Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2013-08-31.

MISSING IMAGE

Material Information

Title:
Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2013-08-31.
Physical Description:
Book
Language:
english
Creator:
Gibson,Daniel James
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences, Biochemistry and Molecular Biology (IDP)
Committee Chair:
Schultz, Gregory S
Committee Co-Chair:
Frost, Susan C
Committee Members:
Yang, Thomas P
Bungert, Jorg
Lewin, Alfred S
Tuli, Sonal S

Subjects

Subjects / Keywords:
Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Genre:
Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility:
by Daniel James Gibson.
Thesis:
Thesis (Ph.D.)--University of Florida, 2011.
Local:
Adviser: Schultz, Gregory S.
Local:
Co-adviser: Frost, Susan C.
Electronic Access:
INACCESSIBLE UNTIL 2013-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043264:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 A CELLULAR AND MOLECULAR STUDY OF CORNEAL SCA R RING By DANIEL JAMES GIBSON 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 2011

PAGE 2

2 2011 Daniel James Gibson

PAGE 3

3 To Prometheus thanks for the fire

PAGE 4

4 ACKNOWLEDGEMENTS I thank the chair and members of my supervisory committee for their mentoring and guidance through the jungle of ambiguity that medical research can be. I would like to thank the National Eye Institute for its financial support through a Ruth L. Kirschstein N ational Research Service Award pre doctoral training g rant I would like to give special thanks to my in laws, whose tireless efforts helping with my children, granted my wife and I enough time for us both to finish our doctoral research, writing, and dissertation defense. And finally, I would like to thank my wife Soojung S eo, for the two healthy boys she has provided me and the love and support she continues to provide me

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ............................... 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURE S ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 13 1 INTRODUCTION ................................ ................................ ................................ .... 19 Corneal Fibrosis ................................ ................................ ................................ ...... 19 Current Clin ical Interventions ................................ ................................ .................. 22 The Molecular Causes of Fibrosis ................................ ................................ .......... 23 Transforming Growth Factor (TGF ................................ ............................... 23 Connective Tissue Growth Factor (CTGF) ................................ .............................. 25 The Current Molecular Model of Fibrosis ................................ ................................ 26 Nucleic Acid Therapeutics ................................ ................................ ...................... 27 Methods and Limitations of Macromolecular Delivery ................................ ............. 30 Projects ................................ ................................ ................................ ................... 31 2 IONTOPHORESIS OF CTGF ANTISENSE OLIOGNUCLEOTIDES (ASOs) INTO EXCIMER WOUNDED CORNEAS ................................ ............................... 40 Introduction ................................ ................................ ................................ ............. 40 Materials And Methods ................................ ................................ ........................... 41 Optimization of Iontophoretic Delivery of Reporter ASOs ................................ 41 Toxicity of Iontophoresed CTGF ASO ................................ .............................. 42 Therapeutic and Biochemical Efficacy of Iontophoresed CTGF ASO ............... 43 Haze Quantification via Macrophotography ................................ ...................... 44 Macrophotography ................................ ................................ ..................... 44 Haze evaluation and quantification ................................ ............................ 45 CTGF Quantification Via an E nzyme Linked I mmunosorbant A ssay ............... 46 Results ................................ ................................ ................................ .................... 47 Iontophoretic Delive ry of Reporter ASOs ................................ .......................... 47 In Vivo Wounding and Iontophoretic Treatment ................................ ............... 48 Discussion ................................ ................................ ................................ .............. 49 3 THE LOCATION OF CTGF PRODUCTION AND ACCUMULATION IN HEALING CORNEAS ................................ ................................ ............................. 63 Introduction ................................ ................................ ................................ ............. 63 Materials And Methods ................................ ................................ ........................... 64

PAGE 6

6 Excimer Laser Surgery ................................ ................................ ..................... 64 Tissue Harvesting, Processing, and Sectioning ................................ ............... 65 Immunohistochemical Staining ................................ ................................ ......... 65 Rabbit corneas ................................ ................................ ........................... 65 Reporter mouse corneas ................................ ................................ ........... 66 Transgenic Mice ................................ ................................ ............................... 67 Whole Mount Confocal Micrography ................................ ................................ 67 Micrography ................................ ................................ ................................ ...... 68 Gross Corneal Dissection and CTGF Transcript Quantification ....................... 69 Results ................................ ................................ ................................ .................... 70 Excimer Wounding ................................ ................................ ........................... 70 The Location of CTGF Protein Accumulation ................................ ................... 70 The Location of CTGF Promoter Activity ................................ .......................... 71 Gross Corneal Dissection and CTGF Transcript Quantification ....................... 72 Discussion ................................ ................................ ................................ .............. 72 4 HAZE FORMATION TIMELINE ................................ ................................ .............. 89 Introduction ................................ ................................ ................................ ............. 89 Materials and Methods ................................ ................................ ............................ 89 Excimer Wounding ................................ ................................ ........................... 89 Follow Up ................................ ................................ ................................ ......... 90 Macrophotography ................................ ................................ ........................... 90 Evaluation of Haze Development ................................ ................................ ..... 91 Molecular and Histological Analysis ................................ ................................ 91 Frozen sections ................................ ................................ .......................... 91 Paraffin sections ................................ ................................ ........................ 92 Immunohistochemistry ................................ ................................ ............... 92 Results ................................ ................................ ................................ .................... 93 Wound Macrophotography ................................ ................................ ............... 93 Immuno histochemistry ................................ ................................ ...................... 93 Discussion ................................ ................................ ................................ .............. 93 5 THE EMERGENCE OF A NEW THEORY FOR HAZE FIBROGENESIS ............... 99 Introduction ................................ ................................ ................................ ............. 99 Materials and Me thods ................................ ................................ .......................... 101 Cell Tracing Experiment Reporter Mice ................................ .......................... 101 Excimer Laser Wounding ................................ ................................ ............... 102 Reporter Mouse Tissue Harvesting, Processing, and Sectioning ................... 103 Galactosidase Detection ................................ ................................ ............. 103 Molecular and Histological Analysis ................................ ............................... 104 Frozen sections ................................ ................................ ........................ 104 Paraffin sections ................................ ................................ ...................... 104 Immunohistochemistry ................................ ................................ ............. 105 Result s ................................ ................................ ................................ .................. 106 Corneal Wounding ................................ ................................ .......................... 106

PAGE 7

7 Cell Tracing ................................ ................................ ................................ .... 106 Gross Histology ................................ ................................ .............................. 106 Immunohistochemistry ................................ ................................ .................... 107 Discussion ................................ ................................ ................................ ............ 107 6 CONCLUSIONS AND FUTURE DIRECTIONS ................................ .................... 123 A New Therapeutic Modality Validated ................................ ................................ 123 A New Standard in Visualizing and Reporting Corneal Haze ................................ 123 Connective Tissue Growth Factor in Healing Corneas ................................ ......... 124 Haze Fibrogenesis ................................ ................................ ................................ 126 Expected Clinical Impact ................................ ................................ ....................... 127 Closing Remarks ................................ ................................ ................................ .. 128 LIST OF REFE RENCES ................................ ................................ ............................. 129 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 139

PAGE 8

8 LIST OF TABLES Table page 2 1 Antisense oligonucleotides used in the anti fibrotic experiments. ....................... 54 3 1 real time polymerase chain reaction p rimers and probe sequences ................................ ................................ ................................ .......... 75 4 1 Color map used in Photoshop to generate the pseudocolored images. ............. 95 5 1 Primers for the mice with genetically labeled corneal epithelium ..................... 113

PAGE 9

9 LIST OF FIGURES Figure page 1 1 A schematic of the corneal cell layers. ................................ ............................... 35 1 2 Examples of the different classes of corneal opacification. ................................ 36 1 3 functional domains and the observed activities associated with them. ................................ .............. 37 1 4 The simplified fibrotic cascade. ................................ ................................ ........... 37 1 5 A more involved cellular and molecular model of the fibrotic response in the cornea. ................................ ................................ ................................ ............... 38 1 6 A schematic depiction of three general clas ses of functional nucleic acids ....... 38 1 7 General setup and principle behind iontophoresis ................................ .............. 39 2 1 Results from iontophor esis into intact rabbit corneas ................................ ......... 54 2 2 A representative excimer phototherapeutic keratectomy (PTK) wound in an ex vivo cornea. ................................ ................................ ................................ ... 55 2 3 The iontophoresis setup used for the ex vivo experiments. ................................ 55 2 4 Two schematics of the iontophoretic setups used ................................ .............. 56 2 5 Calculation of edema ................................ ................................ .......................... 56 2 6 An example of a band pass filter generated with non viable eyes. ..................... 57 2 7 Reporter single stranded deoxyribonucleic acid (ss DNA ) delivery into ex vivo globes ................................ ................................ ................................ ................. 58 2 8 The area of fluorescein staining was measured daily following PTK wounding with or without iontophoresis ................................ ................................ .............. 59 2 9 Corneal edema following antisense oligonucleotide ( A SO ) treatment and during healing ................................ ................................ ................................ ..... 60 2 10 The paired quantity of total CTGF mass extracted from each 8.0 mm corneal punch ................................ ................................ ................................ .................. 61 2 11 Paired examples of CTGF versu s scrambled ASO treated corneas .................. 61 2 12 Individual paired haze measurements ................................ ................................ 62

PAGE 10

10 3 1 The CTGF synthesis reporter mouse ................................ ................................ 75 3 2 Confocal m icrographs of anti CTGF immunofluorescent stained wounds 30 minutes post wounding ................................ ................................ ....................... 76 3 3 Confocal micrographs of anti CTGF immunofluorescent sta ined wounds 1 day post wounding ................................ ................................ ............................. 77 3 4 Two days post Wounding ................................ ................................ ................... 78 3 5 Day 3 Post Wounding ................................ ................................ ......................... 79 3 6 Day 4 Post Wounding. ................................ ................................ ........................ 80 3 7 Day 5 post wounding ................................ ................................ .......................... 81 3 8 Day 7 post wounding ................................ ................................ .......................... 82 3 9 Day 10 post wounding at the wound margin ................................ ....................... 83 3 10 CTGF promoter activi ty in an unwounded mouse cornea ................................ ... 84 3 11 The corneal endothelium is the primary location o f CTGF promoter activity ...... 85 3 12 The localization of CTGF protein in healing mouse corneas .............................. 86 3 13 Comparison of ribonucleic acid ( RNA ) mass yields and CTGF levels from g rossly dissected rabbit corneas ................................ ................................ ........ 87 3 14 A new theory to explain the different loci of CTGF synthesis versus accumulation and action ................................ ................................ ..................... 88 4 1 Pseudocolored images of day to day haze formation. ................................ ........ 95 4 2 smooth muscle actin localization ................................ ..................... 96 4 3 Higher power micrograph of the day 5 wound ................................ .................... 97 4 4 A schematic of where haze begins, how it spreads, and how it becomes more dense ................................ ................................ ................................ ......... 98 5 1 Evidence that the scar is formed in de novo synthesized tissue ....................... 113 5 2 The cellular representation of the haze generation conundrum. ....................... 114 5 3 An unresolved conundrum with the current prevailing theory ........................... 115 5 4 A schematic representation of de novo stromagenesis that occurs during development ................................ ................................ ................................ ..... 116

PAGE 11

11 5 5 The two transgenes which lead to an irreversible genetically labeled corneal epithelium ................................ ................................ ................................ ......... 116 5 6 Normal and scarred mouse eyes ................................ ................................ ...... 117 5 7 A single strong positive blue mass comprised of at most two cells in the anterior center of the stroma ................................ ................................ ............ 117 5 8 Epithelial invasion of the stroma during re epithelialization .............................. 118 5 9 Epithelial hyperplasia at the wound margin and migration of stroma derived cells into the wound interface ................................ ................................ ........... 119 5 10 The loss of epithelial attachment via a blistering like mechanism. .................... 120 5 11 The distribution of t enascin C during haze formation ................................ ....... 121 5 12 A potential mechanism for the observed epithelial invasion of the stroma. ...... 122 5 13 A new cellular model for the generation of corneal haze. ................................ 122

PAGE 12

12 LIST OF ABBREVIATION S ASO Antisense oligonucleotide a short single stranded oligonucleotide with a sequence which is complimentary to a portion of a targeted mRNA sequence. bp Base pair, referring to a number of hybridized nucleic acids in a polynucleic acid. EDTA Ethylenediaminetetraacetic acid a chelating agent used to inhibit metalloenzymes. ELISA Enzyme linked immuno sorbent assay an assay for quantifying the presence of a particular antigenic substance using antibodies and dye processing enzymes HEPES 4 (2 hydroxyethyl) 1 piperazineethanesulfonic acid a biologically compatible buffering agent. LASIK Laser assisted in situ kerat omileusis, one of many methods for reshaping the cornea with an excimer laser. Consists of creating a corneal flap, which is then lifted, and the underlying stroma is reshaped by the laser. mAb Monoclonal antibody, an immunoreactive solution composed of a single molecular species of an antibody directed against a molecular target. pAb Polyclonal antibody, an immunoreactive solution composed of a mixture molecular species of antibodies directed against a molecular target. pCTGF eGFP A genetically engineered mouse with the promoter of connective tissue growth factor driving the transcription of enhanced green fluorescent protein. PMSF Phenylmethylsulfonyl fluoride a serine proteinase inhibitor which alkylates the serine in the active site of the enzyme. PTK Phototherapeutic keratectomy, one of many methods for reshaping the cornea with an excimer laser. PTK removes a cylindrical or trapezoidal profile of tissue from the anterior surface of the cornea. PBS Phosphate buffered saline, a common biologically tole rable buffer. Tris T ris(hydroxymethyl)aminomethane a biologically tolerable buffer.

PAGE 13

13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Docto r of Philosophy A CELLULAR AND MOLECULAR STUDY OF CORNEAL SCAR R ING By Daniel James Gibson August 2011 Chair: Gregory S. Schultz Co chair: Susan C. Frost Major: Medical Sciences Fibrosis of damaged tissue is the basis for many chronic pathologies. Insights from the scarless healing of mammalian fetuses and marsupial pouch young have revealed that transforming growth factor Addition al work has revealed that connective tissue growth factor (CTGF) is a mediator of TGF like, activities in vitro ; including the key activities of proliferation collagen synthesis, and cellular differentiation. Fibrosis in the cornea re sulting from acute injury can lead to the generation of a light scattering scar. While it is known that CTGF is necessary for TGF fibrotic activities in cell cultures, it is not known whether this is the case for the formation of light sca ttering scars in vivo The experiments reported herein are separated into two general categories; those centered on studying a molecular effector of fibrosis (CTGF) and those studying the process of fibrosis itself. Throughout all of the experiments a s urgical excimer laser is used to generate an acute wound in the cornea of one of three animal models: 1) a normal rabbit, 2) a transgenic reporter mouse with enhanced green fluorescent protein under control of the promoter found upstream of CTGF, and 3) a

PAGE 14

14 transgenic galactosidase activity First, the role of CTGF in corneal fibrosis was tested by down regulation using an iontophoretically delivered antisense oligonucleotide (ASO). Following bilateral wounding, one cornea received a CTGF ASO while the contralateral cornea received a scrambled sequence control ASO. The ability of the CTG F ASO to reduce corneal haze was measured by photographing the wound and measuring the intensity of light reflection by the scar. The down regulation of CTGF was me asured by enzyme linked immuno sorbant assay (ELISA) analysis of homogenized corneas. The i ontophoresis was well tolerated with an immediate increase in edema as the only observed side effect. The source of edema was confirmed to be caused by the transfer of the highly anionic ASOs themselves, and not iontophoresis (p = 0.01). The amount of ha ze was slightly, but not significantly reduced at day 7 (p = 0.23), but was significant by day 14 (p =0.04). Iontophoresis of CTGF ASO was capable of reducing CTGF protein in the cornea at both days 7 (p = 0.05) and 14 (p = 0.002). The reduction of CTGF protein and corneal haze following treatment with a CTGF ASO validated that CTGF has a role in corneal fibrosis. In order to better understand the role of CTGF in fibrosis, the location of its synthesis and accumulation was measured in healing corneas. Fi rst, t he location of CTGF protein was observed during healing in rabbit corneas using immunofluorescent staining with a monoclonal antibody to CTGF. Next, a transgenic reporter mouse with romoter was used to observe which cells are actively transcribing the CTGF gene The corneas of the

PAGE 15

15 reporter mouse were wounded, and the eyes were harvested and histologically sectioned. The location of CTGF protein was observed by staining the reporter mouse sections with a biotinylated form of the same monoclonal antibody used in the rabbit experiments. Since the synthesis was reported with a green fluorescent protein, the CTGF protein was detected using a Texas Red labeled avidin, thereby enabling sim ultaneous observation of CTGF synthesis (green) and protein localization (red). During healing, CTGF protein was detected in all three cellu lar layers of the cornea, but was most abundant in the basal epithelium. Surprisingly, the predominant source of C TGF synthesis in the cornea was the endothelium, with the levels of synthesis in the epithelium and stroma being so low that they could not be observed in the reporter mouse. These observations invalidated the initial hypothesis of fibroblast derived synt hesis and binding of CTGF during fibrosis, and gave rise to a new, more complex, theory that CTGF is synthesized by the endothelium and then bound by the epithelium to have its pro fibrotic effect. Following re epithelialization the cornea is still clear, but within the weeks that follow it can become opaque. Not much is known about how and where the opacification starts and how it matures to its final level of intensity. The next project investigated how and where haze starts, and how it changes with tim e using a novel photographic technique which enables visualization of the entire scar with sta ndard photographic equipment. The emergence of haze was observed by daily photography of wounds as they developed haze Next, a series of haze images were compa red with the immunofluorescent localization of light reflecting myofibroblasts during the period of haze maturation. Haze consistently began as a thin ring at the wound periphery.

PAGE 16

16 Occasionally, another separate region of haze emerged in the center of the wound with a haze less region in between the center region haze and the ring of haze at the margin. The haze spread with time from the these regions of nucleation into the adjacent haze less regions with time The regions with haze present intensified w ith SMA) staining in the wound, mirrored the pattern of haze initiation and spread during haze formation. SMA was also substantially present in the basal epithelium, but h ad a different intracellular distribution that that of the cells in the stroma. Finally, the intensification of haze was reflected in the number of myofibroblast layers that accumulated below the epithelium. These observations required a drastic change i n the theory of haze fibrogenesis following acute injury; directing attention away from the fibroblasts in the underlying stroma towards either the fibroblasts in the peripheral stroma or even the epithelial cells themselves. Finally, s ome previously known facts about wound healing and the new data generated by some of the work presented herein led to a conundrum concerning the current theory of fibrosis in the cornea Given that epithelial cells are the first cells to contact the non w ounded stromal surface, that they are the primary source of CTGF SMA, and that the formation and spread of haze mirror the pattern of re epithelialization, a hypothesis that the epithelial cells themselve s differentiate into the stromal myofibroblasts was formed to resolve the conundrum. An alternative hypothesis, that the peripheral stromal cells migrate into the wound area and displace the epithelium in a manner reminiscent of de novo stromagenesis was also formed. A reporter mouse with a genetically labeled epithelium

PAGE 17

17 was used t o observe whether cells from the epithelium are present in the stromal scar following acute injury. Concurrently, gross histological analysis was performed on hematoxylin and e osin stained rabbit corneal wounds from the period prior to haze formation to seek evidence of any interchange of cells between the epithelium and stroma or of epithelial displacement by the stroma derived cells. Finally, the presence and localization of tenascin C, a candidate marker for the process of epithelial to mesenchymal transition (EMT), was observed in wounded rabbit corneas during the period of haze formation via immunofluorescent staining. Epithelium derived cells were found below the scar in the wounded reporter mouse cornea but were surrounded on all sides by stroma derived cells. The gross histological analysis revealed that during re epithelialization some epithelial cells invade d the residual stroma along incompletely removed stromal la mellae (days 1 and 2). The gross histological data also provided evidence for the migration of peripheral fibroblasts into the wound (day 3) and displacement of the epithelium by a blistering like mechanism (day 5). Tenascin C was present during haze for mation and its distribution and spread mirrored both the spread SMA. In total, these data support the hypothesis that epithelial cells do become part of the stroma, but only to a minor degree. The data supp ort a mechanism similar to the de novo stromagenesis with the peripheral fibroblasts as the primary source of cells that become light reflecting haze. The data also cast the use of protein markers as indicators of the EMT process. In conclusion, the work reported here validates the proposed role of CTGF in the generation of light reflecting scars, but it calls into q uestion the source of the pro fibrotic CTGF and its locus and mechanism of action. The current evidence suggests that

PAGE 18

18 CTGF synthesized by the endothelium could possibly diffuse through the edematous stroma and then bind to the basal epithelium. While, ep ithelial cells were found in the stroma following wound healing, they were vastly outnumbered by stroma derived cells, undercutting an EMT based theory for haze formation. With the new data presented here, the best supported theory explaining the formatio n of haze is that the peripheral SMA positive cells. The epithelium is then displaced by a blistering like mechanism and the haze myofibroblasts populate the sub epithelial space and proliferate, leading to the spread and intensification of the light reflecting scar.

PAGE 19

19 CHAPTER 1 INTRODUCTION Corneal Fibrosis Fibrosis is the biological response to damage or disruption of cellularized tissue whereby the cells prolifera te, differentiate, and synthesize new extracellular matrix (ECM) molecules to patch the damage. While this response does replace or bridge lost or rent tissue, the newly rege nerated tissue typically does not possess the same structure as the uninjured tis sue, nor does it regain the function of the once intact tissue. Fibrotic healing is responsible for prolonged pathological conditions in the kidneys 1 2 liver 3 heart 4 5 peritone um 6 skin 7 and in the eye 8 11 While not life threatening, fibrosis anywhere in the eye as a result of injury, infection, or surgery is highly problematic for an estimated 80% of all sensory input 12 The eye provides at least three visual functions, it refracts and focus es incident light, it transduces the light into a neurological signal, and it provides some initial visual information processing. The eye is functionally divided into two segments, the posterior segment where photo transduction and information processing take place, and the anterior segment where refraction and focus/accommodation take place. While fibrosis can impair each of these functions 9 13 15 the scope of this project will be restricted to fibrosis affecting the clarity of the cornea. The cornea is a highly innervated, avascular, tissue composed of three cellular layers ( Figure 1 1 ). Each of the layers has a critical f unction in the overall health and/or stability of the cornea. The epithelium provides a highly impermeable barrier to external contaminants, the stroma is largely responsible for the mechanical strength and stability

PAGE 20

20 of the cornea, while the endothelium i dehydrated state, and thereby its clarity. Trauma to the cornea can lead to detrimental consequences to both the shape and clarity of the cornea. The shape of the cornea can be perturbed by the irregular replacement of the excised or injured tissue; this process is largely blamed for refractive regression following refractive surgery. Trauma can also lead to decreased light transmission by one of three mechanisms: 1) turbidity caused by edema, 2) matrix protein denaturation /coagulation and/or chemical modification or 3) by light occlusion and scattering due to phenotypic changes in the cells in the cornea In its normal clear state, the cornea is highly dehydrated. The dehydrated state is maintained primarily though the active transportation of ions out of the cornea by the endothelium. Diseases or defects that impact the endothelium can decrease the ion transport activ ity and cause the cornea to retain water and become less clear 16 18 While chronic cases do exist, edema following stroma penetrating wounds ( Figure 1 2A) tends to resolve once the epithelial bar rier is re established (Figure 1 2B ). This class of light interference does not require a viable cornea and can occur on the time scale of minutes in ex vivo corneal tissue that is permitted to absorb water. The turbidity that accompanies e dema is transi ent and reversible if the cornea is dehydrated once again Another class of light interfering lesions in the cornea can be attributed to denaturation of the proteinaceous matter within the cornea. Alkali chemical or thermal burns to the cornea can create light obscuring lesions within a minute or two in both viable (Figure 1 2C) and non viable tissue (not shown) Coagulative opacification can also occur as a consequence of bacterial keratitis, though the mechanism is drastically

PAGE 21

21 slower than the chemical and thermal burn based wound. At present, coagulative lesions are the most severe corneal opacifying lesions and are at present completely irreversible. While the previous two types of opacification can interfere with light transmission through the cornea neither is an example fibrosis. The fibrotic response requires a specific level of damage to the tissue and is a process that requires weeks of wound The formation of a cellular scar is a much slo wer process than the other two classes and is therefore more amenable to therapeutic intervention In the cornea, lesions only affecting the corneal epithelium T hey do however, result in apoptosis of the s tromal fibroblasts directly beneath the injury 19 20 The scrape is healed by proliferation of the epithelial cells and the newly acellularized zone in the stroma is once again repopulated by proliferation and migration of fibroblasts leaving no vision impairing artifacts 21 22 Injuries which penetrate into the stroma also result in apoptosis of the immediately juxtaposed fibroblasts H owever the cells which eventually repopulate the acellularized zone synthesize excessive extracellular matrix and can transdifferentiate into the cells with light reflecting subcellular structures 23 24 forming what is clinically referred to as sub epithelial haze (Figure 1 2D) Minimally, sub epithelial haze disrupts normal vision by interfering with the transmission of light, ability to drive at night 25 26 However, if the scarring is severe, the scar can preclude the transmission of light through the cornea leading to corneal blindness. The degree of haze following a stromal penetrating injury has been found to be directly proportional to

PAGE 22

22 the v olume of stromal tissue removed 27 the concentration of TGF 28 degree of basement membrane disruption 29 and the degree of correction attempted during laser surgery 30 31 Current Clinical Interventions At present, the methods used to decrease corneal haze are topically applied steroids or anti metabolite drugs which indiscriminately respond to growth factor and cytokine signaling. Given their safety profile, anti inflammatory steroids have garnered attention as potential anti fibrotic agents. Steroid s have been demonstrated to reduce cell proliferation in vitro 32 and stromal collagen synthesis in rabbit refractive surgery models 33 Initially, a prospective, double masked clinical trial to te st the efficacy of topically applied dexamethasone 0.1% in preventing corneal haze, found no significant effect on scar reduction 34 However, additional investigations have found that the efficacy of steroids appears to vary with the method 33 and timing 35 of installation, and the magnitude of refractive correction sought 35 37 suggesting that the corneal fibrosis in a sub set of patients is addressable by the immediate installation of steroids. The current gold standard of anti fibrotic drugs is topically applied mitomycin C (MMC). Mitomycin C becomes a strong alkylating and cross linking agent after it is biologically reduced by intercellular enzymes, and it is through this activity that it is proposed to have its drug effect 38 39 At pres ent, it is hypothesized that MMC prevents fibrosis by crosslinking the primary amine moieties on guanosines present in 5' CpG 3' base pairs and thereby preventing DNA synthesis and subsequent cellular proliferation. While MMC has been found to be effective at reducing corneal haze, there exist conflicting reports about its toxicity to the corneal endothelium 40 In addition, MMC is

PAGE 23

23 associated with significant toxicity of the ocular surface and narrowing or blocking the tear drainage ducts (a k a punctal stenosis ) 41 43 Given the high function and fragility of the endothelium, lower risk compounds and techniques are actively being sought. A key strategy to mitigating the risk of future anti fibrotic therapeutic s is to have a more targeted approach; nullifying the specific molecular pathways that give rise to the scar. However, i ncreased targeting precision requires more knowledge about the molecular pathways unique to, and necessary for, the generation of fibrotic lesions. The Molecular Causes of Fibrosis Knowledge about the molecular biology of fibrosis began with insights from the scarless healing of mammalian fetuses and marsupial pouch young (collectively 44 45 Initially, investigators sought to determine if the womb environment bestowed anti fibrotic pro tection upon the fetus by grafting and wounding maternal ovine skin onto fetal lambs. The adult tissue still scarred and immunohistochemistry revealed that the adult tissue, but not the fetal tissue contained transforming growth factor ound the scar. Further experiments revealed that a fetus The experiments collectively revealed that the presence of increased levels of TGF triggers a fibrotic versu s a regenerative wound healing response and that the womb does not provide a protective role against fibrosis. The results from these experiments indicated that the appropriate therapeutic approach to prevent scarring would need to mitigate TGF fi brotic activities. Transforming Growth Factor The transforming growth factor 1 2 ) are known to trigger the fibrotic wound healing response, the third

PAGE 24

24 3 ) is proposed to inhibit the activities of the other two 46 1 has been shown to autoinduce its own expression 47 and TGF 2 has been shown to induce the expression of all three TGF 48 Addition of TGF 1 to cultures has been demonstrated to be sufficient to induce transdifferentiation of fibroblasts 49 and lens epithelial cells 15 to the myofibroblast phenotype. TGF unwounded mouse epidermis 50 the unwounded corneal epithelium in adult mice 51 and rats 52 and in human tear fluid 28 53 54 Following photorefractive keratectomy (PRK) of adult rat eyes, whole corneal TGF 2 mRNA was found to be 300 fold higher than in untreated corneas 52 Hypothetically, to prevent corneal haze the myofibroblast differentiation which follows TGF stimulation is key However, given the eff icacy of ant i mitotic agents like mitomycin C targeting its prolifer ative activity may also have beneficial effects. Initial work to mitigate TGF fibrotic activities in vivo via t argeting its receptor led to positive effects on the targeted tissue, but with substan tial side effects on epithelial tissues 55 The key ef fect that TGF hyperphosphorylation of the retinoblastoma protein (pRB) and thereby inhibit G 1 /S transition 56 In the presenc e of epidermal growth factor (EGF), renal epithelial cells proliferate into hyper cellular masses ( hyperplasia ). W hile these cells exposed simultaneously to EGF ant TGF will become enlarged (hypertrophic) but will not proliferate 56 The TGF inactivating proteins 56 The effects seen with TGF on in vivo and knowledge of its epithelial growth inhibiting function leads to the conclusion that strong inhibition of TGF activity essentially recapitulates a retinoblastoma like disease state leading to

PAGE 25

25 the generation non transformed epithelial tumors Given this observation, therapies targeting downstream effectors of TGF have become the dominant strategy. Connective tissue growth factor (CTGF) is one downstream effector of TGF fibrotic activities which has been i dentified 57 59 In fibroblast cultures, CTGF has been demonstrated to be both nece ssary and sufficient for TGF contraction 60 61 proliferation 58 59 collagen production 58 59 and the differentiation of cells into myofibroblasts 58 59 and is therefore at the forefront of gene targeted anti fibrotic therapies. Connective Tissue Growth Factor Connective tissue growth factor (CTGF) is a 38kDa protein and the founding member of the CTGF/Cyr61/Nov (CCN) fami ly of secreted cytokines which has been demonstrated to mediate TGF 58 59 62 Connective tissue growth factor was first discovered as a mitogen that was co purified with anti platelet derived growth factor (PDGF) antibodies 62 The investigators discovered that CTGF was directly bound to the PDGF antibody, not PDGF itself, and that CTGF itself was directly responsible for the majority of the PDGF like mitogenic act ivity. In both skin and corneas, CTGF and CTGF mRNA levels were both low in unwounded tissue, but were both elevated in wounded tissues at time point s following TGF 7 12 57 63 64 These observations strengthen ed the relationship between TGF models and in humans. Proteolytically processed fragments of CTGF have been found in biological fluids obtained from porcine uterine secretory fluids and in the vitreous of patients with proliferative di abetic retinopathy 65 67 Connective tissue growth factor has fo ur domains with sequence similarity to other growth factors (Figure 1 3). Experiments with

PAGE 26

26 s y nthetically bisected CTGF have demonstrated that the individual fragments possess distinct activities 68 69 The amino terminal fragment comprised of the insulin like growth factor binding (IGFB) and von Willebrand factor binding (VWB) domains, p ossesses the observed differentiation and collagen synthesis stimulating activities T he c arboxyl te rminal half of CTGF, comprised of the thrombospondin 1 (TSP1) and cysteine knot (CT) domains, possesses the observed fibroblast mitogen activity. It has yet to be determined whether the fragmentation has a regulatory role, but the activities found in both halves can result in detrimental clinical outcomes following vision correcting laser surgery. A simple hypothetical molecular model of how the differential activities of CTGF might be regulated through proteolytic processing is presented in Figure 1 4. The Current Molecular Model of Fibrosis Building a model of the molecular pathway responsible for the fibrotic wound healing response requires a combination of macroscopic observation of the type of wound necessary for scarring and the known localization a nd activities of pro fibrotic molecules. As previously mentioned, a lesion which does not penetrate into the corneal stroma will not scar For lesions that do penetrate into the stroma, fibrosis is constrained to the location where the strom a is both dis rupted and is placed into communication with the tear film and/or epithelium The restriction of fibrosis sol ely to the region of epithelium to stroma contact is most apparent when comparing scars from transepithelial laser surgery with th ose from laser a ssisted in situ keratomileusis like ( LASIK like ) procedures In the trans epithelial lesion, the entire wound region is capable of scarring whi le in lamellar keratectomy surgeries where the removed button of tissue is replaced, or in LASIK procedure s on ly the margins where the trephine cut or LASIK flap cut through the epithelium form a scar 33 70 The exact nature of this limitation

PAGE 27

27 is still not understood, but it is also seen in other tissues In the retina, exogenously added TGF could not induce fibrosis in intact retinal detachments, but did if the detach ments had an accompanying tear 71 In the retinal detachments with a tear, the fibrosis was restricted to the region of the tear indicating that tissue disruption is a co requisite of TGF fibrotic activity. When these facts are integrated together, a more complex cellul ar and molecular model begins to emerge ( Figure 1 5 ) where TGF derived from the tears or the epithelium serves as the initial signal and the stromal penetration serves to sensit ize the corneal cells to TGF Shortly after the stroma penetrating injury, it has been reported that the underlying keratocytes apoptose, leaving an ac ellular region below the wound 16 17 With time, the acellular re gion can become re populated by stromal keratocytes which chemotax into the wounded region It is hypothesized that these cells are the locus of the first wave of the pro fibrotic TGF response (endocrine) and are the loci of subsequent TGF nthesis and activity (autocrine) Ultimately, these are the cells that are proposed to form the light reflecting haze. Given this model, the residual stromal keratocytes have become the primary target of CTGF targeted mRNA ablation therapeutics. Nucleic Acid Therapeutics While n ucleic acids are renowned for their capacity to store information they have also been found to be capable of many of the same functions that proteins possess including catalysis 72 79 and macromolecular recognition 74 80 82 Many functional nucleic acids were discovered in naturally occurring biochemical pathways (ribozymes, antisense, RNAi) 83 whil e others were synthetically discovered (deoxyribozymes, aptamers) 79 82 and in the case of aptamers, were found later to occur naturally 74

PAGE 28

28 Below, these nucleic acids will be discussed in terms of their therapeutic application, not in terms of their natural biochemistry. There are two general classes of therapeutic nucleic acids, th ose that can act on their own and those that require endogenous macromolecular complexes. Ribozymes, deoxyribozymes, and RNA or DNA aptamers (Figure 1 6A&B, respectively) are examples of nucleic acids that can, themselves, have therapeutic activity requir ing at most a divalent cation (e.g. Mg 2+ ) 72 75 79 84 Anti sense oligonucleotides (ASOs) and the various RNAi and macromolecular complexes which when combined, possess the therapeutic activity 85 (Figure 1 6C) The general classes are further subdivided based on their mode of action. Most target mRNA for destruction or translational silencing (siRNA, miRNA, ribozymes, deoxyribozymes, and antisense). Others have been demonstrated to guide heterochromatic gene silencing, DNA methylation, or even genomic DNA excision and exclusion ( other RNAi based) 86 Finally, aptamers have the most disparate function of the therapeutic nucleic aci ds in that they do not employ Watson Crick base pairing, but specific geometric arrangement of hydrogen bonds, ionic bonds, and ring stacking interactions 87 The importance of considering the mode of therapeutic action is that any the best target(s). If the target is a pre existing (i.e. already transcribed) protein, employing mRNA targeting therapies would not h ave any effect on that pre existing effector. Whereas, if the target is a yet to be transcribed/translated gene, using an

PAGE 29

29 aptamer to neutralize the final product, instead of neutralizing the mRNA transcript would suffer from the target amplification which occurs as a result of a single mRNA molecule being translated multiple times. The final important distinguishing characteristic is the location where the nucleic acid therapy can have its effect. Simply put, the targeted molecule, its co requisites, an d the drug must be co located for therapeutic effect. For instance, the mRNA ablating technologies are only active intracellularly, since that is where the target (i.e. mRNA) is located, whereas aptamers can be directed to targets on the cell surface or i n the extracellular milieu. Finally, drug delivery is a significant hurdle and therefore the locus of activity constraint can have a significant impact on the efficacy of a candidate therapy if effective concentrations of the drug cannot be delivered when and where it needs to be. Nucleic acid therapies, as a whole, have the same strategic limits as all pharmacological agents. If the target of interest is a highly stable protein (i.e. a structural protein) with a long half life, a therapeutic reduction of the target may difficult o r impossible even with drastic reductions in mRNA levels. Alternatively, if the target is a highly expressed and highly active enzyme, even a drastic reduction in the total target levels may not have any biological effect. In short, some biological casca des have built nucleic acid therapies. The overall sensitivity of the targeted biological response to levels of the targeted macromolecule must be considered very e arly on in selecting a protein for therapeutic reduction.

PAGE 30

30 To date, one antisense oligonucleotide (ASO) therapy has been approved by the cytomegalovirus (CMV) antisense oligonucleotide approved for treati ng CMV retinitis in patients with acquired immune deficiency syndrome ( AIDS ) 88 Anti fibrotic ASO s targeting TGF been used to control fibrosis in rat kidneys 89 mouse and rat livers 90 95 and mouse skin 96 in a research setting More recently, an anti fibrotic CTGF ASO has been tested in phase II clinical trials for its ability to reduce the visual appearance of scars from acute surgical skin wounds in humans 97 Methods and Limitations of Macromolecular Delivery Due to the size and highly negative charge of nucleic acids, delivery and cellular uptake are difficult using traditional means such as systemic delivery or local injections. Additionally, due to the small tissue volume and high function of the cornea, injections are not desirable. Using iontophoresis, which is based on the same principles behind gel electropho resis, the highly charged nucleic acids can be selective ly delivered into the corneal tissue. The factors affecting migration in an electrophoresis gel are the same that can impact the migration of a drug into the tissue; larger, lower charged molecules m igrate slower than smaller, higher charged molecules and molecules move slower in a more dense tissue than a less dense one. Currently, iontophoresis is used clinically for transdermal delivery of small molecule analgesics such as lidocaine. Initial trial s using the conditions published for use with small molecules such as lidocaine and epinephrine were unable to deliver ssDNA oligonucleotides into ex vivo rabbit corneas 98 The increased mass o f oligonucleotides led to the need of an the drug into the tissue. The use of a higher electric field (or current density) is needed to impart the force

PAGE 31

31 necessary to surmount the barrier posed by the cornea Consequentially there is an accompanying increase in electrolytic alkylization of the delivery reservoir (and acidification of the receiving chamber) which n ecessitates a robust buffering strategy Finally, the use of a dual phase system in which the drug is solubilized in a dense, non ionic solution (i.e. 15% sucrose) enables the drug to sink closer to the target tissue, phase system also allows the drug, not the buffer ion, to be predominan tly delivered. When all of the aforementioned modifications to the standard form of iontophoresis are employed, iontophoresis is capable of delivering a 20 nucleotide single stranded DNA reporter oligonucleotide into the cornea without significant, observ able, damage 98 101 A model of the modified iontophoretic setup is depicted in Figure 1 7A Schematics of t he basic mechanics behind why topical application fails, and why iontophoresis succeeds, as a means of delivery are depicted in ( Fig ure 1 7 B&C, respectively) Projects Much work has been done to test the role of CTGF in scarring using fibroblast cell cultures with the fibrosis related activities of proliferation, collagen synthesis, and differentiation as the primary readouts. No work has been reported on the role of CTGF in actually modulating the formation of light scattering entities in corneas following acute mechanical injury. The first attempts with a topically applied CTGF ASO had no observable effect on haze formation, but topical application is not expected to lead to efficient uptake of the oligonucleotide. The first project described herein seeks to deliver a CTGF ASO into acutely wounded rabbit corneas and determine 1) whether CTGF protein is red uced and 2) whether there is a decrease in haze formation in the same eyes. This project will test whether CTGF is necessary for scar formation and will

PAGE 32

32 test the feasibility of iontophoretically delivered ASOs as an effective therapeutic modality. The se cond project seeks to identify the locus or loci of CTGF synthesis and activity within corneas which are healing from acute excimer laser wounds. To date, fibroblasts have been hypothesized to be both the source of synthesis and activity of CTGF indicatin g an autocrine or regional endocrine signaling mechanism. Initial attempts to measure CTGF protein localization suggested a heavy presence in the epithelium 63 but signi ficant immunohistochemic al over exposure of the sections synthesis of the observed CTGF during h ealing in the cornea is known. A combination of improved immunofluorescent staining and confocal micro scopy of healing corneas will reveal the distribution of CTGF protein during healing of acute wounds. In order to reveal the locus or loci of CTGF synthesis, a mouse model with enhanced green fluorescent protein under control of a CTGF promoter will be used. The green fluorescent protein transgene does not possess a secretion signal, leading to an accumulation of green fluorescence in cells with CTGF promoter activity. The results from this reporter mouse were further validated by q uantitative real time reverse transcriptase polymerase chain reaction ( q RT rtPCR ) analysis for CTGF transcripts in corneas which were grossly dissected into the three separate cellular layers The data provided by these experiments are necessary for two key reasons: 1) to provide a better understanding of the role CTGF has in wound healing (protein location) and 2) to ensure that the anti mRNA therapies are targeted to the correct cells (synthesis location).

PAGE 33

33 The next project sought to investigate the pattern, location, and timing of haze formation and how the distribution of biochemical markers for light reflecting cells were likewise distributed during the formation of the light reflecting scar. A time series of macrophotographic images of the wounded corneas was used to visualize the initiation and change of haze as it developed. Standard immunofluorescent staining and gross histology were used to observe where and when these markers arise and wh ether their distributions mirror the pattern of the light reflecting scar. T he final project sought to test the growing body of evidence that the epithelium has a greater role in fibrosis than initially appreciated. The two possible roles id entified are that the epithelial cells provide a the observed restriction of fibrosis to only the disrupted tissue or that the epithelial cells themselves directly contribute cells and/or ECM to the fibrotic mass. For thi s final project the hypothesis that the epithelium directly contributes cells to the fibrotic mass by the process of epithelial to mesenchymal transition (EMT) was tested The key experiment used to test the EMT hypothesis was a mouse model with genetically labeled corn eal epithelial cells which would determine whether any epithelium derived cells are present within the stroma after the light reflecting scar has been established. Concurrently, g ross histology of wounded rabbit corneas during the time prior to haze forma tion was used to investigate whether there was any grossly observable evidence of epithelial cells crossing into the stroma or vice a versa. Finally, immunofluorescent staining for markers associated with, but not necessarily indicative of, EMT was perfor med during the time period of haze formation

PAGE 34

34 In total, the projects seek to provide a better understanding of the role of CTGF in fibrotic would healing and the cellular and molecular biological processes that give rise to pathological scarring Improv ements in understanding the molecules that are believed to communicate and orchestrate the pro fibrotic wound healing response are necessary to improve design and testing of novel anti fibrotic therapies Knowledge of the cellular players and the changes they undergo during fibrosis is key to identifying new potential therapeutic modalities. Only by integrating knowledge about the source and localization of the key molecular effectors and the timing and location of the cellular responses to them, is a com prehensive, well integrated, systemic understanding of fibrosis possible. Only after the fibrotic response is known in such detailed terms will scientists and engineers be able to design highly targeted therapies that can steer a wound away from fibrosis and towards regeneration.

PAGE 35

35 Figure 1 1. A schematic of the cornea l cell layers

PAGE 36

36 A B C D Figure 1 2 E xamples of the different classes of corneal opacification. A) Edema 1 day after a wound, B) the same eye as in A), but 2 days after the wound. Note the annular region that has regained clarity after being re epithelialized while the center remains turbid. C ) C oagulation and modification of stromal proteins from a n alkali burn D ) cellular h aze that takes a week or two to form.

PAGE 37

37 Figure 1 3. and the observed activities associated with them. Figure 1 4 The simplified fibrotic cascade.

PAGE 38

38 Figure 1 5 A more involved cellular and molecular model of the fibrotic response in the cornea. A B C Figure 1 6. A schematic d epiction of three general classes of functional nucleic acids. A ) Ribozymes and deoxyribozymes possess enzymatic activities themselves. B ) DNA or RNA aptamers are capable of selective macromolecular recognition by themselves A) RNAi and anti sense oligonucleotides require macromolecular complex es.

PAGE 39

39 A B C Figure 1 7. General setup and principle behind iontophoresis. A) Schematic of a transepithelial iontophoresis chamber. The blue is the oligonucleotide dissolved in sucrose, the yellow is a suitable buffer, the grey is a buffer soaked gauze in contact with the eyel id. B) The basic mechanics of topical delivery, there are forces resisting delivery (F R ), and forces facilitating delivery (F D ) due to diffusion. C) With iontophoresis, an additional force (F E ) is applied to the drug which enables it to overcome the tis sue barrier.

PAGE 40

40 CHAPTER 2 IONTOPHORESIS OF CTGF ANTISENSE OLIOGNUCLEOTIDES INTO EXCIMER WOUNDED CORNEAS Introduction Oligonucleotide based technologies are a relatively new class of drugs capable of highly targeted therapies that are expected to be drasticall y less toxic than small molecule drugs. If the presence of a protein has been implicated in a particular pathological process, these oligonucleotide technologies can be used to degrade the covery that CTGF was necessary for TGF fibroblasts to differentiat e into myofibroblasts made it a prime target for a potential anti fibrotic oligonucleotide therapy. Cell culture experiments further strengthened CTGF as a candidate anti fibro tic target when the use of CTGF antisense oligonucleotides (ASOs) on TGF capable of reducing the resultant contraction 60 61 collagen synthesis 63 and myofibroblast differentiation 60 The excitement from these early experiments was greatly diminished when unpublished preliminary in vivo experiments of topical applied oligonucleotides to excimer wounded rabbit corneas had no therapeutic effec t (2004). Translating the CTGF ASO cell culture results into in vivo models has been met with the same key obstacle as most other oligonucleotide therapies: delivery. referred to as iontophoresis has garnered attention as a means of oligonucleotide delivery. Berdugo and others were the first to report successful delivery and short term bioavailability of fluorescently labeled reporter ASOs into rat corneas 102 To date, no reports have described pharmacologically or therapeutic ally efficacious delivery of oligonucleotide therapies into the cornea. However, Kigasawa and others have recently

PAGE 41

41 reported success in both delivering a reporter siRNA into rat epidermis and in reducing IL 10 mRNA levels using an iontophoretically deliver ed anti IL 10 siRNA 103 While they were successful in obtaining the desired biochemical effect of reducing IL 10 mRNA, no mention was made about the therapeutic effect of the obser ved reduction. We decided to test the ability of an i ontophoretically delivered CTGF ASO to reduce CTGF protein and light scattering myofibroblasts in a model which uses a surgical excimer laser to create reproducible acute wounds in rabbit corneas. Mater ials And Methods Optimization of Iontophoretic Delivery of Reporter ASOs In previous work, it was found that iontophoresis of 5.0 mA for 5 minutes using our 1.1 cm diameter eye cup was sufficient to deliver a reporter oligonucleotide (ISIS 13920) 104 105 into the deepest layers of intact cornea s in the eyes of live rabbits (Figure 2 1 ). Given that for this wound healing model we will be delivering into an ablated cornea, decreased delive ry current and/or time is possible to obtain more optimal results. To this end a series of iontophoretic deliveries were performed in the corneas of ex vivo rabbit eyes acutely injured with an excimer laser First, frozen whole rabbit globes (Pel Freez, LLC., Rogers, AR) were thawed and immobilized on a piece of polystyrene foam. A central wound was created on the immobilized globes Figure 2 2 depictes a represe ntitve PTK wound in an ex vivo rabbit eye. Next, using the setup depicted in Figure 2 3 FAM ) labeled reverse phase high performance liquid chromatography ( rpHPLC ) purified ssDN A with the same sequence as ISIS 13920 ( Table 2 1, Integrated DNA Technologies, Inc., San Diego, CA) was subjected to iontophoresis for 5 minutes using currents of 0.0, 1.0, 2.0, or 3.0

PAGE 42

42 mA. The corneas were fixed overnight, then cryoprotected, and finally embedded in optimal cutting temperature (OCT) compound for cryosectioning. The sections were mounted with 4',6 diamidino 2 phenylindole ( DAPI ) counterstain containing medium and the labeled ssDNA was observed by direct fluorescence microscopy. All micro graphs were obtained using the same imaging parameters allowing relative quantitative comparisons amongst the micrographs. Toxicity of Iontophoresed CTGF ASO All of the rabbits used herein were treated in a manner consistent with the Association for Research in Vision and Op hthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. In order to establish whether either the iontophoresis or the ASOs themselves were well tolerated we conducted some preliminary experiments mea epithelialization and edema. Each rabbit was anesthetized using inhaled isoflurane and each eye was topically anesthetized using one drop of tetracaine (Bausch & Lomb, Tampa, FL). The eye was exposed and the central t hickness of the cornea was measured using a clinical grade ultrasonic pach ymeter (n = 5 central measurements per eye). Each eye then thickness of the ablated cornea was measure d by ultrasonic pachymetry again, then iontophoretically treated as described below. A total of 3 rabbits received bi lateral central corneal wounds and were treated with i ontophoretically delivered CTGF ASO 95 (ISIS 124189) in on e eye and a scrambled ASO (ISIS 29848) in the contralateral eye. Details about the se ASOs may be found in Table 2 1. Based on preliminary data with unwounded eyes, iontophoresis was performed using a current of 5.0 mA for 5 minutes with the apparatus set up as depicted

PAGE 43

43 in Figure 2 4 A treatment and scrambled control ASOs. An additional three rabbits received the same wound in one eye and no further intervention. Re epithelialization was measured by daily fluorescein staining and macrophotography. Re epithelialization was analyzed two different ways. First, the day the cornea fully epithelialized was noted (clinical). Second, the daily area that remained wounded was measured using digital photographs of the fluores cein stained corneas (technical). In these same rabbits, corneal edema was quantified via daily pachymetry measurements of the central cornea until re epithelialization. In order to understand the source of edema following iontophoresis, 6 additional ra bbits received the same bi lateral wounds as described before. One eye received iontophoresis with only the 15% sucrose vehicle and the other received 1.0 mg of CTGF ASO. The corneal thickness was measured before and after excimer laser PTK and then agai n immediately following iontophoresis. In all cases, the amount of edema was calculated daily by the difference between the corneal thickness on that day and the immediate post wounding thickness. The percent difference in thickness was calculated using the formula in Figure 2 5 Therapeutic and Biochemical Efficacy of Iontophoresed CTGF ASO In order to determine if io ntophoresis of CTGF ASOs could both reduce the target protein and whether there was also a reduction in corneal haze, twelve rabbits were anesthetized and received bilateral excimer wounds to the center of their cornea as described above. Immediately following excim er wounding, 0.125 mg of either CTGF ASO or a scrambled ASO were delivered into the wound by iontophoresis (3.0 mA for 5 min) using the setup depicted in Figure 2 4 B During follow up, the rabbits were grossly

PAGE 44

44 observed with no contact or fluorescein installation due to previous observations of epithelial instability with frequent observation an d fluorescein instillation (2009 ). The rabbits were divided into two groups of 6 each for evaluation at days 7 and 14. At day 7 post wounding, one rabbit was removed from the study due to an unrelated injury, the remaining 5 rabbits had their eyes photographed for haze quantification, wer e euthanized, and their corneas were collected for CTGF quantification. Briefly, the cornea with a scleral rim was harvested and placed in a cornea punch block. An 8 mm punch was used to harvest the wound and a 2.0 mm rim of uninjured cornea. The excise d tissue was placed in a tube, snap frozen in liquid nitrogen, and stored at 80C until processed. The remaining 6 rabbits were similarly processed at day 14, thus creating two time points with coordinated haze measurements and CTGF quantification. Haze Quantification via Macrophotography At present, a subjective scale of haze density is the accepted method of measuring corneal haze. A value between 0 and 4 is chosen based on qualitative factors such as the amount of detail visible in the iris through the scar. In order to improve the objectivity and to provide a better means of communicating the effect(s) of a given anti fibrotic therapy, a novel macrophotography based method for haze imaging and quantification was developed. Macrophotography In order to both quantify the amount of light reflecting haze, and to generate a lasting image for complete visualization of the corneal scar, a macrophotography method was developed. Prior to general anesthesia, each eye was topically anesthetized with p roparacaine and each pupil was dilated with phenylephrine 2.5% and tropicamide eye drops. Each rabbit was then generally anesthetized with inhaled

PAGE 45

45 isoflurane as described earlier. The eyelids were held open and out of the way with either an eyelid specul um or a pair of cotton swabs. Either a Nikon D40 DSLR, or in later experiments a Nikon D7000, was outfitted with a macro lens capable of native 1:1 reproduction (either a 100mm Tokina or 60mm Nikkor) and the Nikon R1C1 Creative Lighting System (CLS) flash 200, manual exposure with a shutter speed of 1/500 second and f/16. T he D 7000 was second and f/18. To visualize a nd measure haze, the flash power was set manually (1/16 th D40, 1/6.4 th D7000) and neither the flash nor lens had a filter. For all images, the lens was set to manual focus and pre focused to a 1:1 reproduction ratio and the camera was focused by moving t he camera closer or further from the subject. Guide lights on the flash heads were used to facilitate haze visualization and focusing. H aze evaluation and q uantification As with any measurement system, there is noise present in the macrophotographic techn ique used herein to quantify the amount of light a corneal wound is reflecting. A set of images from normal corneas and from wounded or pseudo wounded corneas were used to identify the signature of the red reflex of the retina (black level) and the surfac e report of the macro flash (false positive). First, in order to blacken the retinal red reflex, all of the images were subjected to anti red grayscale conversion by using only the unmodified data in the blue channel. The overall contrast between the wou nd and retina was increased by overexposing the image to the point where the sclera became saturated, but the wound was not. The same global exposure correction was applied to every image. A circular region in the center of the cornea which encompassed t he wound and/or the flash reports was

PAGE 46

46 of the selected area. Using the recorded histograms, the signal due to the retina reflex and the flash reports were identifiable by co mparing the histograms from normal corneas versus the wo unded or pseudo wounded corneas, an example is presented in Figure 2 6 From these comparisons, a band pass filter was generated which excluded pixels with values lower than 80 and greater than 168 to preclude pixels outside of the wound (i.e. reflex, <80) and pixels from the flash reflection (>168) respectively. The pixels within the band were integrated and this band pass filtered integrated density served as a quantitative measurement of the inte nsity of corneal haze (haze score). CTGF Quantification V ia E nzyme L inked I mmunosorbant A ssay (ELISA) In order to link the observed effects with changes in our proposed molecular target, the total mass of CTGF was measured by a quantitative sandwich ELISA The total extracted CTGF mass per biopsy was chosen as the basis of compar ison Since CTGF has been observed to induce both significant cellular proliferation and protein synthesis normalization to the total mass of protein extracted is expected to skew the results. First, the tissues were removed from deep freeze storage and thawed on ice. extraction buffer (PBS pH 7.4 + 0.1% Triton X100, 5mM EDTA, 2mM PMSF, 0.24 mg/ml lev amisole) and ground with a dounce homogenizer. Finally, the base of the tube was submerged in iced saline and the homogenate and remaining tissue were subjected to ultrasonication. The homogenate was cleared by centrifugation and the supernatant was tran sferred to a fresh tube.

PAGE 47

47 Finally, total extractable CTGF was measured via a sandwich ELISA in a 96 well Biological (C7978 ernight at 4C. Diluted SuperBlock (1/10th) was used as the reagent dilutent. Three replicate wells per s plated in duplicate using recombinant human CTGF (rhCTGF) at concentrations of 400, 200, 100, 50, 25, 12.5, 6.25, and 0.0 ng/ml. The plate was sealed and incubated for 1 hour at 30C. The red saline with 0.05% (v/v) Tween 20 (TBST). Fifty microliters of 200 ng/ml biotinylated probe antibody from US Biological (C7978 25D) was applied to each well then sealed and incubated at 30C for 1 hour. The plate was then washed as before. The plate was finally incubated with streptavidin alkaline phosphatase (Zymed, 1:10,000) at 30C for 1 hour. The plate was ELISA was developed with PNPP in basic ethanolamine. The pla te was continuously monitored, with the absorbance at 450 nm being measured every 10 minutes for an hour. Results Iontophoretic Delivery of Reporter ASOs In the wounded ex vivo corneas, topical delivery of the labeled ASO ( no current for 5 min) resulted in no detectable delivery (Figure 2 7 A). A delivery current of 1.0 mA for 5 min resulted in trace amounts of green fluorescence (Figure 2 7 B). For current values of 2.0 and 3.0 mA, the epithelium at the wound margin and the anterior s troma possessed high levels of green fluorescence homogenously distributed in an

PAGE 48

48 electrophoretic like band (Figure 2 7 C F). The penetration depth and intensity of green fluorescence were both highest in the 3.0 mA treated cornea. In Vivo Wounding and Iont ophoretic Treatment All of the eyes that received iontophoresis tolerated the procedure well without any observable signs of irritation or damage at the time of surgery. In the days following surgery, some rabbits which received iontophoresis, and some wh ich did not, showed mild signs of inflammation and irritation at levels consistent with corneal wounding. Most wounds were re epithelialized by day 3, the remainder re epithelized by day 4 with no clinically significant difference amongst the CTGF ASO or scrambled ASO treated, and the eyes receiving no further intervention (Figure 2 8 ). Eyes that received the CTGF ASO were significantly thicker immediately after iontophoresis than those receiving the sucrose vehicle (Figure 2 9 A, p = 0.01). The corneas t hat received the CTGF ASO had less swelling than the opposing eye that received the scrambled ASO. A paired analysis of the differences in edema revealed a statistical trend for days 1 and 4 (Figure 2 9 B, p = 0.13 & 0.10 respectively ) and significant dif ferences for days 2 and 3 (Figure 2 9 B, p = 0.04 & 0.03 respectively ). One rabbit from the day 7 set was removed from the study due to a pre existing wound discovered on the contralateral eye immediately following surgery and treatment of the first eye. Another sample from the day 7 set was lost precluding paired CTGF analysis of that pair. In each pair of corneas, the total extractable CTGF protein contained in the analyzable biopsies was reduced in the treated corneas compared to the mock treated corne as by varying amounts (Figure 2 10 ). The reduction of CTGF protein was statistically significant for both days 7 (p = 0.05) and 14 (p = 0.002) with an average reduction of 15.7% and 30.7% respectively.

PAGE 49

49 The images in Figure 2 1 1 represent paired eyes from two rabbits in two separate experiments. These are examples of the images that are used to quantify the intensity of the scar as measured by the return of light by the scar ; Figure 2 1 1 A&B were imaged at day 7 while Figure 2 1 1 C&D were imaged at day 2 8 T he haze is most prominent at ich are most easily seen in the CTGF ASO treated cornea at day 28 (Figure 2 1 1 C). The total integrated density was not always reduced in each CTGF ASO treated eye compared to the contralateral scrambled control (Figure 2 1 2 ). However, in aggregate, the average paired difference indicated a decrease in total integrated density with anti CTGF treatment The decreased in haze was not statistically signific ant at day 7 post wounding (p = 0.23) but was significantly reduced at day 14 (p = 0.04). Discussion In the work reported here we have demonstrated that iontophoresis is a viable m eans of delivering oligonucleotides into the cornea. We have also demonstrated that iontophoretically delivered, biologically active, ASOs can both reduce the targeted gene product and have therapeutic effects consistent with that gene's known functions. These findings support the use of iontophoresis both as a therapeutic modality, and as a research tool capable of temporally and spatially controlling a gene product. The observation of a statistically significant reduction in corneal thickness during re epithelialization, which is strictly due to reduction in edema, was not expected and potentially reveals a novel role for CTGF in the cornea. In the lung, TGF been found to have a role in the modulation of fluid transport where it has been demonstrated to decrease the fluid transport across alveolar epithelia 106 As CTGF is a

PAGE 50

50 known mediator of TGF fibrotic acti vities, it is possible that CTGF may also mediate this novel function of TGF majority of fluid transport out of the cornea and the epithelium has a minor role. It is conceivable that the presence o f CTGF during re epithelialization decreases the fluid transporting capacity of the migrating epithelial cells. Therefore, a reduction of CTGF may enable these migrating epithelial cells to maintain their capacity to move fluid out of the cornea. While o ur data here do not rule out other possible explanations, including the possibility that the presence of CTGF could have effects on the barrier function of the migrating epithelial cells, the lack of a significant difference in the re epithelialization rat e between these same groups is evidence against a mechanism based on the re establishment of the epithelial barrier. Overall, the amount of edema was elevated in the iontophoretically treated corneas versus wounded corneas receiving no further intervention, though only the increase in the scrambled ASO treated corneas was significant. It is possible that the iontophoretic treatment, or the sucrose vehicle, could stimulate edema above what is normally seen in wounded corneas. The lack of signif icant difference in the anti CTGF treated cornea versus the non intervention control may be explained by our hypothesized improved fluid transport compensating for the treatment induced edema leading to an overall lack of significant increase. While the i ontophoresis may have increased the post operative edema, it appears to be short lived and does not appear to be clinically significant. The i ontophoretically delivered CTGF ASO consistently had the desired biological effect of reducing total CTGF protein present in the healing corneas. Overall, the

PAGE 51

51 amount of CTGF present in the cornea was higher at day 7 than at day 14. There are currently two conflicting reports about the expression levels of CTGF during cor neal wound healing. Blalock et al. report ed a fairly consistent rise in CTGF levels up through day 21 post wounding in rat corneas 63 while Yang et al. report ed a peak level of CTGF expression at day 3 post wounding in rabbit corneas 107 which then declines to normal levels by day 21. Our data support the expression profile more in keeping wi th the observations reported by Yang and others In this experiment, it is likely that the concentration of intracellular ASO was insufficient at day 7 to handle the increased concentration of CTGF mRNA and generate an equivalent magnitude of effect of t hat seen in the day 14 corneas. Even with the moderate reduction of CTGF, an observable and statistically significant difference in light scattering haze was measured. Additional in vivo work with increased doses must be performed to determine whether gr eater levels of protein reduction are possible and whether greater reductions in corneal haze will follow. In these series of experiments, only a few rabbits had a robust scarring response. Similar to humans, the robustness of scarring is not homogenous throughout the population of experimental rabbits. While the quantitative technique employed was sensitive enough to measure differences in the moderate to mild scarring rabbits, the effect was dramatically more profound in those with a robust fibrotic r esponse. Concurrently with the work reported here, Excaliard, Inc., a c ompany which is testing an CTGF ASO as anti fibrotic drug in skin, has had success in improving the appearance of hypertrophic skin scars in a phase II clinical trial 97 While these results were obtained in a different tissue and via a different delivery method (injection), they do further

PAGE 52

52 reinforce our f indings of the ability for CTGF ASOs to control hypertroph ic scarring. In addition to the positive therapeutic effects, the effects observed in the trial followed a pattern similar to what we observed. Subjects with mild scarring had less of a reduction in the appearance of the scar compared to subjects who had robust scarring; suggesting that the reduction of fibrosis may reach some basal level not addressable by anti CTGF therapy. Experimentally, these observations suggest the presence of a significant heterogeneity in scar formation in experimental populatio ns which needs to be considered during the design of experiments or clinical trials. The observation that not all treatments reported here resulted in a decrease in haze might be explained by limitations of the measurement system. At present, not enough s amples have been generated to determine the variance in measurements made using this photogrammetric technique. Additionally, the current haze measurement only takes into account the reflectivity of the wounded cornea, which may not be the sole measuremen t to consider. Using the eyes depicted in Figure 2 1 1 A& B as an example, the mock treated eye appears worse due to the more rapid progression of the spread of the haze from the wound margin compared to the anti CTGF treated right eye. However, the treatment (Figure 2 1 1 A) had an overall higher amount of integra ted density than the scrambled, ASO treated eye (Figure 2 1 1 B). The rate of haze generation and rate of spread, not just total reflectivity, may need to be measured at these early time points during formation of corneal haze in order to measure the effica cy of any anti fibrotic therapy. The results reported here add to the evidentiary base of that CTGF is necessary for the formation of cornea opacifying cellular scars. Additionally, the novel observation

PAGE 53

53 of a difference in corneal edema due to treatment w ith CTGF ASOs implicates a new role for CTGF in the regulation of fluid transport. Here we presented further evidence which demonstrates that iontophoresis has the capacity to deliver biologically active oligonucleotide therapies into the cornea. In addit ion, for the first time, we provide evidence of combined biochemical and therapeutic effects by reducing the amount of CTGF as well as the amount of clinically relevant, and directly observable, corneal haze following corneal wounding. The results present ed here indicate that iontophoresis can be used to surmount the significant barriers currently impeding the translation of oligonucleotide therapies into clinical practice.

PAGE 54

54 Table 2 1. Antisense oligonucleotides used in the anti fibrotic experiments. ASO Target Molecular Mass Sequence ISIS 13920 Reporter /hRAS 7233.21 Da TCCGTCATCGCTCCTCAGGG ISIS 124189 CTGF 7235.29 Da GCCAGAAAGCTCAAACTTGA ISIS 29848 Scrambled 7233.20 Da Random A B C D E F Figure 2 1 Results from iontophoresis into intact rabbit corneas. A macrophotograph of the eye treated with a directly labeled reporter oligonucleotide A) white light illumination and B) blue excited, yellow filtered fluorescent macrophotography of the same eye. D irect fluorescence micrographs C ) immediately after delivery and D ) 24 hours later. Micrographs of a nuclease resistant oligonucleotide revealed by immunohistochemistry E ) immediately after delivery and F ) 24 hours later. Approximately 40% of all cells i n panel F ) are ASO positive by direct counting (216 Red / 547 Blue = 39.4 % )

PAGE 55

55 Figure 2 2 A representative excimer PTK wound in an ex vivo cornea. Figure 2 3. The iontophoresis setup used for the ex vivo experiments.

PAGE 56

56 A B Figure 2 4 Two schematics of the iontophoretic setups used. In the cut away view, the yellow represents the slightly acidic HEPES, the blue represents the ASO/sucrose solution. In A) the anode is surrounded by a buffer soaked sponge (blue) and gauze (grey). In B) a sche matic of a rabbit ear with slightly basic buffer soaked gauze (blue) wrapped around it and held in place with the anode. The anode would itself then be wrapped with another buffer soaked gauze. Figure 2 5 i 0 wounding, thickness.

PAGE 57

57 Figure 2 6 An example of a band pass filter generated with non viable eyes A different filter was designed for use on images of viable eyes.

PAGE 58

58 A B C D E F Figur e 2 7 Reporter ssDNA delivery into ex vivo globes A) 0.0 mA (topical), B) 1.0 mA C,D), 2.0 mA or E,F) 3.0 mA.

PAGE 59

59 Figure 2 8 The area of fluorescein staining was measured daily following PTK wounding with or without iontophoresis. ANOVA indicates that the differences in wound area during healing amongst the groups were not significantly different. 0 5 10 15 20 25 30 35 40 45 50 0 1 2 3 4 5 Fluorescein Stained Area (mm 2 ) Day Post Wounding Wound Area vs. Time Average Iontophoresis Scrambled Average Iontophoresis anti-CTGF Average No Intervention n = 3 Bars = 1 Standard Error ns ns ns

PAGE 60

60 A B Figure 2 9 Corneal edema following ASO treatment and during healing. A) The d ifference in corneal edema immediately following iontophoresis of eith er the sucrose vehicle or the vehicle and CTGF ASO. B) Edema in eyes receiving CTGF ASO or scrambled ASO in the contralateral eye. A statistical trend in edema between the CTGF and scrambled ASO treated corneas was present for days 1 & 4 post wounding (p = 0.13 & 0.10) and statistically significant differences were present for days 2 & 3 (p = 0.04 & 0.03). 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Central Corneal Edema (% Difference in Thickeness) Post Iontophoresis Edema SUCROSE (n = 6) CTGF ASO (n = 5) p = 0.01 0% 50% 100% 150% 200% 250% 0 1 2 3 4 % Thickness Time (Days) Post Wounding Thickness vs. Time Average Iontphoresis Scrambled Average Iontophoresis Anti-CTGF p = 0.13 p = 0.04* p = 0.10 p = 0.03*

PAGE 61

61 Figure 2 10 The paired quantity of total CTGF mass extracted from each 8.0 mm corneal punch. The samples were plated in duplicate, the error bars re present one standard error. A one test analysis revealed that the observed paired differences were statistically significant (Day 7, p = 0.05; Day 14 p = 0.002). A B C D Figure 2 1 1 Pai red examples (A&B, C&D) of CTGF versus scrambled ASO treated corneas. A) Day 7 CTGF ASO treated and B) the opposing eye treated with scrambled ASO. C) Day 28 CTGF ASO treated and D) the opposing eye treated with scrambled ASO from the tolerance experiments. 0 50 100 150 200 250 300 10H16 10H17 10H18 10H19 10H22 10H23 10H24 10H25 10H26 10H27 Total CTGF Extracted (ng / biopsy) Paired Samples' Total Wound CTGF Content Scrambled ASO Anti-CTGF ASO Day 7 Day 14 p = 0.05 p = 0.002

PAGE 62

62 Figure 2 1 2 Individual pa ired haze measurements. A one test analysis revealed that the observed paired differences were not statistically significant on day 7 (p = 0.23), but did reach statistical significance on day 14 (p = 0.04). 0 10 20 30 40 50 60 10H15 10H16 10H17 10H18 10H19 10H22 10H23 10H24 10H25 10H26 10H27 Integrated Density (pixel x pixel value x 10 6 ) Paired Differences in Haze Integrated Density Scrambled ASO Anti-CTGF ASO Day 7 Day 14 p = 0.23 p = 0.04

PAGE 63

63 CHAPTER 3 TH E LOCATION OF CTGF P RODUCTION AND ACCUMULATION IN HEALING CORNEAS Introduction In the preceding chapter, I sought to decrease the presence of CTGF in the wound healing cornea since it had been identified as a key downstream mediator of transforming growth factor 59 The results were positive, but both the reduction of CTGF and haze were clinically marginal. The delivery of the ASOs to the region of interest is reproducible (Figure 2 7 ), which gives rise to the question of whether the right region of the cornea is being treated Given the mechanism of action of the ASOs is the key piece of information needed to ensure that the ASOs are being dosed to the correc t cell layer(s). Additionally, the final locus of activity is necessary both for therapies targeting CTGF protein and/or receptor neutralization and to further understand the biological role of CTGF in fibrosis. Early investigations have focused on fibr oblasts due to fact that they do produce CTGF in response to TGF 59 The first investigations into CTGF localization in healing corneas found that CTGF was present in all cell layers of the cornea, but had hi gher concentrations in the epithelium 63 The levels of CTGF were found to continually rise from the time of wounding up through 28 days post wounding 63 Subsequent rep orts have observed a peak in CTGF protein at day 2 or 3 post wounding which then returned to basal levels by day 21 post wounding 107 Given the ELISA results presented in the previous chapter (Figure 2 10 ) a n expression profile with a peak and subsequent reductio n during the wound healing proces s is currently better supported. Herein, I will describe the distribution of CTGF in the time period prior to, and during the early phase of, haze formation in rabbit corneas as measured by immunofluorescent staining In

PAGE 64

64 order to measure the location of the synthesis of CTGF protein, a CTGF promoter driven eGFP reporter mouse was used to identify the cells with CTGF promoter activity in normal corneas and during wound healing. The promoter activity level found in normal c orneas was validated by measuring the amount of CTGF transcripts in grossly dissected normal rabbit corneas. These experiments are expected to reveal where CTGF is synthesized and where the protein is localized within the healing corneas. Materials And Methods Excimer Laser Surgery Each rabbit was anesthetized using inhaled isoflurane and each eye was topically anesthetized using one drop of tetracaine or proparacaine. The eye was exposed using either an eyelid speculum or was proptosed with a pair of cotton swabs. The central thickness of the cornea was measured with a clinical ultrasonic pachymeter (n = 5 central measurements per eye). Each eye then received a central 6.0 mm diameter, measured again by the ultrasonic pachymeter to verify ablation depth. Each mouse was placed in a sealed box and anesthetized with 3.5% inhaled isoflurane/oxygen. Once anesthetized, it was removed from the box and placed straddling a 50 ml conical tube and held in place with a semi adhesive elastic band gently wrapped around both the mouse and tube. The tube was outfitted with a nose cone fed supply of isoflurane/oxygen. The mouse was laid on its side and supported by a stack of cotton gauze. The eye to be wounded received a drop of proparacain. The whiskers were cut with scissors and the eye lid and lashes were gently pushed out of the way with a cotton swab and held in place with an adhesive strip if needed. The

PAGE 65

65 mouse was then oriented beneath the exc deep PTK excimer wound was created. Tissue Harvesting, Processing, and Sectioning At the terminal time point, the rabbits and mice were anesthetized as before and euthanized by intravenous Beuthanasia D and cervical dislocation, respectively. For rabbits, e ach globe was immediately enucleated using forceps and sharp scissors to cut the conjunctiva, muscles, and optic nerve. For mice, each eye was proctosed and the posterior portion of the globe was held firmly with fine pointed forceps and the eye was forcibly removed. Both the rabbit and mouse eyes were immediately placed in fresh 10% neutral buffered formalin on ice. The globes were punctured with a 25 gauge needle after 1 hour to improve fixative penetration. T he globes were fixed overnight at 4C. The globes were then grossly prepared by cutting open the post erior retina, removing the lens and any residual air bubbles. The globe and lens were placed in 30% sucrose in PBS overnight at 4C to cryoprotect the ti ssues. In order to improve cryosectioning, each globe was injected with embedding medium (OCT) with a blunt tipped needle and syringe. The lens and globe tissues were then submerged in OCT and rapidly frozen. The frozen blocks were stored at 20C until cryosectioning. Ten micron section s were cut with a cryotome and mounted on poly L lysine coated slides. The slides were air dried and then stored at 20C until staining. Immunohistochemical Staining Rabbit corneas In order to visual the localization o f CTGF protein in healing rabbit corneas immunohistochemical staining with an anti CTGF monoclonal antibody (created in a mouse). Briefly, the slides were washed 3 times with wash solution (10 mH HEPES,

PAGE 66

66 150 mM NaCl, 0.02% Na Azide, and 0.05% Tween 20, pH 7.4). The slides were blocked for 30 min with 10% normal horse serum (Vector Labs). The sections were then washed twice for 2 minutes with wash solution. Half of the sections then received either antibody dilution solution (HEPES Buffer) or the anti dilution solution) and were then incubated overnight at 4C. The antibody solution was then carefully removed and the samples were then washed 3 times for 5 minutes each. The sections were then incubated with affinity puri fied fluoresc e in labeled anti mouse IgG (2 0 minutes each. The slides were mounted with 4',6 diamidino 2 phenylindole (DAPI) containing medium and the coverslips sealed with nail h ardener. The slides were kept refrigerated in the dark until imaged. Reporter mouse corneas In order to determine both the cells producing CTGF and where CTGF is binding, I chose to use immunohistochemical staining with a biotinylated anti CTGF antibody o n the wounded reporter mouse corneas Briefly, the slides were washed 3 times with wash solution (10 mH HEPES, 150 mM NaCl, 0.02% Na Azide, and 0.05% Tween 20, pH 7.4). The slides were blocked for 30 min with 10% normal rabbit serum (Vector Labs). The s ections were then treated to block endogenous biotin and avidin according twice for 2 minutes with wash solution. Half of the sections then received either antibody dilution solution (HEPES Buffer) or the anti solution) and were then incubated for 30 min at room temperature. The antibody solution was then carefully removed and the samples were then washed 3 times for 5 minutes each. The se ctions were then incubated with avidin

PAGE 67

67 Vector Labs) for 10 min. The samples were then washed 3 times for 5 minutes each. The slides were mounted with DAPI containing medium and the coverslips sealed with nail hardener. The slides we re kept refrigerated in the dark until imaged. Transgenic Mice These mice possess a transgene composed of the promoter from connective tissue growth factor upstream from an enhanced green fluorescent protein (eGFP) 108 109 (Figure 3 1A) Th is construct results in the accumulation of eGFP in cells that possess activites capable of stimulating transcription from the CTGF promoter. With the additional step of imm unostaining for CTGF protein in these mice the location of production of CTGF (green) and final destination of the protein (red) provides a more detailed picture of the source and locus/loci of activity of CTGF Mice with the transgene were identified us ing a cobalt or cyan light source to illuminate the eye, the lenses of transgene positive mice fluoresce intensely. Using a camera with a dedicated macro flash with a cobalt blue filter on the flash and a deep yellow filter on the lens, the e GFP in the le ns was easily observable in the mice with the pCTGF GFP transgene (Figure 3 1 B ). The mouse pups were chosen by phenotypic observation of the green fluorescence in the lens in place of standard genotyping. Whole Mount Confocal Micrography Enhanced green fluorescent protein positive mice were euthanized and their eyes were immediately enucleated and placed in 10% neutral buffered formalin Following overnight fixation at 4C, the eyes were grossly dissected and the tissues were cut into sma ll pieces and placed into a custom made slide with a reservoir. The tissues were immersed in mounting medium containing DAPI counterstain (Vector Labs) and a

PAGE 68

68 coverslip was affixed in place with nail hardener. The pieces of tissue were then imaged with a spinning disc confocal microscope. Micrography The sections were imaged on a confoca l microscope with either a 10 x objective or a 60 x oil immersion objective. For the whole mount tissues, the entire volume of tissue within the field was imaged. The z s stack representation was exported for each image. For the immunofluorescently stained sections, the brightest field from the brightest section receiving primary antibody was used to determine the exposure to be used for the entire series of rabbit cornea sections. The exposure conditions were chosen by If the image was overexposed with these condit ions, the exposure time was reduced just to the point that the image was no longer overexposed (i.e. that the final overexposed pixels were no longer overexposed). If the initial exposure conditions did not result in an overexposed image, the gain was rai sed to the point just before overexposure. The gain was never raised above 50% of the maximum gain value, if additional exposure was needed, the exposure time or illumination intensity was adjusted. The signal intensity in the control slides was used as a background correction for the entire series of rabbit cornea sections The same overall procedure was used for imaging the mouse cornea sections, but given the different staining conditions, a different set of exposure conditions were determined and use d.

PAGE 69

69 Gross Corneal Dissection and CTGF Transcript Quantification Rabbits without observable corneal wounds were anesthetized and euthanized as described before. A scalpel was used to immediately scrape the epithelium off with care taken to ensure that the scraped mass was r etained on the blade. The scra ped was the n excised from the globe by cutting with a fresh scalpel and scissors at the corneal/scleral boundary. The cornea was plac ed face down and yet another fresh scalpel was used to scrape off and retain the endothelium as was done with the epithelium. The endothelial ossly isolated cellular layer was then subjected to ultrasonication on ice for further tissue disruption. The probe was rigorously washed, rinsed and dried in between each sample. The homogenates were then immediately loaded onto Qiagen gDNA removal colu mns and RNAeasy Qiagen, Inc., Cat. #74104 ). The purified RNA was quantified via ultra violet absorbance using a Nanodrop ND 1000 spectrophotometer set for RNA quantificat ion. Equal masses of RNA were loaded for each sample and 2 well PCR reaction plate. Real Time PCR (RT PCR) was performed on an Applied Biosystems 7900HT Fast Real Time PCR System utilizing the man recommended thermal cycling conditions. The relative gene expression of CTGF versus GAPDH was calculated for each tissue using the primers and probes listed in Table 3 1

PAGE 70

70 This relative metric was used to compare the relative enrichment of CTGF mRNA in each tissue. A few samples were run without reverse transcriptase to measure the quantity of genomic DNA (gDNA) present in the sample Results Excimer Wounding The excimer wounding was well tolerated by both the rabbits and mice with no complications. An example mouse corneal wound is depicted in Figure 3 1 C The wound s were consistently circular and the rough appearance of the ablated surface evidences that the wound penetrated into the stroma. The Location of CTGF Protein Accumulation Within 30 minutes after wounding, a green fluorescent signal was present in the migrating in the epithelial front (Figure 3 2 A,C,E ) but not in the 1 antibody withheld control (Figure 3 2B&D) One day after wounding, there is an increased presence of g reen fluorescence in the healing epithelium, stroma and endothelium (Figure 3 3A,C,E) The green fluorescence in the stroma is also present in the 1 antibody withheld control indicating that it is either non specific binding of the 2 antibody or autofl uorescence (Figure 3 3 B ). Given the hue and homogeneity of the cytoplasmic staining, the stromal signal in these fields is most likely autofluorescence. Day two after wounding the intensity of the green fluorescence peaks, and is found in both the epithe lium and stroma (Figure 3 4). This time, however, the majority of the green more indicative of a positive signal. Again the corneal endothelium also has a positive green fluorescent signal. In the remaining days, the total intensity of staining diminishes

PAGE 71

71 and the major locus of binding is primarily confined to the basal epithelium (Figure 3 5 through Figure 3 9 ) The Location of CTGF Promoter Activity For t his model, the presence of green fluorescence within the cytoplasm of cells is evidence of CTGF promoter activity. In the unwounded cornea, the endothelium is the only cell layer in the cornea with a detectable signal (Figure 3 10). Fluorescent confoca l micrographs of the cryosectioned eyes revealed that within the cornea the endothelium was still the site of highest CTGF promoter activity as was seen in the unwounded corneas However, in the wounded cornea s ome sub epithelial cells in the stroma poss essed detectable green fluorescence (Figure 3 1 1 A red arrows ) But the presence of red fluorescence in these cells in the immunofluorescent control ( Figure 3 1 2 A ) suggests that this signal is a result of autofluorescence. Furthermore, t he location and appearance of the fluorescence is reminiscent of the autofluorescence seen earlier in the rabbit samples (Figure 3 3B) The high red fluorescent background prevented interpretation of the presence or absence of CTGF protein in the mouse endothelium in thes e samples. Despite the high background signal in the control sections a key difference was noted between the sections receiving the primary antibody and those that did not. As was seen in the rabbits, t he basal epithelium of the cornea was the primary l ocation of CTGF antigenicity (Figure 3 1 2 C ) The CTGF protein is predominantly located on the basal portion of the basal epithelium and appears to have an extrac ellular, punctate, distribution as was seen in the rabbit samples.

PAGE 72

72 Gross Corneal Dissection and CTGF Transcript Quantification In the unwounded rabbit eyes, the grossly dissected epithelium was the tissue with the greatest RNA mass recovered (Figure 3 1 3 A ). All three layers possessed CTGF transcript with a cycle threshold value indicative of a good level of confidence. Of the three layers, the endothelium had more CTGF transcripts per GAPDH transcript (Figure 3 1 3 B) mirroring the data that the pCTGF eGFP reporter mouse provided. Discussion Given that our primary theory behind the formation of haze centers on the activity of connective tissue growth factor, and that CTGF is primarily located in the basal epithelium, new questions arise about which cells we should be targeting for therapies. The observation that in the cornea, CTGF promoter acti vity and mRNA quantity are both highest in the endothelium is novel. During wound healing, any change in promoter activity in the epithelium or stroma does not appear to reach a level sufficient for detection under the conditions used. However, additiona l work is currently planned to repeat the gross dissection qRT rtPCR quantification of CTGF mRNA d uring the wound healing process to observe where in the cornea the previously observed cornea wide upregulation occurs. These data is key to refining future delivery methods for anti CTGF mRNA therapies that are in continued development. G iven that the basal epithelium is both the location of high CTGF protein accumulation and it is where the scar will eventually be, significant conflicts have now been introdu ced to t he initially proposed cellular and molecular model f or scar formation (Figure 1 5) Connective tissue growth factor is present in abundance in the unwounded lens (Figure 3 1 B ) and is constitutively expressed in the corneal endothelium in unwounded corneas. These facts indicat e that CTGF presence is not

PAGE 73

73 sufficient to generate a scar, it also suggests that CTGF may have a basal function that is present in th ese tissues that is also necessary for the generation of a scar. These observations raise the specter of potential off target effects given that they are present in unwounded tissues. The function of CTGF in these tissues is currently unknown thereby precluding a prediction of potential side effects. The presence of CTGF in the cornea increase s during wound healing. While increased transcription and translation is the most commonly attributed mechanism for increases in protein levels in a tissue, other mechanisms are possible. One such mechanism is increased accumulation due to a change in a protein in question. Given the evidence that CTGF begins to accumulate in the basal epithelium as soon as 30 minutes after wounding, it may very well be that on e of the first stages in the CTGF response pathway is that contact de inhibited epithelial cells become rapidly sensitized to CTGF thus enabling a rapid response to the basal levels that exist in the intact cornea. Further evidence supporting a differential cellular sensitization are the fact s that fewer suprabasal epit helial cells bind CTGF than the basal epithelium do not accumulate CTGF. During re epithelialization, the epithelium has yet to begin re stratification and the CTGF staining is present throughout the epithelium (up to day 2 post wounding). Once the epith elium begins to re stratify, staining for CTGF is predominantly located in the basal epithelium, not the differentiated cells that form the more apical surface of the epithelium (day 3 and beyond). Th is observed differential sensitization may be another better, therapeutic target given both that the epithelium is accessib le to topically applied agents and that sensitization would

PAGE 74

74 By far, the most interesting finding was that the primary source of CTGF synthesis and protein binding were in separate cell layers on the extremities of the cornea. The effects of strongly iontophoresed CTGF ASO (5.0 mA) on corneal edema (Figure 2 9B) and haze (Figure 2 11C) indicate that the observation warrants further inv estigation. At first consideration a mechanism of synthesis in the endothelium, diffusion through the stroma, and subsequent binding to the epithelium is difficult to accept. The difficulty arises due to the fact that the bulk flow of water is out of th e cornea across the endothelium and that the stroma is a significant barrier to ionic macromolecule diffusion. However, during wound healing the next flux of fluid is into the cornea as is evidenced by the edema during the first several days of healing. It has been reported that an edematous stroma is not as effective of a barrier as the normally dehydrated one 110 In diffusion chamber experiments with excised rabbit corneas under differing degrees of barrier to the diffusion of insulin under highly hydrated (i.e. edematous) conditions 110 (Figure 3 14A ) This reduction in stromal barrier function provides a reasonable mechanism to enable endothelial derived CTGF to diffuse to and bind to the corneal epithelium (Figure 3 14B) This proposed hypothesis is further supported by the observ ation that the peak of CTGF binding of the epithelium occurs during the edematous phase and tapers off in concert with the restoration of the normal corneal hydration state. While these observations and proposed mechanisms do not definitively prove that t he endothelium is the source of pro fibrotic CTGF during wound healing, they do form the basis of a new consistent and testable theory.

PAGE 75

75 Table 3 Growth Factor Species Accession Number CTGF Forward Reverse Probe AGGAGTGGGTGTGTGATGAG CCAAATGTGTCTTCCAGTCG ACCACACCGTGGTTGGCCCT GAPDH Forward Reverse Probe GAGACACGATGGTGAAGGTC ACAACATCCACTTTGCCAGA CCAATGCGGCCAAATCCGTT A B C Figure 3 1 The CTGF synthesis reporter mouse. A) A schematic representation of the reporter transgene. B) The lenses pCTGF e GFP mice have a green fluorescent phenotype enabling easy sorting of reporter mice from their liter mates. C) A representative PTK wound on a pCTGF e GFP mouse cornea.

PAGE 76

76 A B C D E Figure 3 2 Confocal microgra phs of anti CTGF (green) immuno fluorescent stained wounds 30 minutes post wounding. A) A wide field (10 x) image of the wound margin. C) & E) Higher power (60 x) micrographs of the migrating epithelial edge. B) & D ) are control s in which the primary antibody was withheld

PAGE 77

77 A B C D E F Figure 3 3 Confocal microgra phs of anti CTGF (green) immuno fluorescent stained wounds 1 day post wounding. A) & C ) represent portions of the epithelium while E) is the endothelium. B) D) & F ) are control s in which the primary antibody was withheld.

PAGE 78

78 A B C D E F Figure 3 4 Two days p ost Wounding: Wound Margin Most likely surface effects magnified by a not very flat tissue. A) & C ) represent portions of the epithelium while E) is the endothelium. B), D), & F ) are controls in which the primary antibody was withheld.

PAGE 79

79 A B C D Figure 3 5 Day 3 Post Wounding: A) & B) Wound body. C) & D) wound margin B) & D) are controls in which the primary antibody was withheld.

PAGE 80

80 A B C Figure 3 6 Day 4 Post Wounding. A) & B) Edge of hyperproliferation. C) a piece of undermined stroma in the body of the wound included as a mass surrounded by epithelial cells. B) is a control in which the primary antibody was withheld.

PAGE 81

81 A B C D Figure 3 7 Day 5 post wounding. A) & B) the wound margin C) & D) the wound body. B) & D) are controls in which the primary antibody was withheld.

PAGE 82

82 A B C D E F Figure 3 8 Day 7 post wounding. A) & B) 10x mag of the margin and surrounding area. C) & D) the wound margin E) & F) the wound body. B), D), & F) are controls in which the primary antibody was wi thheld.

PAGE 83

83 A B Figure 3 9 Day 10 post wounding at the wound margin. A) CTGF in the wound margin, and B) is a control in which the primary antibody was withheld.

PAGE 84

84 Figure 3 1 0 CTGF promoter activity in an unwounded mouse cornea. This z stack merged confocal micrograph is a whole tissue mount of a cut cornea. The epithelium and stroma are visible at the cut face, while the posterior surface of the cornea (endothelium) is vis ible as viewed en face

PAGE 85

85 A B Figure 3 1 1 The corneal endothelium is the primary location of CTGF promoter activity (green). A) The field that did not receive the primary antibody and B) primary antibody present A) There is some green fluorescence in the anterior stromal cells (red arrows).

PAGE 86

86 A B C Figure 3 1 2 The localization of CTGF protein in healing mouse corneas. A) Comparison of the red channel data representing the presence of CTGF (lower panel) or background noise (upper panel). B) T he merged layers of promoter activity and protein accumulation. C) Detail of the merged layers at the basal epithelium.

PAGE 87

87 A B Figure 3 1 3 Comparison of RNA mass yields and CTGF levels from grossly dissected rabbit corneas. Equivalent volume of lysi s buffer was used for each layer. A) RNA mass extracted and B) GAPDH normalized CTGF levels in each grossly dissected corneal layer. 0 100 200 300 400 500 Epithelium Stroma Endothelium Extracted RNA (ng/ l) Average RNA Yield 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Epithelium Stroma Endothelium Fold CTGF vs. GAPDH CTGF vs. GAPDH: Normal Eyes

PAGE 88

88 A B Figure 3 1 4 A new theory to explain the different loci of CTGF synthesis versus accumulation and action. A) A schematic of the observation that a normal (dehydrated) stroma is a barrier to insulin, but an edematous stroma permits the entry of insulin. B) A new theory which integrates the new findings about CTGF synthesis and localization, its timing with r espect to edema, and the observed permissiveness of the edematous stroma.

PAGE 89

89 CHAPTER 4 HAZE FORMATION TIMELINE Introduction To date, studies of light scattering haze have focused on the magnitude of haze that forms in mature scars, while there is curr ently little to no information published abo ut the establishment of haze. S tudying the final stable scar can enable one to ascertain what modulates t he amount of haze t hat forms, but it does not provide much insight about the process of haze formation. B ecause the formation of haze is the therapeutic target, this process deserves greater scrutiny. It is well established that cell migration, proliferation, and differentiation are key processes in the development of haze, but little is known of when and wh ere these processes occur. More detailed information about the when and where of haze formation will not only provide insight into the basic science of tissue repair, it will also provide a more refined target for highly targeted therapies. Materials and Methods Excimer Wounding Each rabbit was restrained in a PLAS Lab rabbit restraint and each eye was topically anesthetized using a single drop of tetracaine. Each rabbit was generally anesthetized using inhaled isoflurane and oxygen. Once sedated, the c entral thickness of the cornea was measured by an ultrasonic pachymeter (n = 5 central measurements per eye). The eye being operated on was proctosed using a pair of cotton swabs. The cornea was then centrally ablated using a Nidek EC 5 000 in narrow beam mode PTK zone.

PAGE 90

90 Follow Up T hree time points were chosen prior to epithelial closure, while the other three were during the time period of haze formation (5 to 10 days). Prior to euthanasia, the pupil was dilated with topical tropicamide and phenylephrine then the wound was photographed. Two rabbits were euthanized at each time point by injection of Beuthanasia D into the lateral ear vein while generally anest hetized with inhaled isoflurane and oxygen. Macrophotography Corneal haze was documented using a Nikon D40 digital single lens reflex ( DSLR ) camera was outfitted with a Nikkor AF D 60 mm macro lens and a Nikon R1C1 Creative Lighting System (CLS) flash syst second and f/18. For all images, the lens was set to manual focus and pre focused to a 1:1 reproduction ratio and the camera wa s focused by moving the camera closer or further from the subject. The built in guide lights were used to facilitate haze visualization and focusing. Pictures were taken until an image with adequate focus and framing of the wound was achieved and the ref lection of the flash heads were not within the wound. Two SB R200 remote flash heads were mounted on the front of the lens and were 1/32nd power for each flash. Each flash w as rotated to the most inward facing angle to provide the most oblique lighting possible.

PAGE 91

91 Evaluation of Haze Development The day to day change in haze was observed using macro photographs The color images were converted to grayscale using Adobe Photosho p CS3 Extended with a custom black and white conversion filter that used only the data in the blue channel of the image and that the flash report was The grayscale image was overexposed by a consistent amount in silico in order to better visualize the detail within the scar. The wound area was selected using a circular selection re gion with a dia meter of 850 pixels and a gradient co lor map was applied to the selected region (Table 4 1) T hese pseudocolored images were assembled in a time line sequence and the progression of haze development was followed. Molecular and Histological Analysis Foll owing euthanization, each cornea was briefly irrigated with 10% neutral buffered formalin (10% NBF) and then each globe was enucleated. In order to preserve the shape of the cornea, each globe was placed cornea down into a 12 well tissue culture dish well that was filled with 10% NBF for one hour. The cornea was then excised and placed in another well with fresh 10% NBF and was fixed for 4 to 18 hours at 4C. Frozen s ections Following fixation, the corneas chosen for cryosectioning were bisected and then cryoprotected in 30% sucrose in PBS overnight at 4C. The corne a halves were then oriented in OCT and quickly frozen on dry ice. The OCT blocks were stored at 20 C

PAGE 92

92 c L lysine coated glass slides and then air dried overnight. The sections were stored at 20C until staining. Paraffin s ections Corneas chosen for paraffin embedding and sectioning were bisected and t hen transferred into 70% ethanol and kept at 4C until submission for processing at the CTAC Histology Core) co L lysine coated glass slides. One slide per cornea was stained with hematoxylin & eosin (H&E) f or gross histological analysis. Immunohistochemistry Paraffin sections were de paraffinized with xylenes and rehydrated though a graded series of ethanol though tap water. Frozen sections were rehydrated with PBS. The rehydrated sections were then r insed with wash buffer (phosphate buffered saline with 0.05% Tween 20 PBST ) The sections were blocked with 10% normal horse serum (NHS) for 30 minutes at room temperature. The blocking solutions were carefully removed from one section group per slide by aspiration and blotting The primary SMA) antibody was d iluted (1:500) in 1% NHS i n PBST and was incubated for 1 hour at room temperature on the one section while the negative control remained in blocking solution. Following this incubation, the slides were washed three times for 5 minutes with wash buffer. The sections were then incubated with a fluoresceinated secondary horse anti mouse IgG for 30 min at room temperature Following the final washes, the slides were mounted with DAPI containing medium and the slide cover was fixed in place with nail hardene r.

PAGE 93

93 Results Wound Macrophotography The first sign of haze was typically present at day 5 post wounding. The first site of haze formation was invariably a ring shaped region of haze at the wound margin. At the same time, there was usually, but not always, a spot of haze in the body of the wound, but its shape and location varied wound to wound. Once initiated, the haze spread from the loci and intensified with time (Figures 4 1 A&B ) Immunohistochemistry SMA was co incident with the beginn ing of haze formation. The localization SMA within the tissue also mirrored the initiation and spread of haze (Figure 4 2A,C,&E) ; beginning with a distinct region at the wound margin (Figure 4 2B), and spreading laterally (Figure 4 2D &F ) with time. Also, t he intensification of haze was mirrored by an in crease SMA positive lamellae (Figure 4 2 D& F) with time. SMA was present in significant quantities in the basal epithelium (Figure 4 3), but its intracellular distrib ution was significantly different (diffuse) from that of the distribution in the stromal cells (sharp, fibrous). Discussion There appear to be two distinct regions of haze initiation (Figure 4 4) The formation of a ring of haze at the wound margin was in variant while both the formation of islands of haze in the body of the wound and the location of the se islands varied Once initiated, the haze appears to first spread laterally and secondly to become more intense with time. The distribution of smooth muscle actin, a putative marker of light reflecting cells, mirrored the spread and intensification of haze with an increase of cells along stromal lamellae increasing with time (spread) and the number of affected

PAGE 94

94 lamellae increasing as well (densi ty/intensity). These observations are the first to describe when and where haze first begins to form, and to describe how it covers the The formation of haze occurs only after re epithelialization and the spread of haze from the wound margin is a pattern similar to the migrating epithelial front. While t he SMA staining in the epithelium was not expected, it was fairly consistent and has been seen in cel SMA is distributed. Following TGF SMA will be arranged in fiber like structures in the fibroblasts while it remains gl obular and diffuse in the epithelial cells 111 The presence of CTGF on the s urface of basal epithelial cells (Figure 3 7A) and SMA present in side the basal epithelial cells (Figure 4 3) supports a new hypothesis that the epithelium may have a more central role in the formation of haze following acute injury to the cornea. The observations reported here all which gave rise to the final project presented in the next chapter.

PAGE 95

95 Table 4 1. Color map used in Photoshop to generate the pseudocolored images. Color % Full Scale Black 2% Blue 5% Green 8% Yellow 11% Orange 14% Red 17% Magenta 20% White 22% A B Figure 4 1 Pseudocolored images of day to day haze formation.

PAGE 96

96 A B C D E F Figure 4 SMA localization. A,C,E) are macrophotographs of the healing SMA localization. A&B) Day 5 post, C&D) day 7 post, and E&F) day 10 post.

PAGE 97

97 A B E F Figure 4 3. Higher power micrograph of the day 5 wound. A) original micrograph a nd B) the primary antibody withheld control. E&F) are zoomed in regions of A) to better show the detail of the staining in the epithelium and stroma.

PAGE 98

98 Figure 4 4 A schematic of where haze begins, how it spreads, and how it becomes more dense. The colors used to represent the coaxial spread of haze are related to the intensification as represented in the pseudocolored images in Figure 4 1 while the layering is related to the build SMA positive cell layers seen in Figure 4 2F

PAGE 99

99 CHAPTER 5 THE EMERGENCE OF A N EW THEORY FOR HAZE F IBROGENESIS Introduction To date, it is believed that the fibro blasts underlying the wound are the source of light reflecting cells that comprise corneal haze based scars The observations presented in the preceding chapter s cast doubt on this theory of haze fibrogenesis The observed timing and localization of CTGF synthesis and accumulation severely undercut the hypothesis that the underlying fibroblasts were the site o f CTGF activity. T he emergence of haze at two distinct regions of the cornea following acute injury is difficult to reconcile with the idea that the underlying fibroblasts migrate by spiraling up How can this theory explain the gap between the haze enc ircling the wound at the In addition to the new data reported here, other information adds to cast further doubt on the current theory. First, i t is generally known in the field o f ophthalmology that the cells in the epithelium are the first cells to come into contact with the surface of the unwounded s troma as they migrate to cover the wound within the first few of days following acute injury By day 5 following wounding in rabbi t corneas, the wound volume is filled with dark eosin staining epithelial cells. However, over the next 9 days, the wound volume becomes filled with a light eosin stai ning, collagenous matrix which is filled with fibroblastic cells This first became app arent when examining gross histological sections of mature scars where the shape of the scar tissue smooth muscle actin were both precisely the shape of the excimer laser wound (F igure 5 1A&B ). This same pattern was repeated in the worst wound scar observed during the experiments reported herein ( Figure 5 1C&D ) While these observations were

PAGE 100

100 independently discovered a search of the literature found that the next, more conclusive, experiment had already been done 33 Tuft and others covalently labeled the fluorescently labeled f ibrotic mass] first appeared in the periphery of wounds after reepithelialization, and had usually extended over the entire wound bed between the 7th and 14th postoperative demonstrating that the location once occupied by the epithelium had be repla ced by new stromal scar tissue In addition to validating the observation that the fibrotic mass fills the wound volume, the pattern of tissue deposition reported mirrors the onset and spread of haze reported in the previous chapter. Given these facts, t he following conundrum arises: onceptualized in two separate ways, in terms of the cells present ( Figure 5 2 ), or in crude terms of the major structural proteins present ( Figure 5 3 ). Given the observed timeline and location of haze emergence and the localization of the key effectors of haze formation, two new hypotheses emerge. The first is centered on the fact that the epithelium is the first cell layer to come into contact with the unwounded stroma, and that it is the first cell layer to fill the wound volume D uring the time period of prior to and during haze formation, CTGF is predominantly localized to the basal epithelium SMA is present there as well. These observations led to the hypothesis that t hese epithelial cells transdifferentiate in place int o mesenchymal like cells to form the new scarred stroma. The second hypothesis is that while the epithelium is the first on the scene, cells from the peripheral stroma can migrate in

PAGE 101

101 between the epithelium and unwounded stroma in a manner reminiscent of t he de novo formation of the stroma during embryonic development 112 113 (Figure 5 4 ) Briefly, the cornea begins as a bilayer with cells from two different sources, one of which gives rise to the epithelium, the other the endothelium. Cells from the neural crest then migrate from the periphery and split the bilayer, effectively displacin g the epithelial layer from the endothelium. Interestingly, the epithelium also begins stratifying in this same period; a process which is also present during corneal wound healing just prior to scar formation. It is important to note, that these might n ot be mutually exclusive mechanisms C ell tracing experiments in the scarring of internal organs found that cells from multiple sources were present in the scar 114 115 In order to elu ci date which cells are ultimately responsible for forming the light reflecting scar an epithelial cell traci ng experiment was conducted. Concurrently, a series of experiments using excimer wounded rabbit eyes sought out evidence of grossly observable cellular interchange in be tween the epithelium and stroma (or vice a versa), and the presence of tenascin C during haze formation since it has been proposed to be a marker for the epith elial to mesenchymal transition 116 (EMT) Materials and Methods Cell Tracing Experiment Reporter Mice In order to test whether the epithelium contribute cells to the final scar, a genetically labeled reporter mouse model was used. The reporter mice employ the Cre recombinase system 117 to de galactosidase reporter enzyme 118 Briefly, these mice galactosidase gene (LacZ) which r ecombinase which provides the cell which

PAGE 102

102 prevents the reporter enzyme from being translated The stop codon is flanked on either side by a cis acting sequence (loxp) which serv e s to mark the stop condon for Cre mediated removal by DNA recombination (Figure 5 5A) For this project, the Cre recombinase is under the transcriptional control of the PAX6 promoter (Figure 5 5B) This promoter is active in various tissues in the eye, but in the cornea it is solely active in the corneal epithelium 119 In the cornea s of the reporter mice c ells derived from the epithelium will p galactosidase activity, while cells from the stroma will not. The pups were genotyped and marked by ear punch prior to weaning (postnatal day 21). Tail biopsies were collected with sharp bead sterilized surgical scissors. Silver nitrate was applied to the cut tail fo r hemostasis. Genomic DNA was isolated from the tail biopsies using a Sigma REDExtract N Amp kit. The primers listed in Table 5 1 were used to detect the presence of both necessary transgenes The polymerase chain reaction (PCR) reactions were al ways ru n with a negative control comprised solely of 20ul of master mix. The PCR amplified samples were resolved o n a 1.5% agarose gel with 100 base pair ladder galactosidase transgene has a 220 base pair amplicon, while the PAX6 Cre has a 270 base pair amplicon. Excimer Laser Wounding Slides from rabbit samples generated for the CTGF localization experiment described in Chapter 3 were used for the H&E analysis, while s lides from the previous chapter on haze formation were used to measure tenascin C for use in this chapter No further rabbits were needed for this experiment s described in this chapter Mice were placed in a sealed box and anesthetized with 3.5% isoflurane/oxygen. Once a mouse was anesthetized, it was removed from the box and placed stra ddling a 50 ml conical tube and held in place with a semi adhesive elastic band gently wrapped around both

PAGE 103

103 the mouse and tube. The tube was outfitted with a nose cone fed supply of isoflurane/oxygen. The mouse was laid on its side and supported by a stac k of cotton gauze. The eye to be wounded received a drop of proparacain. The whiskers were cut with scissors and the eye lid and lashes were gently pushed out of the way with a cotton swab. The mouse was then oriented beneath the excimer laser. A 1.0 m m diameter by 24 the following weeks to determin e whether and to what extent the wound scarred Reporter Mouse Tissue Harvesting, Processing, and Sectioning At the terminal time point, the mice were anesthetized as before and euthanized by cervical dislocation. Each globe was immediately enucleated via blunt dissection with fine tipped forceps and placed in fresh 10% neutral buffered formalin on ice. The globes were punctured w ith a 25 gauge needle after 30 min to improve fixative penetration. The globes were intentionally under fixed for 1 hour at 4C to preserve the galactosidase reporter. The globes were then placed in 30% sucrose in PBS overnigh t at 4C to cryoprotect the tissue. In order to improve cryosectioning, each globe was grossly prepared by cutting open the posterior retina, removing the lens, and injecting embedding medium (OCT) into the globe with a blunt tipped needle and syringe. T he tissue was then submerged in OCT and rapidly frozen. The frozen blocks were stored at 20C until cryosectioning. Ten micron section were cut and mounted on poly L lysine coated slides. The slides were air dried and then stored at 20C until staining. Galactosidase Detection The sections were rehydrated and washed in PBS 3 times in PBS. The slides were then incubated in 5 bromo 4 chloro 3 indolyl beta D galactopyranoside ( XGal )

PAGE 104

104 staining solution (2 mM MgCl 2 (Sigma), 5 mM K 4 (CN) 6 3H 2 O (Sigma ), 5 mM K 3 (CN) 6 (Sigma), 1 mg/ml XGal (Promega), in PBS pH 7.4) at 37C in a humidified chamber overnight. The slides were then washed 3 times with PBS, counter stained with nuclear fast red (Vector Labs, CA). The sections were dehydrated and permanently mounted with Permount (Fisher Scientific). The slides were imaged with a light microscope and mounted Nikon D7000 dSLR. Molecular and Histological Analysis Following euthanization, each cornea was briefly irrigated with 10% neutral buffered formalin (10% NBF) and then each globe was enucleated. In order to preserve the shape of the cornea, each globe was placed cornea down into a 12 well tissue culture dish well that was filled with 10% NBF for one hour. The cornea was then excised and placed in another well with fresh 10% NBF and was fixed for 4 to 18 hours at 4C. Frozen sections Following fixation, the corneas chosen for cryosectioning were bisected and then cryoprotected in 30% sucrose in PBS overnight at 4C. The cornea halves were then oriented in OCT and quickly frozen on dry ice. The OCT blocks were stored at 20C L lysine coated glass slides and then ai r dried overnight. The sections were stored at 20C until staining. Paraffin sections Corneas chosen for paraffin embedding and sectioning were bisected and then transferred into 70% ethanol and kept at 4C until submission for processing at the McKnight CTAC

PAGE 105

105 L lysine coated glass slides. One slide per cornea was stained with hematoxylin & eosin (H&E) for gross histological analysis. Immunohistochemistry Paraffin sections were de paraffinized with xylenes and rehydrated though a graded series of ethanol though tap water. Frozen sections were rehydra ted with PBS. The rehydrated sections were then rinsed with wash buffer (phosphate buffered saline with 0.05% Tween 20, PBST). The sections were blocked with 10% normal horse serum (NHS) for 30 minutes at room temperature. The blocking solutions were ca refully removed from one section group per slide by aspiration and blotting. The tissue was then blocked for endogenous avidin and biotin. The slides were then washed three times for 5 times with wash buffer. The primary tenascin C ( TNC ) antibody was di luted (1:500) in 1% NHS in PBST and was incubated for 1 hour at room temperature on the one section while the negative control remained in blocking solution. Following this incubation, the slides were washed three times for 5 minutes with wash buffer. Th e sections were then incubated with a biotinylated secondary horse anti mouse antibody for 30 min at room temperature. The sections were then washed 3 times for 5 minutes with wash buffer. Finally, the sections were incubated with avidin /ml, Vector Labs) for 10 min. The samples were then washed 3 times for 5 minutes each. The slides were mounted with DAPI containing medium and the coverslips sealed with nail hardener. The slides were kept refrigerated in the dark until imaged.

PAGE 106

106 Results Corneal Wounding The reporter mice appear to have thinner and/or weaker than normal corneas The one eye that di present in this one cornea by day 7 (Figure 5 6 ) Even though the scar appeared to be mature by day 7, the eyes were not harvested until day 32 to ensure that a mature scar had formed Cell Tracing Overall the fixation was insufficient for good tissue integrity, and it did not quench galactosidase activity in the epithelium (Figure 5 7 A) O nly one region in the wound possessed a strong positive signal and upon closer inspection t galactosi dase positive region appears to be two cells in close juxtaposition (Figure 5 7B) The majority of the cells in the wound surrounding the two positive cells are galactosidase negative, N uclear Fast Red positive, cells indicatin g that they were not epith elial derived cells. Gross Histology Within 24 hours following wounding, there is evidence of a significant number of neutrophils and epithelial invasion of incompletely ablated stromal lamellae (Figure 5 8 A&B ). One day later, there is evidence of continu ed migration of epithelial cells between stromal lamellae, and evidence of additional regions of stromal invasion (Figure 5 8 C). On day 4 post wounding, one eye had evidence of the epithelium a stromal mass inclusion within the central epithelium (Figure 5 8 D). Three days after wounding, the basal

PAGE 107

107 epithelium is mostly intact with evidence of peripheral stromal cells migrating towards the wound margin and possibly proliferating there as well (F igure 5 9 ). While the epithelium in the day 3 sample was mostly intact, there is a single blister like structure which appears to be surrounded by nuclear staining. By day 5, the blister like structures have become significantly more numerous and are at this time point the predominant feature at the wound interface (Figure 5 10 ). Immunohistochemistry At day 5 following wounding, tenascin C was present in a distinct region within the wound margin and appears to be limited to a single layer of stromal cells (Figure 5 1 1 A). Two days later, tenascin C staining had spread towards the center of the wound and now appears to cover two layers of stromal cells (Figure 5 1 1 C). By day 10, tenascin C staining is present throughout the wound and encompasses about 4 to 5 cells layers in the stroma (Figure 5 11E) The initiation and spread of tenascin C mirrors the spread of haze as was seen earlier (Figures 5 1 1 B, D, & F ). Discussion The data provided by the cell tracing experiment does not support the hypothesis that the newly synthesized strom al tissue is derived from the epithelium But, the combination of the cell tracing experiment and the gross histology indicate that epithelial cells can penetrate into the stroma, adopt a low fibroblastic profile, and remain emb edded in the stroma up to 32 days after wounding. A study conducted on post mortem corneas from individuals who had undergone laser assisted in situ keratomileusis (LASIK) at various times prior to death found that portions of the reflective s cars in the center of the LASIK treated cornea s were cells in various states of necrosis 23 In LASIK, a microkeratome or laser is used to

PAGE 108

108 make an incomplete cut along the stromal lamellae with a hinge like portion left in place n laid back down on the ablated surface. If the adhe re well, the cut provides an easy conduit for epithelial cells to migrate into the stroma. With the laser surgery techniques which ablate directly through the epithelium, such as phototherapeutic (PTK) and an immediately obvious avenue for invasion into the stroma. However, examination of a cornea one day after PTK surgery (Figure 5 12A) presents a potential mechanism. The migrating epithelial front is clearly visible (white arrows) as is the presence of a proteinaceous slough (green arrows) While lasers might make highly reproducible wounds, the cut cannot account for the heterogeneity of the interwoven structure of the stroma ; e ven a very clean cut will have completely removed stromal lamellae. If the migrating epithelial front migrates into one of these disrupted stromal lamellae at a certain angle, the stromal lamellae will direct the migrating epithelial cells into the stroma Figure 5 12B While epithelia l cells do enter and remain in the stroma and become light reflecting entities 23 t he total area covered by t he sole galactosidase positive cluster observed in the reporter mouse (Figure 5 7B) cannot account for the e ntire area of opacity that was observed (Figure 5 6C) The first hypothesis proposed a mechanism which entailed epithelial cells turning into myofibroblasts via epithelial to mesenchymal transition (EMT). Connective tissue growth factor smooth muscle actin, and tenascin C were all present within the scarring corneas, and they have been implicated as either effectors (CTGF) or markers

PAGE 109

109 of EMT 116 120 121 But the facts presented herein, especially the cell tracing experiment, indicate that the preponderance of cells in the scarring tissue come from the stroma, not large, the result from studies in systems where EMT is known to occur due to the complete lack of fibroblasts (such as the lens) or where there is a paucity of fibrobl asts (like the heart and brain) In effect, these protein s are markers within a highly specific context they are not universal What does appear to be the case given the evidence presented here is that these proteins are markers of wound healing and that in tissues without fibroblasts, epithelial cells can provide the wound healing activities and factors r epresented by these markers. Within the EMT literature, t here are some who are backing away from protein markers of EMT and from the premise that epithelial cells are turning into fibroblasts 111 Instead, the emerging idea is that epithelial cells which take on mesenchymal like characteristics such as loss of tight cell cell contact, flattening of the cellular profile, and migration and/or invasion into surrounding tissue have effectively undergone EMT. On this basis, the invasion and implantation of epithelial cells into the stroma demonstra ted here would classify as EMT. Overall, this shift in EMT paradigm appears to be a mere re labeling of metastasis, and calls into question whether EMT as a separate research focus has any merit. Without question, understanding the process and mechanisms of invasive epithelial cells is important for understanding diseases (particularly in cancer) but to generate a new redundant, concept just serves to dilute the attention of those studying these processes The evidence presented in this chapter greatly supports a mechanism of scar formation which is very similar to the de novo formation of the stroma during

PAGE 110

110 development. In the days preceding the emergence of the first signs of visible haze, stromal fibroblasts are concentrated along the stromal lamellae at the wound margin. Concurrent with the first appearance of haze, the basal epithelium begins losing contact with the unwounded surface of the stroma by a mechanism which most closely resembles blistering. These two observations provide two additional hypothetical targets for anti fibrotic therapy in the cornea. Given that the fibroblasts are migrating to the epithelial stromal interface, there is likely a chemotactic molecule being secreted by the basal epithelium. Identification and neutralization o f the chemotactic factor(s) would thereby prevent the haze cells from ever arising. Alternatively, a therapeutic strategy would enable the clear epithelium to remain in place thus preventing the opaque myofibroblasts from migrating into and accumulating in the wound. Going back to consider the structural protein based conceptualization of the wound healing conundrum, there is a hypothetical proteinase activity responsible for which could possibly give rise to the observed blister like structures Interestingly enough, this concept is the proverbial other side of the coin for the haze formation hypothesis that preceded the current cellular theory of haze formation The initial theory was that the immature collagen bundles were the source of haze and that preventing the synthesis of collagen was the key to preventing haze. In order for there to be room for the collagen producing cells to Given its temporal precedence, one would expect that inhibiting the removal of keratins would

PAGE 111

111 also prevent th e accumulation of collagen, but not vice a versa; making the hypothetical keratinolytic activities a better therapeutic target than those giving rise to collagen synthesis. All together, the data support a mechanism of scar formation that is similar to the process by which the corneal stroma is first generated. A new model which integrates all of the evidence presented herein is presented in Figure 5 13. Following a transepithelial stroma penetrating wound (Figure 5 13A), the epithelial cells migrate into the wound volume covering the stromal surface (Figure 5 13B) A portion of the migrating epithelial front occasionally migrate s into the stroma along incompletely removed stromal lamellae (Figure 5 13 C ). The b asal epithelium where CTGF was predominant ly found, appear s to be the destination for stroma derived cells migrating from the periphery of the wound (Figure 5 13D). Around the time that the stroma derived cells arrive at the basal epithelium, a blister like process displaces the epithelium from t he wound surface thereby making room for the stroma derived cells to populate the wound interface and subsequently form the light reflecting scar (Figure 5 13E) With time, these cells proliferate into multiple layers and form a scar which can fill the en tire wound void (Figure 5 13F). With this new model in hand, one issue is still not clear. If the process is similar to de novo stromagenesis, then why are the cells opaque? Clearly, these cells share a common lineage, but something either happens to them in the time following development, or as was seen in the fetal fibrosis experiments, a factor is present in the adult tissue which causes them to change their phenotype Wor k to identify the key difference(s) between the neural crest cells in development and the resident fibroblasts

PAGE 112

112 in adult tissues, and between the factors present during stromagen e sis versus fibrosis must be identified if a truly regenera tive response is to ever be obtained

PAGE 113

113 Table 5 1 Primers for the mice with genetically labeled corneal epithelium Name Sequence R26R F 5' TTT CCA CAG CTC GCG GTT GAG GAC 3' R26R R Cre F Cre R 5' CTA AAG CGC ATG CTC CAG ACT GCC 3' 5' GCC GTA AAT CAA TCG ATG AGT 3' 5' TGA CGG TGG GAG AAT GTT AAT A B C D Figure 5 1. Evidence that the scar is formed in de novo synthesized tissue. A) & B) SMA, respectively, in a day 28 post wounding scar. The arrows indicate the border between scar and non wounded tissue, not that the scar is in the shape of the wound volume. C) & D) A photograph an SMA, respectively, in a day 14 wound. It is the worst scar observed in this project SMA staining is within the sharply delimited wound boundaries

PAGE 114

114 A B C Figure 5 2. The cellular repre sentation of the haze generation conundrum. A) The initial wound viewed at the wound margin. B) By day 5, the wound is filled with epithelial cells. C) But, by day 14 the epithelial cells have been replaced or displaced by myofibroblasts.

PAGE 115

115 A B C Figure 5 3 An unresolved conundrum with the current prevailing theory The conundrum is represented in the context of the major structural proteins present in the healing wound.

PAGE 116

116 Figure 5 4 A schematic representation of de novo stromagenesis that occurs during development. A similar mechanism might be behind the corneal wound healing process. A B Figure 5 5 The two transgenes which lead to an irreversible genetically labeled corneal epithelium. A) The conditional reporter enzyme transgene. In the presence of Cre enabling translation of the full galactosidase encoded by the LacZ gene. B) Is a scematic of the corneal epithelium specific promote r driving Cre recombinase expression.

PAGE 117

117 A B C Figure 5 6 Normal and sca r red mouse eyes. A) A normal non scared mouse cornea. B) A representative PTK scar at day 7 post wounding in a ROSA26R/PAX6 Cre mouse eye. C) The same eye in B) one week later and emphasized by grayscale conversion. A B Figure 5 7 A single strong positive blue mass comprised of at most two cells in the anterior center of the stroma. A) Low power and B) higher power image with detail from the center of the cornea. There

PAGE 118

118 A B C D Figure 5 8 Epithelial invasion of the stroma during re epithelialization. A) An H&E wide field mosaic of a day 1 post wounding cornea. B) A higher power image of the region highlighted region. C) An autofluorescence and DAPI images of a day 2 post wounding cornea where the initial invasion has progressed significantly. D) An H&E low power mosaic of a day 4 post wounding cornea. The inva ding epithelial front has effective delaminated several lamellae and generated an included mass.

PAGE 119

119 A B Figure 5 9 Epithelial hyperplasia at the wound margin and migration of stroma derived cells into the wound interface. A) A low power H& E image and B) a higher power image from the wound margin in the same section.

PAGE 120

120 A B C Figure 5 10 The loss of epithelial attachment via a blistering like mechanism. A) Five days after wounding, a region of blistering between the basal epithelium and wounded stroma (right of the arrow) begins to appear while the epithelium outside of the wound (left of the arrow) appears fine. B&C) Detail of the blister like structu res.

PAGE 121

121 A B C D E F Figure 5 1 1 The distribution of tenascin C during haze formation. A), C), & E) are immunofluorescent micrographs from days 5, 7, & 10, respectively. B), D), & F) Macrophotographs of the haze present in the corneas just prior to tissue harvesting. These are separate sections from the same corneas that were SMA (Figure 4 2) Note, that the pattern of staining for tenascin SMA.

PAGE 122

122 A B Figure 5 12. A potent ial mechanism for the observed epithelial invasion of the stroma. A) A cornea one day after PTK surgery. The white arrows indicate the migrating epithelial front while the green arrows indicate peeling residual stromal lamellae. B) If the migrating epit helial front aligns with the peeling stromal lamellae, then the epithelial cells can enter and become embedded in the stroma. A B C D E F Figure 5 1 3 A new cellular model for the generation of corneal haze. A) The initial wound. B) The migrating epithelial front (day 1). C) Occasional epithelial invasion of the stroma (days 2 3). D) Epithelial hyperplasia at the wound margin and keratocytes migration into the wound (days 3 4). E) Stacking of n the sub epithelial stroma and loss of epithelial adhesion via a blistering like mechanism. E) The epithelial implant is surmounted by stromal cell derived scar cells and matrix which form the preponderance of light reflecting material

PAGE 123

123 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS A New Therapeutic Modality Validated Iontophoresis of small molecules has been a viable technology for decades, but there are substantive barriers to adapting the process for the delivery of larger macrom olecular therapies like mRNA ablating oligonucleotides A few groups have been working on using iontophoresis to deliver oligonucleotides into skin 103 and corneas 102 122 Both have reported success in delivery, but only the group treating skin had any measurable biochemical effect. Here, I have demonstrated a change in a physiological process as a consequence of iontophoretically delivered, gene targeted, therapy in the cornea. This success justifies continued attention to refining the method and devi ce in for use in subsequent studies of other gene products within the cornea A New Standard in Visualizing and Reporting Corneal Haze The current standard in measuring and reporting corneal haze essentially represents a qualitative assessment of the obs erved scar. The approach reduces a complex geometric distribution across the wounded area into a single scalar number on a scale from 0 to 4. It is currently impossible to objectively compare results from other investigations within this field of study, or e ven from one article to another written by the same group. The new method of recording and quantifying corneal scars presented here has the ability to improve both the communication of and comparison of results amo ngst groups which can drastically improve the rate of progress in research targeted at preventing corneal fibrosis That said, there are still aspects of the method described here that need refinement and standardization. Future work will focus on ensuring that the method is amenable to variations in the choice of imaging platform, sensor size, and

PAGE 124

124 camera placement as well as operator to operator variations Most of these issues will be greatly remedied with the use of external standards placed next to the wound, which can be included within the image and thereby enable corrections for differences in exposure (intensity) and camera placement (distortion) Connective Tissue Growth Factor in Healing Corneas The success in reducing connective tissue growth factor (CTG F) with an ASO and the observed reduction in haze that followed served to greatly support the hypothetical role of CTGF in corneal haze formation. The additional evidence that CTGF protein was primarily localized to the basal epithelium, where the scar wi ll eventually be further supported its hypothetical role in scarring owing to it being in the right place at the right time. Had CTGF not been highly concentrated in the wound prior to the formation of haze, and more importantly in the location of where the scar tissue was growing, then its role in scarring would have been difficult to support. While the evidence does support a role for CTGF in scarring, the fact that it was predominantly found on the epithelium calls into question what, precisely, that role is. The evidence provided here demonstrate that the location of CTGF binding, epithelial blistering, and the destination of the migrating fibroblasts are coincident. Whether CTGF stimulates the blistering, or the secretion of chemotactic signals fo r the fibroblasts remains unclear. T he initial hypotheses about the source of CTGF in the cornea was also undercut by the data presented here. The data revealed unexpectedly that CTGF is predominantly produced in the corneal endothelium not the stromal f ibroblasts. That the locations of maximal synthesis and maximal binding were on opposing sides of the cornea was interesting, but difficult to understand at first. T ypically, the net flow of fluid is out of the cornea across the endothelium giving the a ppearance that the endothelial

PAGE 125

125 However, during healing the net flow of fluid is into the cornea as is evidence d by the corneal swelling during the first few days. Additionally, t he fact t hat the peak of CTGF binding and its decline coincide with the peak and tapering off of edema provide s a plausible mechanism for the proposed trans corneal diffusion The stoma is a significant barrier to diffusion of ionic macromolecules until it becomes edematous 110 The sum of these observations is the new th eory that CTGF synthesized by the endothelium is permitted by the edematous stroma to diffuse and bind to the basal epithelium. Given the proposed mechanism of edema permitted diffusion of CTGF from the endothelium to the epithelium, preventing edema migh t be a viable anti fibrotic target in that it would preclude the introduction of CTGF to the epithelium. While the endothelium was the predominant site of CTGF synthesis, it was not the only site. Both the stroma and epithelium had CTGF synthesis that was measurable by highly sensitive real time PCR analysis. An alternative theory to the edema permitted diffusion theory is that the epithelium undergoes a change during healing which permits it to accumulate the low levels of CTGF synthesis present in the e pithelium and stroma. from more conclusively is essential if any mRNA ablating anti CTGF strategy is chosen Given efficacy of the ASOs to reduce both haze and CTGF CTGF is derived from cells in the cornea as opposed to a s ource outside of the cornea (i.e. tear fluid ) Future work in this area will be done using conditional knockout mice which allow semi selective genomic ablation of CTGF in the cornea using the sam e tissue specific Cre recombinase system as was used in the cell tracing experiments By using a panel

PAGE 126

126 of promoters specific for the different cells layers in the cornea the source of pro fibrotic CTGF in the cornea can be determined more conclusively. Ha ze Fibrogenesis The key observation leading to the new theory of haze fibrogenesis was made possible by the novel haze imaging and measuring technique developed herein. The fact that the haze begin s in two distinct loci within the wound and then spreads from these loci of nucleation calls into question the currently espoused theory that underlying keratocytes spiral upwards into the wound, differentiate, and then adopt a light reflecting phenotype. If this was the case, then it is expected that the haze would have uniformly appeared and gradually increased in intensity The histological evidence of fibroblasts accumulating along the stromal lamellae at the wound margin is evidence that the haze forming cells migrate into the wound interface from the periphery. This mechanism is consistent with the observed lateral spread of haze since the peripheral fibroblasts have to migrate a longer distance to get to the more central regions of the wound. Given a constant rate of migration, haze would appear to spread as the migrating fibroblasts made progress towards the center of the wound. The separate haze initiation at the center of the wound is not as consistent in occurrence or location as the haze at the margin, indicating an element of randomness in the nucleating event. The image data from the CTGF antisense experiments indicated that the haze at the margin was sensitive to CTGF reduction while the central haze was refractory. The gross histological an d cell tracing evidence demonstrated that the migrating epithelial cells can enter the stroma along incompletely removed lamellae. Given that the epithelial invasion of the stroma is expected to be independent of CTGF, it is possible that the epithelial i nvasion serves as another mechanism by which the

PAGE 127

127 central haze is initiated and propagated. In the cell tracing experiment in the mouse cornea, and in study of the corneas from post mortem human LASIK patients 23 the to tal contribution to haze of these epithelial islands appears to be minimal. While the gross histological data from the rabbits indicate that the invading structures can be more substantial it is not clear whether the observed larger epithelial inclusion s persist, and the degree to which they contribute to the central haze. Since the rabbit PTK model is the more relevant model for transepithe lial stroma penetrating lesions, the degree of contribution to the central corneal haze of the epithelial inclusio ns is the focus of studies currently being planned. Expected Clinical Impact The majority of the findings presented here are expected to have the ir initial impact in the research and testing of new haze preventing agents; any clinical impact would be c on tingent upon the outcomes of those research projects Of these findings, the new insights into the mechanism of haze formation is expected to have the largest, long range, clinical impact. That said, some of the other findings have the potential for im mediate clinical impact. T he first of these findings with immediate potential arises from the hypothesis that edema might be necessary for CTGF to be introduced into the woun d While this hypothesis is still immature scientifically, the approach and agents for testing it are already used in the clinic and are relatively benign. Currently, hypertonic saline is commonly used to reduce corneal swelling following the placement of c orneal grafts. Given its clinical availability and nearly non existent toxicity (compared to mitomycin C), a trial of topically applied hypertonic saline as a means of preventing corneal haze in wounds could have an immediate and drastic clinical impact.

PAGE 128

128 While the wound photography and quantification technique is expected to be most useful in a research and testing environment, the technique could be used to make the of the shelf equipme nt, high resolution, highly detailed images of the lesion (fluorescein stained) or of the scar/opacity could be recorded and tracked with time. The greatest anticipated clinical value of including such images in the charts is that the complete coverage of the wound with a high resolution image, or series of images, would greatly facilitate consultation with more experienced colleagues from all over the globe (via telemedicine). Additionally, the inclusion of these images in to medical records would create an invaluable repository for future medical researc h and training. Closing Remarks The work presented here severely undercut some initial hypotheses, validated others and led to several newly generated ones as well While the data presented here have provided further insight into, and immediately testable hypothesis about, the process of haze formation in the cornea, the data did not provide a definitive solution for the prevention or reversal of corneal haze. Whil e the ultimate goal of medical research is to solve problems pertaining to human health and wellness, progress in medical research is measured by a reduction in the ambiguity surrounding any given medical problem. By this standard, I submit that the resea rch presented here in was fruitful.

PAGE 129

129 LIST OF REFERENCES 1. Border, W.A. & Ruoslahti, E. Transforming growth factor beta in disease: the dark side of tissue repair. J Clin Invest 90 1 7 (1992). 2. Ito, Y. et al. Expression of connective tissue growth factor in human renal fibrosis. Kidney Int 53 853 861 (1998). 3. Gressner, A.M., Weiskirchen, R., Breitkopf, K. & Dooley, S. Roles of TGF beta in hepatic fibrosis. Front Biosci 7 d793 807 (2002). 4. Lijnen, P.J., P etrov, V.V. & Fagard, R.H. Induction of cardiac fibrosis by transforming growth factor beta(1). Mol Genet Metab 71 418 435 (2000). 5. Thompson, N.L. et al. Transforming growth factor beta 1 in acute myocardial infarction in rats. Growth Factors 1 91 99 (1988). 6. Chegini, N. The role of growth factors in peritoneal healing: transforming growth factor beta (TGF beta). Eur J Surg Suppl 17 23 (1997). 7. Igarashi, A. et al. Connective tissue growth factor gene expression in tissue sections from localized s cleroderma, keloid, and other fibrotic skin disorders. J Invest Dermatol 106 729 733 (1996). 8. Sherwood, M.B. A sequential, multiple treatment, targeted approach to reduce wound healing and failure of glaucoma filtration surgery in a rabbit model (an Ame rican Ophthalmological Society thesis). Trans Am Ophthalmol Soc 104 478 492 (2006). 9. Connor, T.B., Jr. et al. Correlation of fibrosis and transforming growth factor beta type 2 levels in the eye. J Clin Invest 83 1661 1666 (1989). 10. Moller Pedersen, T., Cavanagh, H.D., Petroll, W.M. & Jester, J.V. Neutralizing antibody to TGFbeta modulates stromal fibrosis but not regression of photoablative effect following PRK. Curr Eye Res 17 736 747 (1998). 11. Esson, D.W. et al. Expression of connective tissue growth factor after glaucoma filtration surgery in a rabbit model. Invest Ophthalmol Vis Sci 45 485 491 (2004). 12. Krachmer, J.H.M., Mark J.; Holland, Edward J. Cornea (Elsevier Mosby, 2005). 13. Gordon Thomson, C., de Iongh, R.U., Hales, A.M., Chamber lain, C.G. & McAvoy, J.W. Differential cataractogenic potency of TGF beta1, beta2, and beta3 and their expression in the postnatal rat eye. Invest Ophthalmol Vis Sci 39 1399 1409 (1998).

PAGE 130

13 0 14. Hales, A.M., Chamberlain, C.G., Dreher, B. & McAvoy, J.W. Intravitreal injection of TGFbeta induces cataract in rats. Invest Ophthalmol Vis Sci 40 3231 3236 (1999). 15. Lee, E.H. & Joo, C.K. Role of transforming growth factor beta in transdifferentiation and fibrosis of lens epithelial cells. Inves t Ophthalmol Vis Sci 40 2025 2032 (1999). 16. Zadunaisky, J.A. Active transport of chloride in frog cornea. The American journal of physiology 211 506 512 (1966). 17. Zadunaisky, J.A. Active transport of chloride across the cornea. Nature 209 1136 1137 (1966). 18. Hodson, S. & Miller, F. The bicarbonate ion pump in the endothelium which regulates the hydration of rabbit cornea. The Journal of physiology 263 563 577 (1976). 19. Dohlman, C.H. The function of the corneal epithelium in health and disease. T he Jonas S. Friedenwald Memorial Lecture. Invest Ophthalmol 10 383 407 (1971). 20. Wilson, S.E. & Kim, W.J. Keratocyte apoptosis: implications on corneal wound healing, tissue organization, and disease. Invest Ophthalmol Vis Sci 39 220 226 (1998). 21. Zieske, J.D. Extracellular matrix and wound healing. Curr Opin Ophthalmol 12 237 241 (2001). 22. Kuo, I.C. Corneal wound healing. Curr Opin Ophthalmol 15 311 315 (2004). 23. Dawson, D.G. et al. Ex vivo confocal microscopy of human LASIK corneas with his tologic and ultrastructural correlation. Ophthalmology 112 634 644 (2005). 24. Moller Pedersen, T. Keratocyte reflectivity and corneal haze. Exp Eye Res 78 553 560 (2004). 25. O'Brart, D.P. et al. Night vision after excimer laser photorefractive keratec tomy: haze and halos. Eur J Ophthalmol 4 43 51 (1994). 26. O'Brart, D.P. et al. Disturbances in night vision after excimer laser photorefractive keratectomy. Eye 8 ( Pt 1) 46 51 (1994). 27. Moller Pedersen, T., Cavanagh, H.D., Petroll, W.M. & Jester, J.V. Corneal haze development after PRK is regulated by volume of stromal tissue removal. Cornea 17 627 639 (1998).

PAGE 131

131 28. Long, Q. et al. Correlation between TGF [beta]1 in tears and corneal haze following LASEK and Epi LASIK. American Journal of Ophthalmology 143 195 195 (2007). 29. Stramer, B.M., Zieske, J.D., Jung, J.C., Austin, J.S. & Fini, M.E. Molecular mechanisms controlling the fibrotic repair phenotype in cornea: implications for surgical outcomes. Invest Ophthalmol Vis Sci 44 4237 4246 (2003). 30. Alio, J.L. et al. Ten year follow up of photorefractive keratectomy for myopia of more than 6 diopters. Am J Ophthalmol 145 37 45 (2008). 31. Alio, J.L. et al. Ten year follow up of photorefractive keratectomy for myopia of less than 6 dio pters. Am J Ophthalmol 145 29 36 (2008). 32. Boulton, M., Singh, A., Wong, HC., and McLeod, D. Dexamethasone directly and indirectly inhibits the proliferation of retinal microvascular endothelial cells. Invest ophthalmol vis sci 29 243 (1988). 33. Tuft, S.J., Zabel, R.W. & Marshall, J. Corneal repair following keratectomy. A comparison between conventional surgery and laser photoablation. Invest Ophthalmol Vis Sci 30 1769 1777 (1989). 34. Gartry, D.S., Kerr Muir, M. & Marshall, J. The effect of topical corticosteroids on refraction and corneal haze following excimer laser treatment of myopia: an update. A prospective, randomised, double masked study. Eye 7 ( Pt 4) 584 590 (1993). 35. Baek, S.H., Chang, J.H., Choi, S.Y., Kim, W.J. & Lee, J.H. The effect of topical corticosteroids on refractive outcome and corneal haze after photorefractive keratectomy. J Refract Surg 13 644 652 (1997). 36. Agren, M.S., Mirastschijski, U., Karlsmark, T. & Saarialho Kere, U.K. Topical synthetic inhibitor of matrix metallop roteinases delays epidermal regeneration of human wounds. Exp Dermatol 10 337 348 (2001). 37. Vetrugno, M., Maino, A., Quaranta, G.M. & Cardia, L. The effect of early steroid treatment after PRK on clinical and refractive outcomes. Acta Ophthalmol Scand 7 9 23 27 (2001). 38. Tomasz, M. Mitomycin C: small, fast and deadly (but very selective). Chem Biol 2 575 579 (1995). 39. Tomasz, M. & Palom, Y. The mitomycin bioreductive antitumor agents: cross linking and alkylation of DNA as the molecular basis of the ir activity. Pharmacology & therapeutics 76 73 87 (1997). 40. Roh, D.S. & Funderburgh, J.L. Impact on the corneal endothelium of mitomycin C during photorefractive keratectomy. J Refract Surg 25 894 897 (2009).

PAGE 132

132 41. Kopp, E.D. & Seregard, S. Epiphora as a side effect of topical mitomycin C. The British journal of ophthalmology 88 1422 1424 (2004). 42. Coppens, G. & Maudgal, P. Corneal complications of intraoperative Mitomycin C in glaucoma surgery. Bulletin de la Societe belge d 'ophtalmologie 19 23 (2010). 43. Zhivov, A., Beck, R. & Guthoff, R.F. Corneal and conjunctival findings after mitomycin C application in pterygium surgery: an in vivo confocal microscopy study. Acta ophthalmologica 87 166 172 (2009). 44. Ferguson, M.H., GF. Marsupial models of scarless fetal wound healing. in Fetal Wound Healing (ed. Adzick, N.L.M.) 95 124 (Elsevier, New York, 1992). 45. Longaker, M.T. et al. Adult skin wounds in the fetal environment heal with scar formation. Ann Surg 219 65 72 (1994). 46. Shah, M., Foreman, D.M. & Ferguson, M.W. Neutralisation of TGF beta 1 and TGF beta 2 or exogenous addition of TGF beta 3 to cutaneous rat wounds reduces scarring. J Cell Sci 108 ( Pt 3) 985 1002 (1995). 47. Van Obberghen Schilling, E., Roche, N.S., F landers, K.C., Sporn, M.B. & Roberts, A.B. Transforming growth factor beta 1 positively regulates its own expression in normal and transformed cells. The Journal of biological chemistry 263 7741 7746 (1988). 48. Bascom, C.C. et al. Complex regulation of transforming growth factor beta 1, beta 2, and beta 3 mRNA expression in mouse fibroblasts and keratinocytes by transforming growth factors beta 1 and beta 2. Mol Cell Biol 9 5508 5515 (1989). 49. Jester, J.V., Barry Lane, P.A., Petroll, W.M., Olsen, D.R. & Cavanagh, H.D. Inhibition of corneal fibrosis by topical application of blocking antibodies to TGF beta in the rabbit. Cornea 16 177 187 (1997). 50. Whitby, D.J. & Ferguson, M.W. Immunohistochem ical localization of growth factors in fetal wound healing. Dev Biol 147 207 215 (1991). 51. Saika, S. TGF beta signal transduction in corneal wound healing as a therapeutic target. Cornea 23 S25 30 (2004). 52. Tuli, S.S. et al. Immunohistochemical loca lization of EGF, TGF alpha, TGF beta, and their receptors in rat corneas during healing of excimer laser ablation. Curr Eye Res 31 709 719 (2006). 53. Vesaluoma, M., Teppo, A.M., Gronhagen Riska, C. & Tervo, T. Release of TGF beta 1 and VEGF in tears foll owing photorefractive keratectomy. Curr Eye Res 16 19 25 (1997).

PAGE 133

133 54. Yoshino, K., Garg, R., Monroy, D., Ji, Z. & Pflugfelder, S.C. Production and secretion of transforming growth factor beta (TGF beta) by the human lacrimal gland. Curr Eye Res 15 615 624 (1996). 55. Laping, N.J. et al. Tumor specific efficacy of transforming growth factor beta RI inhibition in Eker rats. Clinical cancer research : an official journal of the American Association for Cancer Research 13 3087 3099 (2007). 56. Franch, H.A., Shay, J.W., Alpern, R.J. & Preisig, P.A. Involvement of pRB family in TGF beta dependent epithelial cell hypertrophy. The Journal of cell biology 129 245 254 (1995). 57. Igarashi, A., Okochi, H., Bradham, D.M. & Grotendorst, G.R. Regulation of connective tissue growth factor gene expression in human skin fibroblasts and during wound repair. Mol Biol Cell 4 637 645 (1993). 58. Frazier, K., Williams, S., Kothapalli, D., Klapper, H. & Grotendorst, G.R. Stimulation of fibroblast cell growth, matrix production and granulation tissue formation by connective tissue growth factor. J Invest Dermatol 107 404 411 (1996). 59. Grotendorst, G.R. Connective tissue growth factor: a mediator of TGF beta action on fibroblasts. Cytokine Growth Factor Rev 8 171 179 (1997). 60. Garrett, Q. et al. Involvement of CTGF in TGF beta1 stimulation of myofibroblast differentiation and collagen matrix contraction in the presence of mechanical stress. Invest Ophthalmol Vis Sci 45 1109 1116 (2004). 61. Daniels, J.T. et al. Mediation of transforming growth factor beta(1) stimulated matrix contraction by fibroblasts: a role for connective tissue growth factor in contractile scarring. Am J Pathol 163 2043 2052 (2003). 62. Bradham, D.M., Igarashi, A., Potter, R.L. & Grotendorst, G.R. Co nnective tissue growth factor: a cysteine rich mitogen secreted by human vascular endothelial cells is related to the SRC induced immediate early gene product CEF 10. J Cell Biol 114 1285 1294 (1991). 63. Blalock, T.D. et al. Connective tissue growth fac tor expression and action in human corneal fibroblast cultures and rat corneas after photorefractive keratectomy. Invest Ophthalmol Vis Sci 44 1879 1887 (2003). 64. Baldwin, H.C. & Marshall, J. Growth factors in corneal wound healing following refractive surgery: A review. Acta Ophthalmol Scand 80 238 247 (2002). 65. He, S. et al. Connective Tissue Growth Factor as a Mediator of Intraocular Fibrosis. Invest. Ophthalmol. Vis. Sci. iovs.07 1302 (2008).

PAGE 134

134 66. Hinton, D.R. et al. Accumulation of NH2 terminal fragment of connective tissue growth factor in the vitreous of patients with proliferative diabetic retinopathy. Diabetes Care 27 758 764 (2004). 67. Brigstock, D.R. et al. Purification and characterization of novel heparin binding growth factors in uterine secretory fluids. Identification as heparin regulated Mr 10,000 forms of connective tissue growth factor. J Biol Chem 272 20275 20282 (1997). 68. Grotendorst, G.R. & Dun can, M.R. Individual domains of connective tissue growth factor regulate fibroblast proliferation and myofibroblast differentiation. FASEB J 19 729 738 (2005). 69. Grotendorst, G.R., Rahmanie, H. & Duncan, M.R. Combinatorial signaling pathways determine f ibroblast proliferation and myofibroblast differentiation. FASEB J 18 469 479 (2004). 70. Dawson, D.G., Kramer, T.R., Grossniklaus, H.E., Waring, G.O., 3rd & Edelhauser, H.F. Histologic, ultrastructural, and immunofluorescent evaluation of human laser ass isted in situ keratomileusis corneal wounds. Arch Ophthalmol 123 741 756 (2005). 71. Smiddy, W.E. et al. Transforming growth factor beta. A biologic chorioretinal glue. Arch Ophthalmol 107 577 580 (1989). 72. Cuenoud, B. & Szostak, J.W. A DNA metalloenz yme with DNA ligase activity. Nature 375 611 614 (1995). 73. Dass, C.R., Choong, P.F. & Khachigian, L.M. DNAzyme technology and cancer therapy: cleave and let die. Molecular cancer therapeutics 7 243 251 (2008). 74. Famulok, M., Hartig, J.S. & Mayer, G. Functional aptamers and aptazymes in biotechnology, diagnostics, and therapy. Chem Rev 107 3715 3743 (2007). 75. Khan, A.U. Ribozyme: a clinical tool. Clinica chimica acta; international journal of clinical chemistry 367 20 27 (2006). 76. Maruyama, N., F ukuda, N., Okada, K. & Matsumoto, K. [Development of gene therapy for encapsulating peritoneal sclerosis by a chimeric DNA RNA hammerhead ribozyme targeting TGF beta1 mRNA]. Nippon Jinzo Gakkai Shi 49 113 120 (2007). 77. Wu, Y. et al. Inhibition of bcr abl oncogene expression by novel deoxyribozymes (DNAzymes). Hum Gene Ther 10 2847 2857 (1999). 78. Yusa, J. et al. Inhibition of growth of human gingival fibroblasts by chimeric DNA RNA hammerhead ribozyme targeting transforming growth factor beta 1. J Periodontol 76 1265 1274 (2005).

PAGE 135

135 79. Santoro, S.W. & Joyce, G.F. A general purpose RNA cleaving DNA enzyme. Proc Natl Acad Sci U S A 94 4262 4266 (1997). 80. Bock, L.C., Griffin, L.C., Latham, J.A., Vermaas, E.H. & Toole, J.J. Selection of single stranded DNA molecules that bind and inhibit human thrombin. Nature 355 564 566 (1992). 81. Huizenga, D.E. & Szostak, J.W. A DNA aptamer that binds adenosine and ATP. Biochemistry 34 656 665 (1995). 82. Ellington, A.D. & Szostak, J.W. In vitro selection of RNA molecules that bind specific ligands. Nature 346 818 822 (1990). 83. Waring, R.B., Scazzocchio, C., Brown, T.A. & Davies, R.W. Close relationship between certain nuclear and mitochondrial introns. Implications for t he mechanism of RNA splicing. J Mol Biol 167 595 605 (1983). 84. Cotten, M., Schaffner, G. & Birnstiel, M.L. Ribozyme, antisense RNA, and antisense DNA inhibition of U7 small nuclear ribonucleoprotein mediated histone pre mRNA processing in vitro. Mol Cel l Biol 9 4479 4487 (1989). 85. Zamore, P.D., Tuschl, T., Sharp, P.A. & Bartel, D.P. RNAi: double stranded RNA directs the ATP dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 101 25 33 (2000). 86. Matzke, M.A. & Birchler, J.A. RNAi media ted pathways in the nucleus. Nat Rev Genet 6 24 35 (2005). 87. Hermann, T. & Patel, D.J. Adaptive recognition by nucleic acid aptamers. Science 287 820 825 (2000). 88. Advances in CMV management: fomivirsen (Vitravene) approved. Proj Inf Perspect 7 (199 8). 89. Choi, Y.K. et al. Prevention of tissue injury by ribbon antisense to TGF beta1 in the kidney. Int J Mol Med 15 391 399 (2005). 90. Doh, K.O. et al. Prevention of CCl4 induced liver cirrhosis by ribbon antisense to transforming growth factor beta 1. Int J Mol Med 21 33 39 (2008). 91. Liang, Z. & He, Z. [TGF beta 1 antisense gene transfer into ito cells and suppressed extracellular matrix production]. Zhonghua Yi Xue Za Zhi 78 850 852 (1998). 92. Liu, X., Yang, L., Chen, D., Zhang, Z. & Qiang, O. [Inhibition of the activation and collagen production of cultured rat hepatic stellate cells by antisense oligonucleotides against transforming growth factor beta 1]. Hua Xi Yi Ke Da Xue Xue Bao 31 42 45 (2000).

PAGE 136

136 93. Liu, X., Zhang, Z., Yang, L., Chen, D. & Wang, Y. [Inhibition of the activation and collagen production of cultured rat hepatic stellate cells by antisense oligonucleotides against transforming growth factor beta 1 is enhanced by cationic liposome delivery]. Hua Xi Yi Ke Da Xue Xue Bao 31 133 135, 142 (2000). 94. Lu, C.H. et al. [Effects of antisense RNA of connective tissue growth factor expressing plasmid on rat liver fibrosis]. Zhonghua Gan Zang Bing Za Zhi 15 118 121 (2007). 95. Uchio, K., Graham, M., Dean, N.M., Rosenbaum, J. & Desmoulie re, A. Down regulation of connective tissue growth factor and type I collagen mRNA expression by connective tissue growth factor antisense oligonucleotide during experimental liver fibrosis. Wound Repair Regen 12 60 66 (2004). 96. Choi, B.M. et al. Contr ol of scarring in adult wounds using antisense transforming growth factor beta 1 oligodeoxynucleotides. Immunol Cell Biol 74 144 150 (1996). 97. Excaliard Pharmaceuticals, I. Excaliard Announces Positive Data From Phase 2 Clinical Trial of Its Anti Scarri ng Drug, EXC 001. Vol. 2010 A press release about the results of Excaliard Pharmaceuticals, Inc.'s phase 2 clinical trial. (Carlsbad, CA, U.S.A., 2010). 98. Gibson, D.J. Electromotive delivery of oligonucleotides into the cornea. (University of Florida, [ Gainesville, Fla.], 2007). 99. Gibson, D.J. & Schultz, G. Optimization of Electrode Material and Placement for in vivo Iontophoretic Delivery of Oligonucleotides Into the Stroma. Invest. Ophthalmol. Vis. Sci. 49 3195 (2008). 100. Gibson, D.J. & Schultz, G. High Field Computer Controlled Pulse Iontophoresis of Oligonucleotides Into the Cornea. Invest. Ophthalmol. Vis. Sci. 47 1593 (2006). 101. Gibson, D.J. & Schultz, G.S. Trans Corneal HEPES Buffered Iontophoresis for Delivery of Oligonucleotides Into th e Stroma. Invest. Ophthalmol. Vis. Sci. 48 5879 (2007). 102. Berdugo, M. et al. Delivery of antisense oligonucleotide to the cornea by iontophoresis. Antisense Nucleic Acid Drug Dev 13 107 114 (2003). 103. Kigasawa, K. et al. Noninvasive delivery of s iRNA into the epidermis by iontophoresis using an atopic dermatitis like model rat. Int J Pharm (2009). 104. Butler, M. et al. Spinal distribution and metabolism of 2' O (2 methoxyethyl) modified oligonucleotides after intrathecal administration in rats. Neuroscience 131 705 715 (2005).

PAGE 137

137 105. Dokka, S., Cooper, S.R., Kelly, S., Hardee, G.E. & Karras, J.G. Dermal delivery of topically applied oligonucleotides via follicular transport in mouse skin. The Journal of investigative dermatology 124 971 975 (2005). 106. Roux, J. et al. Transforming growth factor beta1 inhibits cystic fibrosis transmembrane conductance regulator dependent cAMP stimulated alveolar epithelial fluid transport via a phosphatidylinositol 3 kinase dependent mechanism. J Biol Chem 2 85 4278 4290 (2010). 107. Yang, Y.M., Wu, X.Y. & Du, L.Q. [The role of connective tissue growth factor, transforming growth factor and Smad signaling pathway during corneal wound healing]. Zhonghua Yan Ke Za Zhi 42 918 924 (2006). 108. Blake, J. et al. The Mouse Genome Database (MGD): Premier Model Organism Resource for Mammalian Genomics and Genetics. Nucleic Acids Res 39 D842 D848 (2011). 109. Kapoor, M. et al. Connective tissue growth factor promoter activity in normal and wounded skin. Fibrogenesi s & tissue repair 1 3 (2008). 110. Kim, J.H., Green, K., Martinez, M. & Paton, D. Solute permeability of the corneal endothelium and Descemet's membrane. Experimental Eye Research 12 231 238 (1971). 111. Chai, J.Y., Modak, C., Mouazzen, W., Narvaez, R. & Pham, J. Epithelial or mesenchymal: Where to draw the line? Biosci Trends 4 130 142 (2010). 112. Ikeda, A., Maisel, H. & Waggoner, D. An immunofluorescent study of cornea development in the chick. Journal of embryology and experimental morphology 33 279 290 (1975). 113. Gage, P.J., Rhoades, W., Prucka, S.K. & Hjalt, T. Fate maps of neural crest and mesoderm in the mammalian eye. Investigative ophthalmology & visual science 46 4200 4208 (2005). 114. Kalluri, R. & Neilson, E.G. Epithelial mesenchymal tran sition and its implications for fibrosis. The Journal of clinical investigation 112 1776 1784 (2003). 115. Neilson, E.G. The Jeremiah Metzger lecture. The origin of fibroblasts and the terminality of epithelial differentiation. Transactions of the America n Clinical and Climatological Association 121 240 250; discussion 250 241 (2010). 116. Dandachi, N. et al. Co expression of tenascin C and vimentin in human breast cancer cells indicates phenotypic transdifferentiation during tumour progression: correlat ion with histopathological parameters, hormone receptors, and oncoproteins. J Pathol 193 181 189 (2001).

PAGE 138

138 117. Sauer, B. & Henderson, N. Site specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proceedings of the Natio nal Academy of Sciences of the United States of America 85 5166 5170 (1988). 118. Soriano, P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nature genetics 21 70 71 (1999). 119. Joo, J.H., Kim, Y.H., Dunn, N.W. & Sugrue, S.P. Disruptio n of mouse corneal epithelial differentiation by conditional inactivation of pnn. Investigative ophthalmology & visual science 51 1927 1934 (2010). 120. Gore Hyer, E. et al. TGF beta and CTGF have overlapping and distinct fibrogenic effects on human renal cells. Am J Physiol Renal Physiol 283 F707 716 (2002). 121. Liu, B.C., Zhang, J.D., Zhang, X.L., Wu, G.Q. & Li, M.X. Role of connective tissue growth factor (CTGF) module 4 in regulating epithelial mesenchymal transition (EMT) in HK 2 cells. Clinica chimica acta; international journal of clinical chemistry 373 144 150 (2006). 122. Bejjani, R.A. et al. Electrically assisted ocular gene therapy. Surv Ophthalmol 52 196 208 ( 2007).

PAGE 139

139 BIOGRAPHICAL SKETCH Daniel James Gibson was born in Sarasota, Florida in 1978. In 2001, he was honorably discharged from the United States Air Force after 4 years of duty as an Imagery Analyst. He attended Edison Community College shor tly after his military service, and in May of 2003, he was awarded an a ssociate of a rts degree in general s tudies. He matriculated to the University of Florida immediately after leaving Edison and earned a b achelor of s cience degree in m echanical e ngineeri ng in May of 2005. Daniel remained at the University and entered a m aster of s cience program in the Department of Molecular Genetics and Microbiology. In May of 200 7, Daniel was awarded a m aster of s cience degree in m edical s ciences for his research in electromotive drug delivery. Currently, he is continuing to further broaden the scope of his knowledge in the field of biomedical sciences where he plans to bring all of his diverse knowledge to bear on problems of human health. His personal interests incl ude his growing family, philosophy, Rachmaninoff, and commercializing technology.