEGDMA hydrogels were created, characterized, topographically modified and tested as cell culture substrates. The solubility parameters were m easured for all three hydrogels. Maximum swelling occurred in chloroform indicating a solubility parameter of 19 MPa1/2 which corresponds well with the ranges of experimental and theoretical values of solubility parameter for crosslinked poly(ethylene oxi de) (PEO) rep orted in the literature, 19 to 40 MPa1/2 and 18 to 31 MPa1/2 respectively (Barton, 1990; Graham et al., 1981; Katz & Salee, 1968). The average molecular weight between crosslinks and Youngs modulus were measured for each hydrogel using swell ing experiments and TMA. Subtraction of ATR FTIR spectra verified that the hydrogels were functionalized by the addition of GMA or HEMA monomer into the prepolymer solution before polymerization. Spectral subtraction and immunofluorescence microscopy bot h confirmed grafting of Fn to the PEGDMA co GMA hydrog el by way of the epoxide ring. Assays with PVECs showed a 1.6fold increase in cells attached to the Fn grafted PEGDMAco GMA surface versus the PEGDMA control. This value corresponds well to the 2fol d increase that was predicted in the hypothesis and is consistent with previously reported results (Markway, et al. 2008) The average number of cells attached to +1SK2x2_n4 was 3.2fold larger than the average number of cells attached to the smooth and +1CH2x2 topography after 24 h. Topographical modification of PEG based

PAGE 147

147 biomaterials was shown to increase fibroblast attachment versus smooth (Schulte, et al. 2009) Channels ranging in size from 5 to 50 m and pillars 3 m in diameter with 3 or 6 m spacing increased attachment and spreading of fibroblasts after 4 h and 24 h (Schulte, et al. 2009) It is possible that the topographical modification led to a change in the amount, type and/or conformation of protein adsorbed to the PEG surfaces. The topographically modified surfaces could also direct or promote cytoskeletal organization which could also lead to better attachment and spreading of the cells. The +1SK2x2_n4 topography showed a clear increase in PVEC attachment and spreading versus smooth a fter 24 h while the +1CH2x2 topography did not. This increase was most likely due to cytoskeletal organization. The small feature spacing did not allow cells to settle within the topographies and this difference may explain why cell attachment did not increase on the channels topography versus a smooth surface as it did in the report by Schulte, et al. Hydrogels based on PEG are known to be nonadhesive surfaces i.e., nonspecific protein adsorption is minimized (Ostuni, et al. 2001, Balamurugan, et al. 2005, Schulte, et al. 2009) These materials were selected as a cell culture substrate so that protein adsorption and therefore cellular attachment could be controlled by chemical and topographical modification. A ll surfaces made of PEGDMA co GMA graft Fn hydrogel showed attachment of HCAECs and HASMCs after 24 h. Minimal cell spreading was observed after 3 d in culture indicating these hydrogels did not provide an ideal substrate for cell culture. Assays on the PEGDMAco GMA graft Fn hydrogels were discontinued after 48 h. Both cell types attached to and spread on all PDMSe surfaces.

PAGE 148

148 Elongation and alignment of HCA ECs and HASMCs were quantified. Topographies with the greatest number of distinct features (n=4 and n=5) in the nseries and +1CH2x2 resulted in the largest elongations and orientations of HCAECs and HCASMCs after 24 h and 7 d. Discontinuous microchannels were shown to increase expression of proteins representative of the contractile phenotype in vascular smooth muscle cells (Cao, et al. 2010) Topography may trigger alignment and elongation, as well as some intracellular signal that triggers the expression of the contractile phenotype (Thakar et al. 2003, Sarkar, et al. 2006a, Beamish et al. 2010) Using topography to direct HCASM C s to express the contractile phenotype could be a strategy to reduce intimal hyperplasia in small diameter vascular graft applications. Protein expression of H C ASMCs should be quantified in cells cultured on both topographically modified and smooth PDMSe to determine if any of the nseries topographies c ould be used to regulate not only orientation, but also phenotype of these cells. Proteins expressed by SMCs in the contractile phenotype include smooth muscle actin and myosin heavy chain. Immunostaining assays could be used to quantify the expression of these proteins by cells cultured on each topography (Beamish, et al. 2010) Functionalized PEGDMA hydrogels are not an ideal cell cul ture substrate for small diameter vascular graft applications. Cells attached to but did not spread on both smooth and topographically modified surfaces created in PEGDMA co GMA graft Fn. It is, however, still feasible to use topographies created in PDMS e to direct cell morphology and possibly phenotype. Further studies should be performed to quantify protein expression of HCAECs and HASMCs on both smooth and topographically modified surfaces. The ability to use topography to select a certain cell phenotype

PAGE 149

149 would advance small diameter vascular graft technology. Studies should also be continued with EPCs to determine the influence of topography on differentiation.

PAGE 150

150 CHAPTER 7 CONCLUSIONS AND FUTURE WORK Conclusions Non Toxic Antifouling Strategies Biof ouling is a technically complex problem that remains an economic and environmental concern. New combinations of material chemistries and topographies were evaluated for antifouling properties with organism s representing three diverse phylogenetic groups, viz. the eukaryotic Plantae ( Ulva ), the eukaryotic Chromista ( Navicula ) and the prokaryotic Bacteria ( Cobetia) (Cavalier Smith 2004) Engineered antifouling topographies in PDMSe inhibited attachment of C. marina up to 99% versus a smooth standard. The factors contributing to the slope of the line created when plotting the attachment model were investigated. The size and motility of bacterial cells and algal spores were incorporated into the attachment model (Long, et al. 2010) by multiplying the ERIII by the Re of the cells. The results showed a negative linear correlation of the normalized transformed attachment densities for both organisms with ERIII*Re (R2 = 0.77) (Magin, et al. 2010b) The same microtopographies also created in PDMSe reduced the attachment density and attachment strength of cells of the diatoms N avicula and S eminavis compared to smooth PDMSe. However, the results did not cor relate with current models, i. e., the attachment mo del and attachment point theory A new analysis showed that t he average normalized, transformed attachment density of Navicula after exposure to shear stress (48 Pa) correlated with the contact area between the diatom and a topographi cally modified surface (R2 = 0.82).

PAGE 151

151 Functionalized PEGDMA hydrogels significantly reduced attachment and attachment strength of Navicula and C. marina. A ll three hydrogel compositions reduced significantly (p=0.05) initial attachment of cells of Navicula (up to 58%) and C. marina ( up to 62%) as well as zoospores of Ulva linza (up to 97%) compared to a smooth PDMSe standard. A shear stress (45 Pa), in a water channel, eliminated up to 95% of the initially attached Navicula cells from the topopgraphicall y patterned surfaces relative to smooth PDMSe surfaces. Compared to the smooth PDMSe standard 79% of the C. marina cells were removed from all hydrogel compositions when exposed to the same shear stress. The Sharklet AF microtopography patterned PEGDMA co HEMA surfaces reduced attachment of Ulva by 97% compared to a smooth PDMSe standard. The attachment of spores of Ulva to engineered microtopographies in PDMSe and PEGDMAco HEMA negatively correlated with the attachment model that includes the engineered roughness index (ERIII) multiplied by the Reynolds number (Re) of the organism. The attachment model was extended with the addition of a surface energy term. Attachment densities of cells from two evolutionarily diverse groups correlated with the attachment model for various engineered topographies replicated in materials other than PDMSe (R2 = 0.80). The extension of the attachment model to incorporate new fouling organisms and new surface chemistries creates an algorithm that can be used to design nontoxic antifouling surfaces. Cell Culture Substrates H ydrogels based on PEGDMA and standards made of PDMSe were topographically modified and evaluated as substratum material s for mammalian cell culture. Capturing endothelial progenitor cells (EPCs) and inducing differentiation into the endothelial cell (EC) phenotype is the ideal way to reendothelialize a small -

PAGE 152

152 di ameter vascular graft. Likewise, modulation of the SMC phenotype is a way to reduce intimal hyperplasia after the implantation of a synthetic graft. Substratum elasticity has been reported to direct stem cell differentiation into specific lineages (Engler, et al. 2006) Functionalized PEGDMA hydrogels provided good compliance, high fidelity of topographic features and sites for surface modification with biomolecules. Fibronectin grafting and topography both increased EC attachment. This combination of adjustable elasticity, surface chemistry and topography has the potential to promote the capture and differentiation of EPCs into a confluent EC monolayer. Engineered microtopographies replicated in PDMSe directed alignment and elongation of HCAECs and HCASMCs compared to smooth surfaces. Both elongation and alignment along the channels of the features in the topographies increased with increasing (n), the number of unique features, on the Sharklet AF patterns. The ability to control cell morphology is important for developing healthy cell cultures that function in the same way as cells grown in vivo Future Work Non Toxic Antifouling Strategies The attachment model can be used t o design new combinations of surface chemistry and engineered microtopography for antifouling applications. However, the model does not completely describe all of the complex factors that contribute to biofouling. The model should be tested to investigat e other factors contributing to the slope of the model and validate its current form. The incorporation of the Re of the organism into the model can be validated by running additional assays using swimming organisms in static assays or nonswimming organi sms under flow conditions Flagellated marine bacteria, for example, would have a different Re than C. marina and

PAGE 153

153 could be used as a test organism. Since diatoms are not motile in the water column the attachment of these cells could be measured under flow conditions so that an average Re could be calculated for the organism. These results may lead to the incorporation of a third marine fouling organism into range described by the current model. Biodiversity within a cell or spore population is another factor that could be incorporated into the attachment model. One way to test this concept is to transfect a population of cells or spores at one age with green fluorescent protein and an older population of the same cells or spores with yellow fluorescent protein and track attachment. Older cells or spores may attach with different affinity or in different preferential locations on topographies. Engineered antifouling topographies should be created with a range of surface energies and evaluated for fouling resistance with a standard Ulva attachment assay. These data can be used to validate the addition of the surface energy term to the current attachment model. Cell Culture Substrates Increased alignment and elongation of HCASMCs could be triggering a s witch from the synthetic to the contractile phenotype. Cells in the contractile phenotype contract in response to molecular signals and have a low proliferative index. Both of these qualities are desired in cells growing on a newly implanted vascular graft. Intimal hyperplasia, a result of excessive smooth muscle cell proliferation, is one of the leading causes of restenosis in small diameter vascular grafts. Controlling the phenotype of HCASMCs that are in contact with a graft could reduce the probabil ity of graft occlusion. The expression of contractile proteins in HCASMCs cultured on engineered

PAGE 154

154 microtopographies, such as smooth muscle actin and myosin heavy chain should be quantified to confirm phenotypic expression. The aim of the cell culture substrate work was to create a cell culture substrate for smalldiameter vascular graft applications that has the potential to re endothelialize in vivo to reduce neointimal hyperplasia and thrombosis. It wa s hypothesized that a combination of surface chemistry and topography on a graft surface would capture circulating EPCs in the peripheral blood and promote their differentiat ion into ECs to create a continuous tissue layer within the lumen of the graft. A biomolecular surface chemistry such as a bioactive peptide or antibody would increase recruitment of endothelial progenitor cells and topographical modification would provide mechanical cues to induce differentiation into the endothelial cell phenotype. This research should be continued in order to examine the ability of physical cues such as engineered microtopographies to induce cell differentiation. Engineered topographies that induce differentiation into the EC phenotype could be combined with those that direct SMCs to remain in the contractile phenotype to create a platform for building vascular graft materials. Since SMCs have a higher aspect ratio than ECs, i.e., SMCs are much longer than they are wide, changing the width and spacing of the topographies could influence differentiation to one phenotype over the other or provide for selection of SMCs and/or E Cs from a co culture population. The influence of engineered m icrotopographies on the formation of an EC monolayer could also be evaluated by staining for a protein present in EC gap junctions such as vascular endothelial caderhin. The formation of functional gap junctions is essential for EC function in vivo

PAGE 155

155 Cel l stiffness has been shown to be influenced by substrate modulus (Engler, et al. 2006, Feinberg, et al. 2009) Cell stiffness may also play a role in how cells interact with engineered microtopographies. A stiffer cell, for example, could deform microtopographic features to settle in between features, while a less stiff cell might spread across the features. To investigate cell stiffness the antimitotic agent, nocodazole, could be added to cells to decrease cell stiffness when seeding cells onto the surfaces. Nocodazole disrupts microtubules by binding to tubulin and preventing the formation of disulfide linkages. This action inhibits micr otubule polymerization within cells. Nocodazole can easily be rinsed from the cells after attachment to the surface to restore a normal degree of stiffness. This treatment could be used to create functional monolayers of cells on engineered microtopographies.

PAGE 156

156 APPENDIX A REYNOLDS NUMBER CALC ULATIONS C. marina: Stationary Phase L = 1 m Logarithmic Growth Phase L = 2 m Flow Cell Parameters: Q = 2x105 L/s h = 1.27 x10 -4 m w = 1.27x102 m v = 9.7x105 L = 1 x102 kg/m s = 1x103 kg/m3 = (A 1) =2 10 5 (0 .001 3 1 ) 1 27 10 2 1 27 10 4 (A 2) = 1 .2 10 3 / (A 3) is the average velocity within the flow cell. The velocity of the bacteria relative to the fluid (V) near the wall was estimated to be 20% of the average fluid velocity in the flow cell. = (A 4) C. marina: Stationary Phase Re = 2.5 x 103 in the flow cell Logarithmic Growth Phase Re = 5 x 103 in the flow cell

PAGE 157

157 Ulva: The diameter of the spore body was taken to be the characteristic length: L = 5 m. The average swimming speed was reported to be 150 m/s (Heydt et al. 2007) This value was taken to be the velocity of the spore relative to the fluid (V). Re = 7.7 x 104

PAGE 158

158 APPENDIX B SUM MARY O F RECENT SMALL DIAMETER VASCULAR GRAFT LITERATURE

PAGE 159

159 Table B. Summary of recent literature on small diameter vascular grafts. Author(s) Year Material(s) Processing Cell Type(s) Advantages Disadvantages (Wu, et al. 2007) Type I Collagen glutaraldehyde crosslinked Bi layer membrane SMC EC Implated in rat Good mechanical properties Regenerated rat vena cava in vivo Have to remove and culture cells ex vivo (Chan, et al. 2007) Type I Collagen photochemically crosslinked with rose Bengal gel Subcutaneous implantation in rat Better mechanical properties than uncrosslinked collagen gels Good stability and tissue compatibility Not tested with vascular cells or in vasculature Only loose connective tissue grew on implant (Cummings, et al. 2004) Type I Collagen Fibrin 3D ma trix gel Rat aortic SMC Fibrin can be used to change mechanical properties of collagen gel No difference in cell proliferation among samples Mechanical properties still do not match artery No cell differentiation / endothelialization (Nguyen and West 2002) PEG co hydroxy acid) Photopolymeriz able hydrogels ECs Implanted in rat SMC Can control modulus, permeability, bioerodability Cells grow and spread Can attach grow factors and adhesion molecules Must optimize mechanical properties (Zhu, et al. 2005) PCL b PEG b PCL DA Micropatterned foldable hydrogel UV embossing 3T3 fibroblasts Biocompatible Biodegradable Patternable High water content Permeability Shrinking/Sw elling of patterned surface Not tested with ECs or SMCs (Stankus, et al. 2006) Poly(l lactide co caprolactone), poly(ester urethane) Electrospun fiber SMCs electrosprayed in hydrogel Strong, flexible, elastic High cell d ensity Have to remove and culture cells ex vivo (He et al. 2005) Poly(l lactide co caprolactone), collagen Electrospun fiber Nanofiber mesh Human coronary artery ECs Enhanced spreading, viability, and attachment Preserved phenotype Have to remove and culture cells ex vivo

PAGE 160

160 (Sell and Bowlin 2008) Synthetic polymers Proteins Electrospun fiber ECs SMCs Good mechanical properties Mimics ECM Good cell growth Have to remove and culture cells ex vivo (Smith, et al. 2008) polydioxanone, elastin, sutures Electrospun fiber SMC and EC Great mechanical properties Biodegradable Bioactivity Tendency to form aneurysm Uneven degradation No mechanism for endothelialization (Lee, et al. 2008a) PCL, colla gen Electrospun fiber SMC and EC Great mechanical properties Resist high pressure over long time Support cell growth Have to remove and culture cells ex vivo No mechanism for endothelialization (Lee et al. 2007) PCL Collagen Elastin Glutaraldehyde crosslinked Electrospun fiber Ovine SMC Improved mech props Cell infiltration Can control size and mech props Have to remove and culture cells ex vivo No mechanism for endothelialization (Chung, et al. 2007a) PLCL Gel spun Implanted in mouse Controllable mech props No mechanism for endothelialization (Zhang, et al. 2007) PLGA, PU layered Bone marrow stroma l cell seeded then implanted in canine Good mech strength 3 month patency in vivo Endothelialized Have to remove and culture cells ex vivo Low degradation rate might increase inflammatory response (Feng, et al. 2007) PCLLGA Collagen Type I hydrogel Layer by layer Microgrooves SMC layered in grooves with or without collagen Good alignment of SMCs Rapid 3D fabrication Have to remove and culture cells ex vivo No mechanism for endothelialization (Choi, et al. 2007) PLAePTFE PLA Layered biodegradable and nonbiodegradab le PLA made porous by gas foaming SMC with pulsatile flow Good mech props Biodegradable Have to remove and culture cells ex vivo No mechanism for endothelial ization (Roh, et al. 2008) PGA or PLA Nonwoven with polyester glue Implanted in mouse No thrombosis or ane urysm Foreign body immune response in 3 weeks

PAGE 161

161 Endothelialization and collagen deposits at 3 weeks Mouse model not representative of human endothelialization (Fidkowski et al. 2005) Poly(glycerol sebacate) Biodegradable elastomer Etched capillary patterns onto silicon for molds EC perfused w ith syringe at physiological flow rate Endothelialized in vitro in 14 days Have to remove and culture cells ex vivo No mechanism for endothelialization (Xu et al. 2008) PU, polyester, spandex Knitted No cell testing yet Improved mech props Controllable dimensions No cell testing (Nieponice et al. 2008) PEUU Porous matrix tube Seeded with muscle derived stem cells Cell proliferation and viability Stem cell phenotype preserved, no differentiat ion Have to remove and culture cells ex vivo No mechanis m for endothelialization (Alm any and Seliktar 2005) PEG fibrinogen Biosynthetic hydrogel scaffolds EC and SMC cured inside scaffold Bioactive Fibrinogen should enhance EC attachment Controllable mech props Proteolytic biodegradation Need to test endothelializtion in vivo (Patel, et al. 2007) Hydrogel poly(acrylamide) poly (ethylene glycol co acrylic acid) GRGDSP on interpenetrating polymer network ECs Increased EC adh esion and spreading Proliferation rate unchanged EC migration inhibited under flow conditions (Lee, et al. 2004) Alginate Hydrogels Microencapsulat ion of growth factors coated with heparin and chitosan VEGF and FGF No cell testing Can locally deliver growth factors Delivery complete after 5 10 days not long enough to enhance cell attachm ent or differentiation (Watanabe, et al. 2007) Autologous ECM and cells Implanted in body to collect tissue Autoimplantation up to 2 months No form ation of aneurysm or rupturing Autologous Must create an implantation site to form tube then graft (Badylak 2002) Small intestine ECM, explanted Xenogeneic and Recruitment of Sourcing

PAGE 162

162 submucosa from porcine allogeneic ECM grafts circulating progenitors ECM remo deling Extensive angiogenesis Rapid scaffold degradation Optimizing mech props Immunologic response to the scaffolds (Wang, et al. 2007) Decellularized xenografts With and without heparin From canine to rabbit Showed cellular reconstruction with EC and SMC after 6 months Without heparin thrombosis rate was 58% Sourcing Immunologic response (L'Heureux, et al. 2006) Completely TE Fibroblasts from skin biopsy made into sheets and wrapped around mandrel Matured 10 weeks Seeded ECs Implanted in nude rats and primates Cell infiltration Good mech props Long ex vivo maturation time Have to remove and culture cells ex vivo Endothelialization models do not mimic human (Aoki, et al. 2005) Stainless Steel stent with CD 34 antibody Covalently coupled polysaccharide intermediate coating with murine monoclonal anti human CD34 antibodies Human C linical Trial (16 patients) Safe and feasible for treatment of coronary artery disease Luminal loss after 6 months ~0.63mm Increased in vitro cell capture >3 fold Intravascular volume obstruction ~27% Intimal hyperplasia not significantly reduced (Co, et al. 2008) Stainless Steel stent with CD 34 antibody Covalently coupled polysaccharide intermediate coating with murine monoclonal anti h uman CD34 antibodies Human Clinical Trial (120 patients) Low rate of major cardiac events at 6 months ~5.8% Low rate of late stent thrombosis ~0.28% No in vitro cell testing No characterization of removed stents or cell differentiation or endothelializati on (Ma rkway, et al. 2008) Glass coverslips Coated w/ HUVECs, KDR is a leading Captured ECs not EPCs

PAGE 163

163 with or without protein G spacer monoclonal mouse anti human KDR IgG1 or anti w/ heat denat. BSA, BSA controls BaSMCs candidate for selection of EPCs Anti KDR with prot ein G spacer captured 2.5 fold more cells No data on differentiation (Rodenberg and Pavalko 2007) Microplate wells or glass slides Coated with recombinant fibronectin peptide fragments HCAECs, HUVECs Both EC types adhered maximally to CS 1 (1.6 fold in crease) Tested only with ECs not EPCs No capture data for EPCs No differentiation data for EPCs (Blindt, et al. 2006) PEUU coated Guidant Tetra stents Controlled release of cRGD Polymer c oated w/ or w/o cRGD and bare metal stents implanted in porcine coronary arteries, infused with EPCs Culture plate coated and perfu sed in parallel plate flow chamber cRGD significant increase over control cRGD stimulates out growth, shear resistant recrui tment, invasion of EPCs Recruited and counted EPCs but did not investigate differentiation of EPCs into ECs (Veleva, et al. 2007) Biopanning 12mer peptide ligands selected with phage display that bind HBOEC Human blood outgrowth ECs Isolated clones display cell specificity No significant binding is observed on panel of other cell types No differentiation studies (Veleva et al. 2008) Peptidemodified terpolymers Peptides incorporate into mehtacyrlic terpolymers via chain transfer free radic al polymerization Human blood outgrowth ECs Ligands do not interfere with EC function Ligand bound HBOEC but not HUVEC No studies on progenitor differentiation (Esguerra et al. 2010) Bacterial cellulose Hydrogel produced by SMCs Implantation in Low inflammatory response Cell ingrowth less than PGA Angiogenesis less

PAGE 164

164 scaffolds bacteria dorsal skinfold chamber of hamsters Increased cell growth versus ePTFE Angiogenesis in border zones pronounced in borders than for ePTFE and PGA (Soletti et al. 2010) P oly(ester urethane) urea Electrospinning Thermally induced phase separation Rotational vacuum seeding of muscle derived stem cells Mechanical properties similar to artery High cell seeding Dynamic culture Mechanical properties not assessed after cell seeding or in wet conditions Have to remov e and culture cells ex vivo before implantation (Nichol et al. 2010) Gelatin methacrylate Hydrogel Immortalized human umbilical vein ECs NIH 3T3 cells Cell attachment, proliferation and spreading on surfaces Can micropattern Mechanical properties not assessed Topographical resolution only 100m (Kibbe et al. 2010) poly(1,8 octanediol citrate) coated ePTFE g rafts Shear spinning Implanted in porcine model POC ePTFE grafts similar extent of neointimal hyperplasia and leukocyte and monocyte/ macrophage infilt ration as control ePTFE grafts POC supported ECs Not tested in small diameter vessels Results were from 28d study (Gui et al. 2009) Decellularized human umbilical arteries Explantation and cleaning HUVECs Implanted into nude rats Re endothelialized with HUVECs Retained function up to 8 weeks in mice Burst pressur e similar to native arteries Obtaining human umbilical arteries is a challenge (Zhou et al. 2009) Decellularized canine carotid arteries coated with heparin and VEGF Explantation and treatment HUVECs Implantation into dog model R educe d in vivo thrombogenicity I mproved early patency rate of grafts T Patent at 6 months H igh degree of reendothelialization Studies performed in dog model for only 6 months I nfection, calcification, aneurismal dilation, and vasoreactivity need to be assessed (Stekelenburg et al. 2009) Fast degrading polyglycolic acid scaffold coated PGA sheet wrapped around mandrel and Myofibroblasts from discarded saphenous Similar mechanical properties to native vessel All ass ays performed in vitro

PAGE 165

165 with poly 4 hydroxybutyrate combined with fibrin gel dip coated veins Grown in bioreactor with dynamic strain Cell infiltration and collagen deposition (Tillman et al. 2009) PCL/Collagen Electrospinning R abbit aort oiliac bypass model S upport ed adherence and growth of vascular cells under physiologic conditions E ndothelialized grafts resisted adherence of platelets I mplanted in Vivo scaffolds retained structural integrity over 1 mo Rabbit model does not mimic human re endthelialization well Short assay time period (Hong et al. 2009) P oly(ester urethane)urea P oly(2 methacryloyloxy ethyl phosphorylcholi neco methacryloyloxy ethyl butylurethane) Electrospinning Rat SMCs Rat abdominal aorta implantation Good mechanical properties Slowed proliferation of rat SMCs Thin neo intimal layer with endothelial coverage and good anastomotic tissue integration after 8 weeks Rat model does not mimic human conditions well 8 week study (Koens et al. 2010) Elastin Collagen C asting, moulding, freezing and lyophilization C arbodiimide cross linked and heparinized P latelet aggregation tests Burst pressures similar to native vessel No platelet aggregation No cell culture in vitro or in vivo has been performed (Zhang et al. 2006) Poly(propylene carbonate) Electrospinning G enetically modified MSCs Cells form 3D network NO produced by grafts seeded with eNOS modified MSCs wa s comparable to that produced by native blood vessels Cell sourcing Genetically modified cells could elicit immune response

PAGE 166

166 LIST OF REFERENCES Abarzua S and Jakubowski S. 1995. Biotechnological investigation for the prevention of bio fouling. I. Biological and biochemical principles for the prevention of biofouling. Marine Ecology Progress Series. 123:301312. Aldred N, Scardino A, Cavaco A, de Nys R and Clare AS. 2010. Attachment strength is a key factor in the selection of surfaces by barnacle cyprids ( Balanus amphitrite) during settlement. Biofouling. 26(3):287299. Almany L and Seliktar D. 2005. Biosynthetic hydrogel scaffolds made from fibrinogen and polyethylene glycol for 3D cell cultures. Biomaterials. 26(15):24672477. Andrewartha J, Perkins K, Sargison J, Osborn J, Walker G, Henderson A and Hallegraeff G. 2010. Drag force and surface roughness measurements on freshwater biofouled surfaces. Biofouling. 26(4):487496. Aoki J, Serruys PW, van Beusekom H, Ong ATL, McFadden EP, Sianos G, van der Giessen WJ, Regar E, de Feyter PJ, Davis HR, Rowland S and Kutryk MJB. 2005. Endothelial Progenitor Cell Capture by Stents Coated with Antibody Against CD34. Jornal of the American College of Cardiology. 45(10):15741579. Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G and Isner JM. 1997. Isolation of Putative Progenitor Endothelial Cells for Angiogenesis. Science. 275:964. Badylak SF. 2002. The extracellular matrix as a scaffold for tissue reconstruction. Seminars in Cell & Developmental Biology. 13(5):377383. Baier RE. 2006. Surface behaviour of biomaterials: The theta surface for biocompatibility. Journal of Materials Science: Materials in Medicine. 17:10571062. Bakker DP, Busscher HJ, van Zanten J, de Vries J, Klijnstra JW and van der Mei HC. 2004. Multiple linear regression analysis of bacterial deposition to polyurethane coatings after conditioning film formation in the marine environment. Microbiology. 150:17791784. Balamurugan S, Ista LK, Yan J, Lopez GP, Fick J, Himmelhaus M and Grunze M. 2005. Reversible Protein Adsorption and Bioadhesion on Monolayers Terminated with Mixtures of Oligo(ethylene glycol) and Methyl Groups. J Am Chem Soc. 127:1454814549. Batschelet E editor 1981. Circul ar Statistics in Biology. London, England: Academic Press, Inc.

PAGE 167

167 Baumann L, Bowditch RD and Baumann P. 1983. Description of Deleya gen. nov. Created to Accommodate the Marine Species Alcaligenes aestus A. pacificus A. cupidus A. venustus and Pseudomonas marina. International Journal of Systematic Bacteriology. 33(4):793802. Beamish JA, He P, Kottke Marchant K and Marchant RE. 2010. Molecular Regulation of Contractile Smooth Muscle Cell Phenotype: Implications for Vascular Tissue Engineering. Tissue E ngineering: Part B. Bechert DW, Bruse M and Hage W. 2000. Experiments with threedimensional riblets as an idealized model of shark skin. Experiments in Fluids. 28:403412. Bers AV, Diaz ER, da Gama BAP, Vieira Silva F, Dobretsov S, Valdivia N, Thiel M, Sc ardino AJ, McQuaid CD, Sudgen HE, Thomason JC and Wahl M. 2010. Relevance of mytilid shell microtopographies for fouling defence a global comparison. Biofouling. 26(3):367377. Bers VA and Wahl M. 2004. The Influence of Natural Surface Microtopographies on Fouling. Biofouling. 20(1):4351. Bettinger CJ, Langer R and Borenstein JT. 2009. Engineering Substrate Topography at the Micro and Nanoscale to Control Cell Function. Angew Chem Int Ed. 48:5406=5415. Bico J, Marzolin C and Quere D. 1999. Pearl drops. Europhys Lett. 47(2):220226. Bico J, Thiele U and Quere D. 2002. Wetting of texture surfaces. Colloids Surf, A. 206(13):4146. Biology of the Arterial Wall. 1999. Dordrecht/Boston/London: Kluwer Academic Publishers. Biomaterials Science: An Introduction to Materials in Medicine. 2004. 2 ed. San Diego: Elsevier Academic Press. Blindt R, Vogt F, Astafieva I, Fach C, Hristov M, Krott N, Seitz B, Kapurniotu A, Kwok C, Dewor M, Bosserhoff A K, Bernhagen J, Hanrath P, Hoffmann R and Weber C. 2006. A Novel Drug Eluting Stent Coated With an IntegrinBinding Cyclic Arg Gly Asp Peptide Inhibits Neointimal Hyperplasia by Recruiting Endothelial Progenitor Cells. Journal of the American College of Cardiology. 47(9):17861795. Bowen J, Pettitt ME, Kendall K, Leggett G J, Preece JA, Callow ME and Callow JA. 2007. The influence of surface lubricity on the adhesion of Navicula perminuta and Ulva linza to aklanethiol self assembled monolayers. J R Soc Interface. 4(473477).

PAGE 168

168 Brady RF and Singer IL. 2000. Mechanical Factors Favoring Release from Fouling Release Coatings. Biofouling. 15(13):7381. Brandup J, Immergut EH and Grulke EA editors. 1999. Polymer Handbook. 4 ed. New York, NY: John Wiley & Sons, Inc. Callow ME, Callow JA, Pickett Heaps JD and Wetherbee R. 1997. Primar y adhesion of enteromorpha (chlorophyta, ulvales) propagules: quantitative settlement studies and video microscopy. J Phycol. 33:938947. Callow ME, Callow JA, Ista LK, Coleman SE, Nolasco AC and Lopez GP. 2000. Use of Self Assembled Monolayers of Different Wettabilities To Study Surface Selection and Primary Adhesion Processes of Green Algal ( Enteromorpha) Zoospores. Applied and Environmental Microbiology. 66(8):32493254. Callow ME, Jennings AR, Brennan AB, Seegert CA, Gibson AL, Wilson LH, Feinberg AW, Baney R and Callow JA. 2002. Microtopographic Cues for Settlement of Zoospores of the Green Fouling Alga Enteromorpha. Biofouling. 18(3):237245. Cao Y, Poon YF, Feng J, Rayatpisheh S, Chan V and ChanPark MB. 2010. Regulating orientation and phenotype of primary vascular smooth muscle cells by biodegradable films patterned with arrays of microchannels and discontinuous mircowalls. Biomaterials. 31(24):62286238. Carman ML, Estes TG, Feinberg AW, Schumacher JF, Wilkerson W, Wilson LH, Callow ME, Callow JA an d Brennan AB. 2006. Engineered antifouling microtopgraphies correlating wettability with cell attachment. Biofouling. 22(1):1121. Carman ML. Engineered Microtopographies to Induce In Vitro Endothelial Cell Morphologies Stable to Shear [Dissertation]. Gainesville FL: University of Florida. Cassie ABD and Baxter S. 1944. Wettability of Porous Surfaces. Trans Faraday Society. 40:546551. Cavalier Smith T. 2004. Only six kingdoms of life. Proc R Soc Lond B. 271:12511262. Chai C and Leong KW. 2007. Biomateri als Approach to Expand and Direct Differentiation of Stem Cells. Molecular Therapy. 15(3):467480. Chan BP, Hui TY, Chan OCM, So KF, Lu W, Cheung KMC, Salomatina E and Yaroslavsky A. 2007. Photochemical Cross Linking for CollagenBased Scaffolds: A Study on Optical Properties, Mechanical Properties, Stability, and Hemetocompatibility. Tissue Engineering. 13(1):7385. Chaudhury MK, Finaly JA, Chung JY, Callow ME and Callow JA. 2005. The influence of elastic modulus and thickness on the release of the soft f ouling green alga Ulva linza (syn. Enteromorpha linza) from poly(dimethylsiloxane) (PDMS) model networks. Biofouling. 2(1):4148.

PAGE 169

169 Chiovitti A, Dugdale TM and Wetherbee R editors. 2006. Diatom Adhesives: Molecular and Mechanical Properties. Berlin: Springer Verlag. Choi YJ, Kim M S and Noh I. 2007. Tissue Regeneration of a Hybrid Vascular Graft Composed of Biodegradable Layers and NonBiodegradable Layer by Static and Pulsatile Flows. Key Engineering Materials. 342343:61 64. Chung E, Kim S H, Ko YG, Kwo n J H, Han J W, Park IS, Han S S and Kim SH. 2007a. Characterization of Gel spun Tubular Scaffold for Cardiovascular Tissue Engineering. Key Engineering Materials. 342343:321324. Chung KK, Schumacher JF, Sampson EM, Burne RA, Antonelli PJ and Brennan A B. 2007b. Impact of engineered surface microtopographies on biofilm formation of Staphylococcus aureus. Biointerphases. 2(2):8994. Co M, Tay E, Lee CH, Poh KK, Low A, Lim J, Lim IH, Lim YT and Tan HC. 2008. Use of endothelial progenitor cell capture stent (Genous BioEngineered R Stent) during primary percutaneous coronary intervention in acute myocardial infarction: Intermediate to longterm clinical followup. American Heart Journal. 155(1):128132. Cordiero A, Nitschke M, Janke A, Helbig R, D'Souza F, Donnelly G, Willemsen P and Werner C. 2009. Fluorination of poly(dimethylsiloxane) surfaces by low pressure CF4 plasma physicochemical and antifouling properties. Express Polymer Letters. 3(2):70 83. Cowling MJ, Hodgkiess T, Parr ACS, Smith MJ and Marr s SJ. 2000. An alternative approach to antifouling based on analogues of natural processes. The Science of the Total Environment. 258:129137. Cummings CL, Gawlitta D, Nerem RM and Stegemann JP. 2004. Properties of engineered vascular constructs made from collagen, fibrin, and collagenfibrin mixtures. Biomaterials. 25(17):36993706. Curtis ASG and Wilkinson CD. 1998. Reactions of cells to topography. J Biomater Sci Polymer Edn. 9(12):13131329. Dalby MJ, Yarwood SJ, Riehle MO, Johnstone HJH, Affrossman S and Curtis ASG. 2002. Increasing Fibroblast Response to Materials Using Nanotopography: Morphological and Genetic Measurements of Cell Response to 13nm High Polymer Demixed Islands. Experimental Cell Research. 276(1):19. Dalby MJ, Childs S, Riehle MO, Johnstone HJH, Affrossman S and Curtis ASG. 2003. Fibroblast reaction to island topography: changes in cytoskeleton and morphology with time. Biomaterials. 24(6):927935. Dalby MJ. 2005. Topographically induced direct cell mechanotransduction. Medical Engineering and Physics. 27:730742.

PAGE 170

170 de Mel A, Jell G, Stevens MM and Seifalian AM. 2008. Biofunctionalization of Biomaterials for Accelerated in Situ Endothelialization: A Review. Biomacromolecules. Epub ahead of print. Drury J, Ashton T, Cunningham J, Maini R and Pollock J. 1987. Experimental and clinical experience with a gelatin impregnated Dacron prosthesis. Ann Vasc Surg. 1(5):542547. Ederth T, Pettitt ME, Nygren P, Du C X, Ekblad T, Zhou Y, Falk M, Callow ME, Callow JA and Liedberg B. 2009. Interactions of Zoospores of Ulva linza with Arginine Rich Oligopeptide Monolayers. Langmuir. 25(16):93759383. Edgar LA. 1982. Diatom Locomotion: A Consideration of Movement in a Highly Viscous Situation. Br phycol J. 17:243251. Edgar LA and Pickett Heaps JD. 1984. Di atom locomotion. Prog Phycol Res. 3:4788. Ekblad T, Bergstrom G, Ederth T, Conlan SL, Mutton R, Clare AS, Wang S, Liu Y, Zhao Q, D'Souza F, Donnelly GT, Willemsen PR, Pettitt ME, Callow ME, Callow JA and Liedberg B. 2008. Poly(ethylene glycol) Containing Hydrogel Surfaces for Antifouling Applications in Marine and Freshwater Environments. Biomacromolecules. 9:27752783. Engler AJ, Sen S, Sweeney HL and Discher DE. 2006. Matrix Elasticity Directs Stem Cell Lineage Specification. Cell. 126(4):677689. Esguer ra M, Fink H, Laschke M, Jeppsson A, Delbro D, Gatenholm P, Menger MD and Risberg B. 2010. Intravital fluorescent microscopic evaluation of bacterial cellulose as scaffold for vascular grafts. Journal of Biomedical Materials Research Part A. 93A:140149. F einberg AW, Gibson AL, Wilkerson W, Seegert CA, Wilson LH, Zhao LC, Baney R, Callow JA, Callow ME and Brennan AB. Investigating the Engergetics of Bioadhesion on Microengineered Siloxane Elastomers: Characterizing the Topography, Mechanical Properties and Surface Engergy and its Effect on Cell Contact Guidance. In: S Clarson, J Fitzgerald, M Owen, S Smith and M Dyke editors: Am Chem Soc. Feinberg AW, Schumacher JF and Brennan AB. 2009. Engineering highdensity endothelial cell monolayers on soft substrate s. Acta Biomaterialia. 5:20132024. Feng J, ChanPark MB, Shen J and Chan V. 2007. Quick Layer by Layer Assembly of Aligned Multilayers of Vascular Smooth Muscel Cells in Deep Microchannels. Tissue Engineering. 13(5):10031012. Fidkowski C, Kaazempur Mofra d M, Borenstein J, Vacanti JP, Langer R and Wang Y. 2005. Endothelialized Microvasculature Based on a Biodegradable Elastomer. Tissue Engineering. 11(1/2):302309.

PAGE 171

171 Finlay JA, Callow ME, Ista LK, Lopez GP and Callow JA. 2002a. The Influence of Surface Wettability on the Adhesion Strength of Settled Spores of the Green Alga Enteromorpha and the Diatom Amphora. Integr Comp Biol. 42:11161122. Finlay JA, Callow ME, Schultz MP, Swain GW and Callow JA. 2002b. Adhesion Strength of Settled Spores of the Green Alga Enteromorpha. Biofouling. 18(4):251256. Finlay JA, Krishnan S, Callow ME, Callow JA, Dong R, Asgill N, Wong K, Kramer EJ and Ober CK. 2008. Settlement of Ulva Zoospores on Patterned Fluorinated and PEGylated Monolayer Surfaces. Langmuir. 24:503510. Finla y JA, Bennett SM, Brewer LH, Sokolova A, Clay G, Gunari N, Meyer AE, Walker GC, Wendt DE, Callow ME, Callow JA and Detty MR. 2010. Barnacle settlement and the adhesion of protein and diatom microfouling to xerogel films with varying surface energy and water wettability. Biofouling. 26(6):657666. Freischlag J and Moore W. 1990. Clinical experience with a collagenimpregnated knitted Dacron vascular graft. Ann Vasc Surg. 4(5):499454. Genzer J and Efimenko K. 2006. Recent developments in superhydrophobic sur faces and their relevance to marine fouling: a review. Biofouling. 22(5):339360. Gudipati CS, Finlay JA, Callow JA, Callow ME and Wooley KL. 2005. The Antifouling and FoulingRelease Performance of Hyperbranched Fluoropolymer (HBFP) Poly(ethylene glycol) (PEG) Composite Coatings Evaluated by Adsorption of Biomacromolecules and the Green Fouling Alga Ulva Langmuir. 21:30443053. Gui L, Muto A, Chan S, Breuer CK and Niklason LE. 2009. Development of Decellularized Human Umbilical Arteries as Small Diameter Vascular Grafts. Tissue Engineering: Part A. 15(9):26652676. Harrison R. 1914. The reaction of embryonic cells to solid structures. The Journal of Experimental Zoology. 17(4):521544. Hatcher BM, Seegert CA and Brennan AB. 2002. Polyvinylpyrrolidone modi fied bioactive glass fibers as tissue constructs: In vitro mesenchymal stem cell response. Journal of biomedical Materials Research Part A. 66A(4):840849. He W, Ma ZW, Yong T, Teo WE and Ramakrishna S. 2005. Fabrication of collagencoated biodegradable polymer nanofiber mesh and its potential for endothelial cells growth. Biomaterials. 26:76067615. Heart Disease & Stroke Statistics Dalls, Texas: American Heart Association

PAGE 172

172 Heydt M, Rosenhahn A, Grunze M, Pettitt ME, Callow ME and Callow JA. 2007. Digital In line Holography as a ThreeDimensional Tool to Study Motile Marine Organisms During Their Exploration of Surfaces. The Journal of Adhesion. 83:417430. Hoagland KD, Rosowski JR, Gretz MR and Roemer SC. 1993. Diatom Extracellular Polymeric Substances: Function, Fine Structure, Chemistry and Physiology. J Phycol. 29:537566. Hoffman AS. 2002. Hydrogels for biomedical applications. Advanced Drug Delivery Reviews. 54(1):312. Holland R, Dugdale T, Wetherbee R, Brennan A, Finlay JA, Callow JA and Callow ME 2004. Adhesion and Motility of Fouling Diatoms on a Silicone Elastomer. Biofouling. 20(6):323329. Holm ER, Schultz MP, Haselbeck EG, Talbott WJ and Field AJ. 2004. Evaluation of hydrodynamic drag on experimental fouling release surfaces, using rotating disks. Biofouling. 20(4/5):219226. Holm ER, Kavanagh CJ, Meyer AE, Wiebe D, Nedved BT, Wendt D, Smith C, Hadfield MG, Swain G, Wood CD, Truby K, Stein J and Montemarano J. 2006. Interspecific variation in patterns of adhesion of marine fouling to silicone surfaces. Biofouling. 22(4):233243. Holzapfel GA, Gasser TC and Stadler M. 2002. A structural model for the viscoelastic behavior of arterial walls: Contimuum formulation and finite element analysis. European Journal of Mechanics A/Solids. 21:441463. H ong Y, SangHo Y, Nieponice A, Soletti L, Vorp DA and Wagner WR. 2009. A small diameter, fibrous vascular conduit generated from a poly(ester urethane)urea and phospholipid polymer blend. Biomaterials. 30:24572467. Hu DSG and Tsai C e. 1996. Correlation Between Interfacial Free Energy and Albumin Adsorption in Poly(acrylonitrileacrylamideacrylic acid) Hydrogels. Journal of Applied Polymer Science. 59:18091817. Ista LK, Mendez S and Lopez GP. 2010. Attachment and detachment of bacteria on surfaces with tunable and switchable wettability. Biofouling. 26(1):111118. Itano N, Okamoto S i, Zhang D, Lipton SA and Ruoslahti E. 2003. Cell spreading controls endoplasmic and nuclear calcium: A physical gene regulation pathway from the cell surface to the nucleus Proc Natl Acad Sci USA. 100(9):51815186. Jain A and Bhosle NB. 2009. Biochemical composition of the marine conditioning film: implications for bacterial adhesion. Biofouling. 25(1):13 19.

PAGE 173

173 Joint I, Tait K, Callow ME, Callow JA, Milton D, Williams P and Camara M. 2002. Cell to Cell Communication Across the ProkaryoteEukaryote Boundary. Science. 298:1207. Ju H, McCloskey BD, Sagle AC, Kusuma VA and Freeman BD. 2009. Preparation and characterization of crosslinked poly(ethylene glycol) diacrylate hydrogels as foulingresistant membrane coatings. Journal of Membrane Science. 330:180188. Kibbe MR, Martinez J, Popowich DA, Kapadia MR, Ahanchi SS, Aalami OO, Jiang Q, Webb AR, Yang J, Carroll T and Ameer G. 2010. Citric acidbased elastomers provide a biocompati ble interface for vascular grafts. Journal of Biomedical Materials Research Part A. 93A:314324. Koens MJW, Faraj KA, Wismans RG, van der Vliet JA, Krasznai AG, Cuijpers VMJI, Jansen JA, Daamen WF and van Kuppevelt TH. 2010. Controlled fabrication of tripl e layered and molecularly defined collagen/elastin vascular grafts resembling the native blood vessel. Acta Biomaterialia. Krishnan S, Weinman CJ and Ober CK. 2008. Advances in polymers for anti biofouling. Journal of Materials Chemistry. 18:34053413. L'H eureux N, Dusserre N, Konig G, Victor B, Keire P, Wight TN, Chronos NAF, Kyles AF, Gregory CR, Hoyt G, Robbins RC and McAllister TN. 2006. Human tissueengineered blood vessels for adult arterial revascularization. Nature Medicine. 12(3):361365. Lee KW, Y oon JJ, Lee JH, Kim SY, Jung HJ, Kim SJ, Joh JW, Lee HH, Lee DS and Lee SK. 2004. Sustained release of vascular endothelial growth factor from calcium induced alginate hydrogels reinforced by heparin and chitosan. Transplantation Proceedings. 36(8):24642465. Lee SJ, Yoo JJ, Lim GJ, Atala A and Stitzel J. 2007. In vitro evaluation of electrospun nanofiber scaffolds for vascular graft application. Journal of biomedical Materials Research Part A. 83A(4):9991008. Lee SJ, Liu J, Oh SH, Soker S, Atala A and Yoo JJ. 2008a. Development of a composite vascular scaffolding system that withstands physiological vascular conditions. Biomaterials. 29:28912898. Lee SJ, Oh SH, Liu J, Soker S, Atala A and Yoo JJ. 2008b. The use of thermal treatments to enhance the mechani cal properties of electrospun poly([var epsilon] caprolactone) scaffolds. Biomaterials. 29(10):14221430. Liliensiek SJ, Wood JA, Yong J, Auerbach R, Nealey PF and Murphy CJ. 2010. Modulation of human vascular endothelial cell behaviors by nanotopographic cues. Biomaterials. 31:54185426.

PAGE 174

174 Lim JY and Donahue HJ. 2007. Cell Sensing and Response to Microand Nanostructured Surfaces Produced by Chemical and Topographic Patterning. Tissue Engineering. 13(8):18791891. Long CJ. Analytic Methods for Predicting Bi osettlement on Patterned Surfaces [Dissertation]. Gainesville, FL: University of Florida. Long CJ, Schumacher JF, Robinson P, II, Finaly JA, Callow ME, Callow JA and Brennan AB. 2010. A Model that Predicts the Attachment Behavior of Ulva linza Zoospores on Surface Topography. Biofouling. 26(4):411419. Lutolf MP and Hubbell JA. 2005. Synthetic biomaterials as instructive extracellular microenvironments for morphogenesis in tissue engineering. Nature Biotechnology. 23(1):4755. Magin CM, Cooper SP and Brennan AB. 2010a. NonToxic Antifouling Strategies. Materials Today. 13(4):3644. Magin CM, Long CJ, Cooper SP, Ista LK, Lopez GP and Brennan AB. 2010b. Engineered Antifouling Microtopographies: The role of Reynolds number in a model that predicts attachment of zoospores of Ulva and cells of Cobetia marina. Biofouling. 26(6):719727. Markway BD, McCarty OJT, Marzec UM, Courtman DW, Hanson SR and Hinds MT. 2008. Capture of Flowing Endothelial Cells Using SurfaceImmobilized AntiKinase Insert Domain Receptor Ant ibody. Tissue Engineering. 14(2). Marmur A. 2006. Super hydrophobicity fundamentals: implications to biofouling prevention. Biofouling. 22(2):107115. Miglionico M, Patti G, D'Ambrosio A and Di Sciascio G. 2008. Percutaneous Coronary Intervention Utilizin g a New Endothelial Progenitor Cells Antibody Coated Stent: A Prospective SingleCenter Registry in High Risk Patients. Catheterication and Cardiovascular Interventions. 71:600604. Molino PJ and Wetherbee R. 2008. The biology of biofouling diatoms and their role in the development of microbial slimes. Biofouling. 24(5):365379. Molino PJ, Campbell E and Wetherbee R. 2009a. Development of the initial diatom microfouling layer on antifouling and fouling release surfaces in temperate and tropical Australia. Biofouling. 25(8):685694. Molino PJ, Childs S, Hubbard MRE, Carey JM, Burgman MA and Wetherbee R. 2009b. Development of the primary bacterial microfouling layer on antifouling and fouling release coatings in temperate and tropical environments in Eastern Australia. Biofouling. 25(2):149162.

PAGE 175

175 Moon JJ, Hahn M, Kim I, Nsiah B and West JL. 2009. Micropatterning of Poly(Ethylene Glycol Diacrylate Hydrogels with Biomolecules to Regulate and Guide Endothelial Morphogenesis. Tissue Engineering: Part A. 15(3):579 5 85. Murosaki T, Noguchi T, Kakugo A, Putra A, Kurokawa T, Furukawa H, Osada Y, Gong JP, Nogata Y, Matsumura K, Yoshimura E and Fusetani N. 2009. Antifouling activity of synthetic polymer gels against cyprids of the barnacle ( Balanus amphitrite ). Biofouling. 25(4):313320. Nerem RM and Seliktar D. 2001. Vascular Tissue Engineering. Annu Rev Biomed Eng. 3:225243. Nguyen KT and West JL. 2002. Photopolymerizable hydrogels for tissue engineering applications. Biomaterials. 23(22):43074314. Nichol JW, Koshy ST, Bae H, Hwang CM, Yamanlar S and Khademhosseini A. 2010. Cellladen microengineered gelatin methacrylate hydrogels. Biomaterials. 31:55365544. Nieponice A, Soletti L, Guan J, Deasy BM, Huard J, Wagner WR and Vorp DA. 2008. Development of a tissueengineer ed vascular graft combining a biodegradable scaffold, musclederived stem cells and a rotational vacuum seeding technique. Biomaterials. 29(7):825833. Ostuni E, Chapman RG, Holmlin RE, Takayama S and Whitesides GM. 2001. A Survey of Structure Property Rel ationships of Surfaces that Resist the Adsorption of Protein. Langmuir. 17:56055620. Otani M, Oumi T, Uwai S, Hanyuda T, Prabowo RE, Yamaguchi T and Kawai H. 2007. Occurrence and diversity of barnacles on international ships visiting Osaka Bay, Japan, and the risk of their introduction. Biofouling. 23(3/4):277286. Owens DK and Wendt RC. 1969. Estimation of Surface Free Energy of Polymers. J Appl Polym Sci. 13(8):17411747. Padera RF and Schoen FJ editors. 2004. Cardiovascular Medical Devices. San Diego: Elsevier Academic Press. Patel S, Tsang J, Harbers GM, Healy KE and Li S. 2007. Regulation of endothelial cell function by GRGDSP peptide grafted on interpenetrating polymers. journal of biomedical Materials Research Part A. 83A(2):423433. Perkins KJ, And rewartha JM, McMinn A, Cook SS and Hallegraeff GM. 2010. Succession and physiological health of freshwater microalgal fouling in a Tasmanian hydropower canal. Biofouling. 26(6):637644.

PAGE 176

176 Pettengill JB, Wendt DE, Schug MD and Hadfield MG. 2007. Biofouling li kely serves as a major mode of dispersal for the polychaete tubeworm Hydroides elegans as inferred from microsatellite loci. Biofouling. 23(3/4):161169. Pfister PM, Wendlandt M, Neuenschwander P and Suter UW. 2007. Surfacetextured PEG based hydrogels wit h adjustable elasticity: Synthesis and characterization. Biomaterials. 28(4):567575. Pickford AR, Smith SP, Stauton D, Boyd J and Campbell ID. 2001. The hairpin structure of the (6)F1(1)F2(2)F2 fragment from human fibronectin enhances gelatin binding. EMB O J. 20(7):15191529. Piola RR and Johnston EL. 2008. The potential for translocation of marine species via smallscale disruptions to antifouling surfaces. Biofouling. 24(3):145155. Quere D. 2008. Wetting and Roughness. Annu Rev Mater Res. 38:7199. Rals ton E and Swain G. 2009. Bioinspiration the solution for biofouling control? Bioinsp Biomim. 4. Rasmussen K, Willemsen PR and Ostgaard K. 2002. Barnacle Settlement on Hydrogels. Biofouling. 18(3):177191. Rasmussen K and Ostgaard K. 2003. Adhesion of the marine bacterium Pseudomonas sp. NCIMB 2021 to different hydrogel surfaces. Water Research. 37:519524. Reilly GC and Engler AJ. 2010. Intrinsic extracellular matrix properties regulate stem cell differentiation. Biomaterials. 43:5562. Rodenberg EJ and P avalko F. 2007. Peptides Derived from Fibronectin Type III Connecting Segments Promote Endothelial Cell Adhesion but Not Platelet Adhesion: Implications in TissueEngineered Vascular Grafts. Tissue Engineering. 13:26532666. Roh JD, Nelson GN, Brennan MP, Mirensky TL, Yi T, Hazlett TF, Tellides G, Sinusas AJ, Pober JS, Saltzman WM, Kyriakides TR and Breuer CK. 2008. Small diameter biodegradable scaffolds for functional vascular tissue engineering in the mouse model. Biomaterials. 29(10):14541463. Rosenhahn A, Schilp S, Kreuzer HJ and Grunze M. 2010. The role of "inert" surface chemistry in marine biofouling prevention. Physical Chemistry Chemical Physics. 12:42754286. Rumisek J, Wade C, Brooks D, Okerberg C, Barry M and Clarke J. 1986. Heat denatured albumin coated Dacron vascular grafts: physical characteristics and in vivo performance. Journal of Vascular Surgery. 4(2):136143.

PAGE 177

177 Saha K, Pollock JF, Schaffer DV and Healy KE. 2007. Designing synthetic materials to control stem cell phenotype. Current Opini on in Chemical Biology. 11(4):381387. Sarkar S, Lee GY, Wong JY and Desai TA. 2006a. Development and characterization of a porous micropatterned scaffold for vascular tissue engineering applications. Biomaterials. 27:47754782. Sarkar S, Sales KM, Hamilt on G and Seifalian AM. 2006b. Addressing Thrombogenicity in Vascular Graft Construction. journal of Biomedical Materials Research Part B: Applied Biomaterials. 82B(1):100 108. Scardino AJ, Harvey E and De Nys R. 2006. Testing attachment point theory: diatom attachment on microtextured polyimide biomimics. Biofouling. 22(1):5560. Scardino AJ, Guenther J and de Nys R. 2008. Attachment point theory revisited: the fouling response to a microtextured matrix. Biofouling. 24(1):4553. Scardino AJ, Hudleston D, P eng Z, Paul NA and de Nys R. 2009a. Biomimetic characterisation of key surface parameters for the development of fouling resistant materials. Biofouling. 25(1):8393. Scardino AJ, Zhang H, Cookson DJ, Lamb RN and de Nys R. 2009b. The role of nanoroughness in antifouling. Biofouling. 25(8):757767. Schilp S, Kueller A, Rosenhahn A, Grunze M, Pettitt ME, Callow ME and Callow JA. 2007. Settlement and adhesion of algal cells to hexa(ethylene glycol) containing self assembled monolayers with systematically chan ged wetting properties. Biointer phases 2(4):143 150. Schilp S, Rosenhahn A, Pettitt ME, Bowen J, Callow ME, Callow JA and Grunze M. 2009. Physicochemical Properties of (Ethylene Glycol) Containing Self Assembled Monolayers Relevant for Protein and Algal C ell Resistance. Langmuir. 25(17):1007710082. Schulte VA, Diez M, Moller M and Lensen MC. 2009. Surface Topography Induces Fibroblast Adhesion on Intrinscially Nonadhesive Poly(ethylene glycol) Substrates. Biomacromolecules. 10:27952801. Schultz MP, Finla y JA, Callow ME and Callow JA. 2000. A Turbulent Channel Flow Apparatus for the Determination of the Adhesion Strength of Microfouling Organisms. Biofouling. 15(4):243251. Schultz MP. 2007. Effects of coating roughness and biofouling on ship resistance and powering. Biofouling. 23(5):331341.

PAGE 178

178 Schumacher JF, Aldred N, Callow ME, Finaly JA, Callow JA, Clare AS and Brennan AB. 2007a. Species specific engineered antifouling topographies: correlations between the settlement of algal zoospores and barnacle cypr ids. Biofouling. 23(5):307317. Schumacher JF, Carman ML, Estes TG, Feinberg AW, Wilson LH, Callow ME, Callow JA, Finlay JA and Brennan AB. 2007b. Engineered antifouling microtopographies -effect of feature size, geometry, and roughness on settlement of zoospores of the green alga Ulva Biofouling. 23(1):5562. Schumacher JF, Long CJ, Callow ME, Finaly JA, Callow JA and Brennan AB. 2008. Engineered Nanforce Gradients for Inhibition of Settlement (Attachment) of Swimming Algal Spores. Langmuir. 24:49314937. Sell SA and Bowlin GL. 2008. Creating small diameter bioresorbable vascular grafts through electrospinning. J Mater Chem. 18:260263. Shen JY, ChanPark MB, He B and Zhu AP. 2006. ThreeDimensional Microchannels in Biodegradable Polymeric Films for Cont rol Orientation and Phenotype of Vascular Smooth Muscel Cells. Tissue Engineering. 12(8):22292240. Shin H, Temenoff JS and Mikos AG. 2003. In Vitro Cytotoxicity of Unsaturated Oligo[poly(ethylene glycol) fumarate] Macromers and Their Cross Linked Hydrogel s. Biomacromolecules. 4(3):552560. Smith M, J., McClure MJ, Sell SA, Barnes CP, Walpoth BH, Simpson DG and Bowlin GL. 2008. Suturereinforced electrospun polydioxanone elastin small diameter tubes for use in vascular tissue engineering: A feasibility study. Acta Biomaterialia. 4:5866. Soletti L, Hong Y, Guan J, Stankus JJ, El Kurdi MS, Wagner WR and Vorp DA. 2010. A bilayered elastomeric scaffold for tissue engineering of small diameter vascular grafts. Acta Biomaterialia. 6:110122. Stankus JJ, Guan J, Fujimoto K and Wagner WR. 2006. Microintegrating smooth muscle cells into a biodegradable elastomeric fiber matrix. Biomaterials. 27:735744. Stegemann JP, Kaszuba SN and Rowe SL. 2007. Review: Advances in Vascular Tissue Engineering Using ProteinBased B iomaterials. Tissue Engineering. 13(11):26012613. Stekelenburg M, Rutten MCM, Snoeckx LHEH and Baaijens FPT. 2009. Dynamic Straining Combined with Fibrin Gel Cell Seeding Improves Stregth of TissueEngineered Small Diameter Vascular Grafts. Tissue Engineering: Part A. 15(5):10811089.

PAGE 179

179 Thakar RG, Ho F, Huang NF, Liepmann D and Li S. 2003. Regulation of vascular smooth muscle cells by micropatterning. Biochemical and Biophysical Research Communications. 307:883890. Thomas CH, Collier JH, Sfeir CS and Healy KE. 2001. Engineering gene expression and protein synthesis by modulation of nuclear shape. Proc Natl Acad Sci USA. 99(4):19721977. Thompson SEM, Taylor AR, Brownlee C, Callow ME and Callow JA. 2008. The role of nitric oxide in diatom adhesion in relation to substratum properties. J Phycol. 44:967976. Tillman BW, Yazdani SK, Lee SK, Geary RL, Atala A and Yoo JJ. 2009. The in vivo stability of electrospun polycaprolactonecollagen scaffolds in vascular reconstruction. Biomaterials. 30:583588. van Kooten TG and von Recum AF. 1999. Cell adhesion to textured silicone surfaces: The influence of time of adhesion and texture on focal contact and fibronectin fibril formation. Tissue Engineering. 5(3):223 240. Veleva AN, Cooper SL and Patterson C. 2007. Selection and Initial Characterization of Novel Peptide Ligands that Bind Specifically to Human Blood Outgrowth Endothelial Cells. Biotechnology and Bioengineering. 98(1):306312. Veleva AN, Heath DE, Cooper SL and Patterson C. 2008. Selective endothelial cell attac hment to peptidemodified terpolymers. Biomaterials. 29(27):36563661. Verran J and Boyd RD. 2001. The Relationship Between Substratum Surface Roughness and Microbiological and Organic Soiling: a Review. Biofouling. 17(1):5971. Wang X N, Chen C Z, Yang M and Gu YJ. 2007. Implantation of Decellularized Small caliber Vascular Xenografts With and Without Surface Heparin Treatment. Artificial Organs. 31(2):99 104. Watanabe T, Kanda K, Ishibashi Ueda H, Yaku H and Nakayama Y. 2007. Development of biotube vascul ar grafts incorporating cuffs for easy implantation. J Artif Organs. 10:10 15. Weinberg CB and Bell E. 1986. A Blood Vessel Model Constructed from Collagen and Cultured Vascular Cells. Science. 231(4736):397400. Weiss P. 1945. Experiments on cell and axon orientation in vitro : the role of collodial exudates in tissue organization. Journal of Experimental Zoology. 100(3):353386.

PAGE 180

180 Wendt DE, Kowalke GL, Kim J and Singer IL. 2006. Factors that influence elastomeric coating performance: the effect of coating thickness on basal plate morphology, growth and critical removal stress of the barnacle Balanus amphitrite. Biofouling. 22(1):19. Wenzel RN. 1936. Resistance of Solid Surfaces to Wetting by Water. Industrial and Engineering Chemistry. 28(8):988994. Wetherbee R, Lind JL, Burke J and Quatrano RS. 1998. The First Kiss: Establishment and Control of Initial Adhesion by Raphid Diatoms. J Phycol. 34:915. Wu H C, Wang TW, Kang P L, Tsuang Y H, Sun J S and Lin FH. 2007. Coculture of endothelial and smooth mus cle cells on a collagen membrane in the development of a small diameter vascular graft. Biomaterials. 28:13851392. Xu W, Zhou F, Ouyang C, Cui W, Yao M and Wang X. 2008. Small Diameter Polyurethane Vascular Graft Reinforced by Elastic Weft Knitted Tubular Fabric of Polyester/Spandex. Fibers and Polymers. 9(1):7175. Xue L and Greisler HP. 2000. Blood Vessels. In: Principles of Tissue Engineering. 2 ed. New York: Academic Press. Yamaguchi T, Prabowo RE, Ohshiro Y, Shimono T, Jones D, Kawai H, Otani M, Oshino A, Inagawa S, Akaya T and Tamura I. 2009. The introduction to Japan of the Titan barnacle, Megabalanus coccopoma (Darwin, 1854) (Cirripedia: Balanomorpha) and the role of shipping in its translocation. Biofouling. 25(4):325333. Yebra DM, Kiil S and Dam Johansen K. 2004. Antifouling technology -past, present and future steps towards efficient and environmentally friendly antifouling coatings. Prog Org Coat. 50:74104. Yim EKF, Pang SW and Leong KW. 2007. Synthetic nanostructures inducing differentiation of human mesenchymal stem cells into neuronal lineage. Experimental Cell Research. 313:18201829. Zar JH. 1984. Biostatistical Analysis. Second ed. Englewood Cliffs, N.J.: PrenticeHall, Inc. Zhang J, Qi H, Wang H, Hu P, Ou L, Guo S, Li J, Che Y, Yu Y and Kong D. 2006. Engineering of Vascular Grafts with Genetically Modified Bone Marrow Mesenchymal Stem Cells on Poly(Propylene Carbonate) Graft. Artifical Organs. 30(12):898905. Zhang L, Zhou J, Lu Q, Wei Y and Hu S. 2007. A Novel Small Diameter Vascular G raft: In Vivo Behavior of Biodegradable ThreeLayered Tubular Scaffolds. Biotechnology and Bioengineering. 99(4):10071015.

PAGE 181

181 Zhang Z, Finlay JA, Wang L, Gao Y, Callow JA, Callow ME and Jiang S. 2009. PolysulfobetaineGrafted Surfaces as Environmentally Benign Ultralow Fouling Marine Coatings. Langmuir. 25(23):1351613521. Zhao C, Brinkhoff T, Burchardt M, Simon M and Wittstock G. 2009. Surface selection, adhesion and retention behavior of marine bacteria on synthetic organic surfaces using self assembled monolayers and atomic force microscopy. Ocean Dynamics. 59:305315. Zhou M, Liu Z, Wei Z, Liu C, Quiao T, Ran F, Bai Y, Jiang X and Ding Y. 2009. Development and Validation of Small diameter Vascular Tissue From a Decellularized Scaffold Coated with Heparin and Vascular Endothelial Growth Factor. Artifical Organs. 33(3):230239. Zhu AP, ChanPark MB and Gao JX. 2005. Foldable Micropatterned Hydrogel Film Made from Biocompatible PCLb PEG b PCL Diacrylate by UV Embossing. Journal of Biomedical Materials Resear ch Part B: Applied Biomaterials. 76B(1):7684. Zhu C, Ying D, Mi J, Li L, Zeng W, Hou C, Sun J, Yuan W, Wen C and Zhang W. 2008. Development of anti atherosclerotic tissueengineered blood vessel by A20regulated endothelial progenitor cells seeding decell ularized vascular matrix. Biomaterials. 29(17):26282636. Zilla P, Bezuidenhout D and Human P. 2007. Prosthetic vascular grafts: Wrong models, wrong questions and no healing. Biomaterials. 28:50095027.

PAGE 182

182 BIOGRAPHICAL SKETCH Chelsea Marie Magin, daughter of Gregory Allen Magin and Tina Marie Capizzi, was born in Ocala, Florida on a Saturday, in November Growing up, Chelsea enjoyed playing softball, building model rockets with her father and swimming with her family in the cool, clear waters of the Silver River. She was a safety patrol and a proud Hornet at Ward Highlands Elementary School, and a Fort King Middle School Falcon. In 1999, she entered the International Baccalaureate program at Ocalas Vanguard High School. While there, Chelsea maintained an unweighted 4.0 and played varsity softball. After graduating high school, Chelsea attended the University of Florida, where she maintained her excellent GPA, earning a bachelors degree with highest honors in m aterials science and e ngineering. During her undergraduate career, she served for four years in the Society of Women Engineers, holding many leadership positions in that organization including Vice President. Chelsea also established the inaugural Introduce a Girl to Engineering Day, a fixture at U Fs Engineers Week. She also founded the UF chapter of the Phi Sigma Rho Engineering Sorority. Towards the end of her undergraduate career, Chelsea interned for two summers with Kimberly Clark Corporation in Neenah, Wisconsin, submitting three patent appli cations during her time there. Chelse a entered the J. Cr ayton Pruitt Family Department of Biomedical Engineering at UF for her graduate studies, working in Dr. Anthony Brennans research group. Her graduate work focused on the use of topographically modif ied surfaces in applications ranging from marine antifouling to directed differentiation of human cells. Chelsea published several papers during her time in the Brennan research group. Her research has taken her all over the world, presenting posters and t alks at conferences

PAGE 183

183 from England to Japan. During her graduate studies, Chels ea founded the SWE Graduate Organization of Women Engineers, served as President of the University of Florida chapter of the Society for Biomaterials (SFB) and served on SFBs Nat ional Bylaws Committee. Chelsea is also a member of the Alachua County Habitat for Humanity Women Build Groups steering committee. She has been recognized for her contributions as the recipient of the 2010 UF Womens Leadership Council Phyllis M. Meek Spirit of Susan B. Anthony Award and the 2010 Attributes of a Gator Engineer Recognition Award for Leadership. Chelsea enjoys cycling, swimming, throwing pottery and gardening as a member of UFs Organic Garden Coop. Chelsea is engaged to Steve Kirschner, an attorney and fellow University of Florida graduate. They have two cats, a dog named Yoshi, matching road bikes and a love for Stevie Ray Vaughan. After graduating from the University of Florida, Chelsea will be working as a postdoctoral researcher in Dr. Kristi Anseths research group in the Department of Chemical and Biological Engineering at the University of ColoradoBoulder.