Glutaredoxin Regulation of Pro-Inflammatory Responses In

Home , GLRX, Glutaredoxin

GLUTAREDOXIN REGULATION OF PRO-INFLAMMATORY RESPONSES IN

A MODEL OF DIABETIC RETINOPATHY

MELISSA DEANN SHELTON

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Advisor: Dr. John J. Mieyal, Ph.D.

Department of Pharmacology

CASE WESTERN RESERVE UNIVERSITY

January, 2009

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Melissa D. Shelton

candidate for the Ph.D. degree *.

(signed) Dr. Anthony J. Berdis, Ph.D.

(chair of the committee)

Dr. John J. Mieyal, Ph.D.

Dr. Michael E. Maguire, Ph.D.

Dr. Timothy S. Kern, Ph.D.

______

(date) August 4, 2008

*We also certify that written approval has been obtained for any proprietary material contained therein.

Dedication

This thesis is dedicated to Mom, Dad, Lisa, Ethan, Sadie, Ely, and Shayla for the

personal sacrifices and compromises they have made over the years for the sake of my career development, for being understanding through many difficult life choices, for the

patience for absorbing my frustrations with graduate school and science, and for their

unconditional love and support.

Table of Contents

Table of Contents 1

List of Tables 3

List of Figures 4

Acknowledgements 10

List of Abbreviations 13

Abstract 17

Chapter 1 Background and Introduction 19

1.1 Protein S-glutathionylation

1.2 Glutaredoxin-mediated catalysis of deglutathionylation

1.3 Diabetes

1.4 S-glutathionylation of proteins in the NFκB signaling pathway

1.5 Cytokines and glutaredoxin

1.6 Glutaredoxin in diabetes

1.7 Diabetic retinopathy

1.8 Retinal Müller glial cells and diabetic retinopathy

1.9 Introduction to thesis

Chapter 2 Glutaredoxin1 (Grx1) regulates activation of Nuclear Factor κappa-B

(NFκB) and production of intercellular adhesion molecule-1 (ICAM-1) in

retinal Müller cells in diabetic-like conditions 47

2.1 Abstract

2.2 Introduction

2.3 Results

2.4 Discussion

2.5 Materials and Methods

Chapter 3 Glutaredoxin1 regulates ICAM-1 production and Interleukin-6 secretion

via IκB Kinase (IKK) in retinal Müller cells, mediating autocrine and

paracrine responses 109

3.1 Abstract

3.2 Introduction

3.3 Results

3.4 Discussion

3.5 Materials and Methods

Chapter 4 Discussion and future directions 153

4.1 Discussion

4.2 Future directions

Appendix 202

A.1 Approaches to Quantitative Analysis of the S-glutathionylation of IKKβ.

Α.2 Attempts to measure the effects of Grx1 on cell death

Bibliography 239

List of Tables

Table 1.1 Effects of cytokine treatments on glutaredoxin and/or glutathionylation of

proteins 45

Table 3.1 Expression and induction of cytokines and vascular endothelial growth

factor (VEGF) in the conditioned medium from rMC-1 cells over-

expressing Grx1 144

List of Figures

Figure 1.1 Trends in the number of PubMed publications on Grx1, glutathionylation,

or Grx1 and glutathionylation 42

Figure 1.2 Leukostasis, and signal transduction pathways; glutathionylation of

proteins that are involved in inflammatory responses 43

Figure 1.3 The NFκB signaling pathway, and proteins that are potential sites for

regulation by glutathionylation and Grx1 44

Figure 1.4 The retinal Müller glial cell 46

Figure 2.1 Grx activity in homogenates of diabetic rat retinae 91

Figure 2.2 Total disulfide reducing capacity of rMC-1 cells cultured in normal or

high glucose medium 92

Figure 2.3 Glutathione reductase activity, thioredoxin reductase activity, and thiol

content in lysates from rMC-1 cells cultured in normal or high glucose

medium 93

Figure 2.4 Western blot analysis of Grx1 in lysate of rMC-1 cells cultured in normal

or high glucose medium 94

Figure 2.5 Grx activity in lysate from retinal Müller glial (rMC-1) cells cultured in

normal or high glucose medium 95

Figure 2.6 Western blot analysis of ICAM-1 in lysate of rMC-1 cells cultured in

normal or high glucose medium 96

Figure 2.7 Nuclear translocation of NFκB in rMC-1 cells cultured in normal or high

glucose medium 97

Figure 2.8 Western blot analysis of actin and GAPDH content found in nuclear

fractions of rMC-1 cells cultured in normal or high glucose medium 98

Figure 2.9 Grx activity in lysate of rMC-1 cells over-expressing Grx1 99

Figure 2.10 Western blot analysis of Grx1 and ICAM-1 in rMC-1 cells over-

expressing Grx1 100

Figure 2.11 ELISA of ICAM-1 in lysate from rMC-1 cells over-expressing Grx1 101

Figure 2.12 Nuclear translocation of the p50 subunit of NFκB in rMC-1 cells over-

expressing Grx1 102

Figure 2.13 Nuclear translocation of the p65 subunit of NFκB in rMC-1 cells over-

expressing Grx1 103

Figure 2.14 NFκB-promoter luciferase assay of lysate from rMC-1 cells over-

expressing Grx1 104

Figure 2.15 Effects of sn50 (inhibitor of NFκB nuclear translocation) on ICAM-1

production in rMC-1 cells over-expressing Grx1 105

Figure 2.16 Western blot analysis of Grx1 and ICAM-1 in lysate of rMC-1 cells

treated with siRNA directed towards Grx1 and cultured in high glucose

medium 106

Figure 2.17 Detection by extracellular membrane-bound ICAM-1 107

Figure 2.18 Restriction digest of NFκB and Renilla plasmids 108

Figure 3.1 Western blot analysis of ICAM-1 from lysates of rMC-1 cells cultured in

normal or high glucose medium in the absence or presence of an IKK

inhibitor 137

Figure 3.2 Western blot analysis of ICAM-1 from lysates of rMC-1 cells over-

expressing Grx1 and cultured in normal glucose medium in the absence or

presence of an IKK inhibitor 138

Figure 3.3 Nuclear translocation of the p50 and p65 subunits of NFκB in lysates of

rMC-1 cells over-expressing Grx1 and cultured in normal glucose medium

in the absence or presence of an IKK inhibitor 139

Figure 3.4 Mass spectral analysis of the S-glutathionylated form of IKKβ from

rMC-1 cells 140

Figure 3.5 Peptide sequences that correspond to masses identified in the mass

spectral analysis of IKKβ 141

Figure 3.6 Western blot analysis of Grx1 and ICAM-1 in lysates of rat TRiBRB

endothelial cells cultured in the conditioned medium of rat rMC-1 cells

over-expressing Grx1 142

Figure 3.7 Western blot analysis of Grx1 and ICAM-1 in lysates of rMC-1 cells

cultured in the conditioned medium of rMC-1 cells over-expressing Grx1

143

Figure 3.8 IL-6 ELISA of lysates from rMC-1 cells over-expressing Grx1 in normal

glucose medium or cultured in high glucose medium 145

Figure 3.9 ELISA analysis of the effects of centrifugal concentration of cell culture

medium on the recovery of recombinant IL-6 146

Figure 4.1 Proteins that are involved in insulin secretion and signaling and implicated

for regulation by glutathionylation 194

Figure 4.2 Changes in protein glutathionylation and glutaredoxin in inflammatory

diseases 195

Figure 4.3 Method of processing recombinant IKKβ to determine the Grx1 activity

attributed to glutathionylation 196

Figure 4.4 Method of processing IKKβ in cell lysates to determine the Grx1 activity

attributed to glutathionylation 197

Figure A1 Western blot analysis of the effects of sulfhydryl modifying agents on the

immunoprecipitation of IKK 203

Figure A2-1 Optimization of the immunoprecipitation of IKK; Part I 207

Figure A2-2 Optimization of the immunoprecipitation of IKK; Part II 212

Figure A3 Slot blot comparison of IKKβ isolated from lysates of rMC-1 cells and

concentrated via multiple different techniques 218

Figure A4-1 Western blot analysis of anti-GSH in lysates of rMC-1 cells; Part I 228

Figure A4-2 Western blot analysis of anti-GSH in lysates of rMC-1 cells; Part II 229

Acknowledgements

I would like to extend sincere gratitude to my mentor and friend, John Mieyal, Ph.D., for

his meticulous scientific thinking and rigorous and challenging training. He is truly a

respectable and respectful leader who maintains high moral and ethical standards and beliefs above all else. John is relentless in his love for the pursuit of science and moreover in his passion and enthusiasm for life. I’ve been blessed with his compassion, great conversations, and fun times over the years. Thank you for believing and trusting in me.

A special thanks to my collaborating advisor, Timothy Kern, Ph.D., for the patience, forgiveness, and support to make this a successful body of work and an enjoyable collaboration. Your straight-forward advice and opinions on science and my career search have been invaluable.

I would also like to extend my gratitude to Michael Maguire, Ph.D. for many good ideas and for helping to guide me into a productive and critically thinking scientist.

I appreciate Ruth Keri, Ph.D. and David Davis, Ph.D. (NIH) for being good scientific counsel. I admire you as scientists and just as much for your fun-loving personalities.

Thank you for taking time to offer guidance, support, and an experienced yet youthful outside perspective on experimental and/or career issues.

I would like to express a very special thanks to a number of colleagues for being great

scientific aids, incredible people and amazing friends. Krisztina Papp-Wallace Ph.D. has

been one of my closest friends since my start at Case and has been invaluable to the very

end. Sue English Ospina and Jared Niedenthal have provided critical support and

friendship over many years. Payal Gandhi, Ph.D., Rick Gibson, Marjorie Montanez-

Wiscovich, and Yee-Hsee Hsieh Ph.D. have been invaluable friends, comics, and

personal therapists. Thanks to all these wonderful people for sharing their lives, experiences, homes, and families, and for being my Cleveland “framily.” I don’t know if

I could have made it without them. I am leaving Case with a degree but moreover with

the valuable gift of many life-long friendships.

Many thanks to Esther Chee-Johnson for being the foundation of my faith in taking

chances on strangers, for having the patience to be my roommate, for providing a non-

scientific yet very down to earth and honest perspective on all aspects of life and for the

unconditional reliance and love that can only come from such a warm, wonderful person

and friend.

Thank you, Heather Goffinet, M.D. and Erin Pilcher-Owen, for over 12 unforgettable

years together through the laughs and tears, bonding, breakups, moves, and the internal

and external struggles of finding ourselves and choosing the appropriate life path. All of

which seemed to be resolved readily after some time spent at our local Tumbleweed.

Thank you for all the support from that first day of my scary new independent life on

Bellfield, throughout college and graduate school, and through many years to come.

Your friendships changed my life.

I am incredibly grateful for the most recent blessing in my life, Ben. I never thought it could be so exciting to see someone day after day after day. I am happy to have met you and happier each day I spend with you. Thank you for being such a wonderful, loving, and considerate person. I am fortunate.

List of Abbreviations

ACN Acetonitrile

Ad-Empty Adenovirus containing empty vector

Ad-Grx1 Adenovirus containing Grx1 vector

AGE Advanced glycation end product

BSA Bovine serum albumin

BSA-SSG Bovine serum albumin(S-carboxymethyl)-NH2-CO-CH2-CH2- CH2-S-SG

BSO L-buthionine sulfoximine

CFAR Center for Aids Research

COPD Chronic Obstructive Pulmonary Disease

DMF Dimethyl fumarate

DTNB 5, 5’-Dithiobis-(2-Nitrobenzoic acid)

DTT Dithiolthreitol

ECGS Endothelial cell growth supplement

ECL Enhanced chemiluminescence

ELISA Enzyme-linked ImmunoSorbent Assay

FITC Fluorescein Isothiocyanate

FMAP Fluorokine MultiAnalyte Profiling

GAPDH Glyceraldehyde 3-phosphate dehydrogenase

GSH Glutathione

GRase Glutathione reductase

Grx Glutaredoxin

GSSG Glutathione disulfide

HCl Hydrogen chloride

HBSS Hanks buffered salt solution

HEDS Hydroxyethyldisulfide

HEK Human embryonic kidney (cells)

HMGB1 High mobility group protein B1

HRP Horse radish peroxidase

IAM Iodoacetamide

ICAM-1 Intercellular adhesion molecule-1

IL-6 Interleukin-6

IL-1β Ιnterleukin1-β

IP Immunoprecipitation

IκB Inhibitor of NFκB

IKK IκB Kinase

IRS-1 Insulin receptor substrate-1

LC-MS-MS Liquid chromatography-mass spectrometry-mass spectrometry

MEKK MAPK/ERK kinase kinase

MOI Multiplicity of infection

MS Mass spectrometry mw Molecular weight

NADPH Nicotinamide adenine dinucleotide phosphate

NIK NFκB inducing kinase

NEM N-ethylmaleimide

NFκB Nuclear factor κappa B

NFκB-luc Nuclear factor κB luciferase plasmid

NO Nitric oxide

OxLDL Oxidized low density lipoprotein

PBS Phosphate buffered saline

PI3K Phosphoinositide-3 kinase

PKC Protein kinase C

PMA Phorbol 12-myristate 13-acetate

PMSF Phenylmethylsulfonyl fluoride

PP2A Protein phosphatase 2A

Protein-SSG Protein mixed disulfide glutathione (aka glutathionylated protein)

PTEN Phosphatase and tensin homolog

PVDF Polyvinyl difluoride

PTP-1B Protein tyrosine phosphatase-1B

RAGE Receptor for advanced glycation end products

RIPA Radio immuno precipitation assay rMC-1 Rat retinal Müller (glial) cell line

RNS Reactive nitrogen species

ROS Reactive oxygen species

RT Room temperature

SDS Sodium dodecyl sulfate

STZ Streptozotocin

TCA Trichloroacetic acid

TFA Trifluoroacetic acid

TDOR Thiol disulfide oxidoreductase

TLR Toll like receptor

TNFα Τumor necrosis factor alpha

TRase Thioredoxin reductase

TR-iBRB Immortalized blood retinal barrier endothelial cells from transgenic rats

Trx Thioredoxin

UPP Ubiquitin-protease pathway

VEGF Vascular endothelial growth factor

YY1 Yin Yang 1, a nuclear protein

Glutaredoxin Regulation of Pro-inflammatory Responses in a Model of Diabetic

Retinopathy

Abstract

MELISSA DEANN SHELTON

Protein S-glutathionylation is a reversible post-translational modification that is important in redox signal transduction and cellular defense against oxidative stress. S- glutathionylation occurs when a cysteine residue forms a mixed disulfide with glutathione. Glutaredoxin 1 is an enzyme that catalyzes the specific and efficient reversal of glutathionylation (de-glutathionylation). Roles for Grx1 in diabetic retinopathy, a disease characterized by oxidative stress and inflammation, had not been explored prior to the current study. Grx activity was found to be increased in retinae from diabetic rats.

Incubation of rat retinal Müller glial cells (rMC-1) in normal glucose medium (5 mM) or

diabetic-like glucose medium (25 mM, high glucose) led to corresponding increases in

Grx content and activity. High glucose also led to increased nuclear translocation of

NFκB and production of ICAM-1 (intercellular adhesion molecule-1), a transcriptional product of NFκB and known pro-inflammatory mediator in diabetic retinopathy. To

evaluate the role of Grx1 in mediating these changes, Grx1 in rMC-1 cells was

upregulated in normal glucose medium via infection with an adenoviral Grx1 construct

(Ad-Grx1). Ad-Grx1 treatment led to increased NFκB activity and ICAM-1 production.

Treatment of rMC-1 cells in high glucose medium with siRNA targeted to Grx1

prevented the increase in Grx1 and coincidentally blocked the increase in ICAM-1. The

site of regulation was localized to the cytoplasm, and IκB kinase (IKK) is a master

cytosolic regulator of NFκB activation. Inhibition of IKK activity abrogated the increase

in ICAM-1 induced by high glucose or by Ad-Grx1. Conditioned medium from the

rMC-1 cells over-expressing Grx1 was added to fresh cultures of rMC-1 cells and

induced Grx1 and ICAM-1 (autocrine regulation). Similarly, Grx1 and ICAM-1 were

elevated in retinal endothelial cells cultured in conditioned medium from the rMC-1 cells

over-expressing Grx1 (paracrine regulation). These effects correlate with a novel finding

that secretion of IL-6 was elevated in the conditioned medium. Furthermore,

IKKβ isolated from Müller cells in normal glucose medium was found to be glutathionylated on Cys179. Hence Grx1-mediated activation of IKK via

deglutathionylation may play a critical role in pro-inflammatory responses in diabetic

complications in vivo where Grx1 is increased.

Chapter 1: Background and Introduction

1.1 Protein S-glutathionylation

Protein S-glutathionylation (Protein-SSG) is the formation of a mixed disulfide bond

between the cysteine-sulfhydryl moiety of glutathione (GSH) with a cysteine-sulfhydryl moiety of a protein. Analogous to phosphorylation, glutathionylation is a reversible post-

translational modification that regulates many proteins important in cellular signal

transduction. Also similar to the modification of a protein with a phosphate, glutathione is relatively small (a tripeptide of glutatmate-cysteine-glycine; 307 Da) and has a net

negative charge. However, unlike phosphorylation, S-glutathionylation is a redox-

dependent mechanism of regulation that occurs on active cysteine residues.

Protein S-glutathionylation is likely the predominant mechanism of reversible cysteine modification because it is promoted in an environment of relatively high concentrations of GSH (0.5-20 mM), i.e., the physiological milieu (Meister and Anderson, 1983;Liu and

Hannun, 1997). Protein cysteine-sulfhydryls also can form mixed disulfide bonds with cysteine residues or other non-physiological thiols, modifications collectively known as

S-thiolation. However, greater than 85% of all intracellular mixed disulfides are attributed to protein S-glutathionylation (Klatt and Lamas, 2000).

In addition to mixed disulfide formations, thiol groups on cysteine residues can be

oxidized to protein sulfenic (protein-SOH), sulfinic (protein-SO2H) and sulfonic (protein-

SO3H) acids upon reaction with reactive oxygen/nitrogen species (ROS/RNS) (Claiborne

et al., 1999). Sulfinates and sulfonates are irreversible, whereas sulfenic acid formation

is a reversible process. Protein-sulfenates however are typically not stable and are readily

oxidized to sulfinates and sulfonates. Protein-sulfenates can react with GSH to form

protein-SSG mixed disulfides. Thus, besides being recognized as a potential regulatory

mechanism, glutathionylation is thought to be a homeostatic protective mechanism by

preventing cysteine oxidation to irreversible forms (Barrett et al., 1999a;Caplan et al.,

2004;Davis et al., 1996;Klatt et al., 1999;Mieyal et al., 1995;Thomas et al., 1995;Zheng

et al., 1998).

Accumulation of glutathionylated proteins (protein-SSG) has been reported in different

cell types under a variety of oxidative conditions (Chai et al., 1994a;Klatt and Lamas,

2000;Rokutan et al., 1994;Schuppe et al., 1992). Glutathione is the predominant redox

buffer in the cell, and the ratio to its oxidized form (GSH/GSSG) typically dictates the

cellular redox status. Under basal conditions, this ratio is approximately 100, generating a reducing environment in the intracellular milieu. Normal physiological redox signaling of growth factors and cytokines appears to elicit intracellular oxidative signals that activate transduction pathways without necessarily changing the global redox status of the cell (Finkel, 2000;Thannickal and Fanburg, 2000). Likewise the concentration of disulfide glutathione (GSSG) in hepatocytes is not substantially altered upon exposure to the oxidative burst of neutrophils e.g., (Chai et al., 1994b). This lack of perturbation in

GSH/GSSG in redox signaling contrasts with the condition of overt oxidative stress where the cell is overwhelmed with excessive oxidants, diminishing substantially the intracellular ratio of GSH/GSSG. Oxidative stress typically is thought to lead to a 10-

20 fold transient decrease in GSH/GSSG, but the ratio also has been reported as low as 1

(Klatt and Lamas, 2000). Thus, oxidative modifications of cysteine residues that occur under normal physiological redox signaling conditions need to be distinguished from those that occur under conditions that mimic pathophysiological oxidative stress, although the progression from one condition to the other likely represents a continuum.

An oxidative stimulus is necessary for glutathionylation of some proteins (e.g. protein tyrosine phosphatase-1B (PTP-1B)) , while persisting paradoxically for other proteins

(e.g. actin) in the reducing environment of the cell (Mieyal et al., 2008). Furthermore, the effects of glutathionylation on the activities of proteins are dependent on the individual protein. For example, the transcription factor NF1 (Bandyopadhyay et al.,

1998), PTP-1B (Barrett et al., 1999a;Barrett et al., 1999b), IκB kinase (IKK) (Reynaert et al., 2006),and phosphofructokinase (Mieyal et al., 1991;Zheng et al., 1998) are inactivated by glutathionylation, whereas HIV-1 protease (Davis et al., 1997), Ras

(Adachi et al., 2004a) and microsomal glutathione S-transferase (Dafre et al., 1996) are activated by glutathionylation. Thus, S-glutathionylation represents a mechanism of reversible redox regulation of protein activity and cell signaling (Gallogly and Mieyal,

2007;Shelton et al., 2005;Mieyal et al., 2008)

The catalytic mechanism of formation of protein S-glutathionylation in cells is still largely unknown, despite the attribution to several enzymes including Grx1 itself under certain conditions (Gallogly and Mieyal, 2007). GSSG is often used as a reagent to drive glutathionylation in vitro and indeed is formed upon oxidation of GSH in cells.

However, it is unlikely that protein glutathionylation by GSSG is a generalized

mechanism of formation within cells because the redox potential of most proteins is

about 1 (Klatt and Lamas, 2000). This means that the intracellular ratios of GSH/GSSG

would have to change markedly from approximately 100:1 to 1:1 in order for there to be

a 50% conversion to the glutathionylated form of the protein (i.e. protein-SH to protein-

SSG) (Gilbert, 1995). However, catalysis of the reversal of glutathionylation (de-

glutathionylation) is well characterized by glutaredoxin (Grx1).

1.2 Glutaredoxin-mediated catalysis of deglutathionylation

Glutaredoxin (Grx1), formerly known as thioltransferase (TTase), is a 11.7 kDa protein

that efficiently and specifically catalyzes deglutathionylation of proteins, and it reverses

intracellular formation of protein-SSG (Jung and Thomas, 1996;Chrestensen et al.,

2000;Gravina and Mieyal, 1993;Mieyal et al., 1995;Yang et al., 1998;Thomas et al.,

1995;Yoshitake et al., 1994;Jao et al., 2006). There has been an explosion of literature

on studies involving observations and correlations of changes in either Grx1 or

glutathionylation. Unfortunately, most studies do not address the mechanism of

regulation and do not test the reversibility of glutathionylation by Grx1 (Fig. 1.1., pg.42).

Grx1 is in the family of enzymes known as thiol disulfide oxidoreductases (TDORs).

Cellular sulfhydryl homeostasis balance is maintained via thiol-disulfide exchanges catalyzed by TDORs. The predominant cytosolic TDORs are Grx1 and thioredoxin

(Trx). However, unlike the selective reduction of protein glutathione mixed disulfides

(protein-SSG) that occurs in a GSH-dependent fashion by Grx1, the mechanisms of the

more promiscuous catalysis by Trx of protein sulfenic acids, intermolecular disulfides,

and intramolecular disulfides are not fully understood.

1.3 Diabetes

According to the Center for Disease Control, nearly 10% of Americans have diabetes.

Contributing factors such as those in current lifestyle trends (e.g. dietary and exercise habits) indicate that the number will continue to increase rapidly. Diabetes develops when the beta cells of the pancreas fail to release sufficient insulin (Type I) or when cells stop taking up glucose in response to insulin (i.e., insulin-resistant) (Type II or adult- onset diabetes). In both cases, the corresponding chronic elevation of circulating blood glucose can lead to complications and tissue damage. Reactive oxygen species (ROS) and inflammation are key mediators in the pathophysiology of diabetes and will be discussed below.

Mechanisms of glucose damage and ROS- Cellular and tissue damage from prolonged exposure to elevated glucose concentrations occurs through multiple mechanisms. Four primary pathways that mediate glucose damage are 1) increased advanced glycation end products (AGEs), 2) polyol pathway activation, 3) protein kinase C (PKC) activation and

4) hexosamine pathway activation (Brownlee, 2001). Activation of all four pathways correlates with increased superoxide from the mitochondrial electron-transport chain.

Diabetes-induced activation of the first three of these pathways can be blocked by

inhibition of the diabetes-induced overproduction of mitochondrial superoxide

(Brownlee, 2001;Du et al., 2003a). ROS from the mitochondria seem to be a primary source of oxidative stress/signaling in diabetes, though NADPH oxidases also contribute

(Coughlan et al., 2007). In addition, high glucose and ROS, as well as oxidized low density lipoprotein (oxLDL) are implicated in stimulation of adhesion molecules and

chemokines, with subsequent macrophage accumulation, supporting the categorization of

diabetes as a disease of inflammation (Tesch, 2007). High glucose leads to increases in

AGEs which upon binding to their corresponding receptor RAGE increase leukostasis and upregulation of vascular cellular adhesion molecule-1 (VCAM-1) via activation of

the redox sensitive transcription factor nuclear factor κappa B (NFκB) (Fig. 1.2, pg. 43)

(Morigi et al., 1998).

Signal transduction pathways in inflammation- Transduction of an inflammatory

response can be initiated either by exogenous or endogenous danger signals that are

released in response to tissue damage, called damage (or danger)-associated molecular

patterns (also known as alarmins) (e.g. high mobility group protein B1(HMGB1), ATP,

ROS, oxLDL) (Foell et al., 2007;Harris and Raucci, 2006;Petrilli et al., 2007). The

binding of danger signals to receptors on the surface of cell membranes leads to

activation of the inflammasome, a multi-protein complex which serves as a molecular

scaffold for caspase-1 activation. The activated inflammasome then leads to the

production of pro-IL1β and pro-IL18 via activation of NFκB (Fig. 1.2, inset, pg.43)

(Church et al., 2008;Mariathasan and Monack, 2007). Signal transduction downstream

of toll like receptors (TLRs) and RAGE is carried out by distinct but intersecting

signaling pathways. Briefly, RAGE activates signaling via Ras whereas the PI3K-Akt

pathway mediates TLR signaling (van Beijnum et al., 2008). Notably, both cascades converge on the activation of IκB kinase (IKK), a critical step in the NFκB signaling pathway. A second danger signal can activate an inflammasome via NOD (nucleotide- binding oligomerization domain)-like receptors (NLRs). Within the known inflammasome family, the NALP3 (Nacht domain-, leucine-rich repeat-, and PYD-

containing protein 3) inflammasome is the best characterized and most broadly activated,

representing a complex of proteins including the NLR NALP3 and pro-caspase-1

(Church et al., 2008;Petrilli et al., 2007). Activation of the NALP3 inflammasome leads

to caspase-1 activation and subsequent cleavage of pro-IL-1β and pro-IL-18 to produce

mature cytokines (Church et al., 2008;Mariathasan and Monack, 2007) which are

essential for the recruitment of the circulating monocytes into the tissue.

Important roles of glutathionylation in inflammatory signaling and responses have begun

to emerge. Signal transduction networks (i.e. NFκB) that mediate inflammatory

responses contain a multitude of proteins which have activities that are potential

regulatory sites by glutathionylation. Most recently, glutathionylation on two cysteine

residues was reported to inhibit Caspase 1 activity and subsequent maturation of

cytokines in superoxide dismutase 1 deficient macrophages (Meissner et al., 2008).

These studies provide a launch pad for elucidating the role(s) of glutathionylation and the

importance of these roles in inflammation.

IKK in diabetes and inflammation- Aspirin was first reported in 1876 to have tremendous benefit for lowering blood glucose concentrations even though the anti-thrombotic effects of such high doses prevent its use as a modern day anti-diabetic treatment (Shoelson et

al., 2003). Activation of the IKK complex is at the crux of insulin resistance in diabetes.

For example, IKK can exacerbate insulin resistance by direct phosphorylation of the

insulin receptor substrate-1 (IRS-1) (Gao et al., 2002). Inhibition of the specific subunit

IKKβ (e.g., by salicylate) was found to mediate the anti-inflammatory and anti-diabetic effects of aspirin and aspirin-derivatives and can enhance insulin sensitivity in animals and humans (Shoelson et al., 2003;Arkan et al., 2005;Yuan et al., 2001). The effects of

IKKβ are cell type specific. Mice deficient in IKKβ in hepatocytes develop insulin

resistance in muscle and fat but not the liver. Conversely, transgenic mice with

constitutively active hepatocyte IKKβ develop diabetes readily (Cai et al., 2005). Mice

lacking IKKβ in the myeloid cells remain fully responsive to insulin when they are fed a

high fat diet (Arkan et al., 2005). However, mice with IKKβ conditionally knocked

down in myocytes develop obesity-induced insulin resistance similar to wild-type mice

(Rohl et al., 2004). Therefore, the elucidation of the mechanisms of regulation of IKKβ

activity may uncover new targets for therapeutic intervention in diabetes and other

inflammatory diseases.

1.4 S-glutathionylation of proteins in the NFκB signaling pathway

NFκB signaling is activated in tissues of diabetic animals including the retina (Romeo et al., 2002;Zheng et al., 2004), kidney (Schmid et al., 2006), and liver (Cai et al., 2005). A

26 review of proteins in the NFκB pathway that serve as potential sites for regulation by glutathionylation is presented below.

Cytosolic signaling upstream of IKK- A series of cytoplasmic proteins can initiate nuclear

NFκB activation through the canonical NFκB signaling pathway, and these components present a plethora of potential target sites for regulation by glutathionylation in the cytoplasm (Fig. 1.2, pg. 43). However, signals from multiple upstream mediators converge in the cytoplasm at IKK, a well established master regulator of NFκB. For example, MAPK/ERK kinase kinase (MEKK) can activate IKK, and MEKK is inhibited by glutathionylation in menadione-treated lymph node carcinoma prostrate cells (Cross and Templeton, 2004). Additionally, glutaredoxin enhances NFκB activation through

MEKK in Human Embryonic Kidney (HEK) 293 cells (Hirota et al., 2000). NFκB inducing kinase (NIK) can signal to MEKK and IKK, and over-expression of Grx1 further enhances NFκB activity induced by NIK, phorbol 12-myristate 13-acetate (PMA) or tumor necrosis factor alpha (TNFα) in HEK293 cells (Hirota et al., 2000). Akt signals to IKK, and is regulated by Grx1 in hydrogen peroxide treated cardiac H9c2 cells

(Murata et al., 2003). PTEN (phosphatase and tensin homolog) is a key component of the PI3K-Akt pathway, and is glutathionylated in ATP treated macrophages (Cruz et al.,

2007). PP2A (protein phosphatase 2A) regulates Akt activity, and is inhibited by glutathionylation in hydrogen peroxide treated colon adenocarcinoma Caco-2 cells (Rao and Clayton, 2002). Ras is known to signal to multiple NFκB signaling proteins such as

PTEN, PP2A, Akt, and MEKK. Ras is activated by glutathionylation in vascular smooth muscle cells. Furthermore, glutathionylation of Ras is decreased when Grx1 is over-

expressed in the same type of cells (Adachi et al., 2004a). Figure 1.3 (pg. 44) depicts the complexity of the interacting pathways that are upstream of IKK and the remarkable number of these proteins that could potentially be regulated by Grx1. However, it is also likely that IKK could receive mixed signals from different signaling cascades.

Therefore, it is likely that direct regulation of IKK by Grx1 would have the most impact on the activation of NFκB.

Cytosolic signaling involving IKK- IκB kinase (IKK) is a 600−900 kDa multimeric complex comprised of a core enzyme of two catalytic subunits (IKKα and IKKβ) and one or more regulatory subunits (IKKγ, also known as NEMO). Additional proteins can bind to this central core, and at least 65 individual proteins have been reported to interact with IKK in different contexts (Scheidereit, 2006;Lawrence et al., 2005). While IKKβ is well established as a critical mediator, studies on the role of IKKα in inflammation are not as clear. For example, IKKα has been reported to have anti-inflammatory effects in macrophages and endothelial cells (Li et al., 2002;Li et al., 1999;Lawrence et al.,

2005;Mukherjee et al., 2007), and these effects are generally accepted for most cells

(Scheidereit, 2006). However, IKKα activity has also been reported to be essential for mediating NFκB-driven pro-inflammatory responses in mouse embryonic fibroblasts (Li et al., 2002).

Several post-translational modifications such as phosphorylation and ubiquitination are well established modulators of IKK activity. Analogously, IKKα has been reported to be glutathionylated in human pulmonary aortic endothelial cells under basal conditions, and

28 the extent of glutathionylation was induced by TNFα (Mukherjee et al., 2007).

Furthermore, the increase in glutathionylation was reported to inhibit IKKα activity and thereby block its inhibitory effects on NFκB. Nevertheless, the redox sensitive IKKβ is the best characterized subunit as a mediator of inflammatory responses via NFκB activation (Li et al., 2002;Scheidereit, 2006;Catley et al., 2006;Ghosh and Karin, 2002) and has the highest catalytic activity towards IκB (Scheidereit, 2006). A number of sulfhydryl reactive compounds such as arsenite (Kapahi et al., 2000), gold compounds

(Jeon et al., 2003), and cyclopentenone prostaglandins (Rossi et al., 2000) have been shown to inhibit NFκB activity via specific thiol modification involving Cys179 of

IKKβ. Moreover, glutathionylation of IKKβ has been found previously in a mouse alveolar type II epithelial cell line (C10) over-expressing IKKβ, and mass spectral analysis demonstrated glutathionylation of a synthetic peptide containing Cys179 of

IKKβ (Reynaert et al., 2006). Hydrogen peroxide plus TNFα or over-expression of

NADPH oxidase 1 and its coactivators increased the extent of glutathionylation in C10 cells over-expressing IKKβ, and an increase in the extent of glutathionylation was reported to correspond to a decrease in IKK activity (Reynaert et al., 2006).

Furthermore, over-expression of Grx1 led to enhanced activation of NFκB under these conditions, suggesting that Grx1 can activate NFκB via de-glutathionylation of IKKβ.

The expression of the chemokines keratinocyte-derived chemokine and macrophage inflammatory protein 2 in the medium of primary tracheal epithelial cells from Grx1 knockout mice was decreased (Reynaert et al., 2006). These data suggest that Grx1 is important in the regulation of IKKβ/NFκB-mediated inflammatory responses in lung cells and that inhibition of Grx1 may serve as a beneficial therapeutic tool. However,

whether these events occur under physiological conditions (i.e., non-over-expressing cells, physiological stimuli) has not been addressed previously. For example, whether endogenous IKKβ becomes glutathionylated within the cell in the absence of a stimulus or in response to a naturally occurring stimulus such as high glucose is the initial step critical for determining the physiological occurrence and relevance of the modification.

Cytosolic signaling downstream of IKK- Following activation of the centrally acting IKK, several proteins in the subsequent cytoplasmic signaling cascade have been reported to be glutathionylated. IKK phosphorylation of IκB (inhibitor of NFκB) promotes subsequent ubiquitination and proteolytic degradation of IκB. The loss of IκB releases the NFκB

dimer for nuclear translocation where it binds DNA and activates transcription (Fig. 1.3,

pg. 44). Reports of glutathionylation within the ubiquitin-protease pathway (UPP)

regulating ubiquitination and degradation of IκB include the ubiquitin-activating (E1) and

ubiquitin-carrier (E2) enzymes in retinal pigmented epithelial cells (Obin et al.,

1998;Jahngen-Hodge et al., 1997) and the 20S proteasome in S. cerevisiae (Demasi et al.,

2003). The 20S proteasome is a component of the 26S proteasome that degrades IκB and cleaves the cytosolic precursor protein p105 to generate the p50 subunit of NFκB

(Moorthy et al., 2006).

Nuclear Signaling- A heterodimer of the p50 and p65 subunits represent the classical and predominant active form of NFκB in most cell types. Hence, “NFκB” is typically meant to refer to the p50-p65 dimer in most studies in the literature. However, it is important to recognize that NFκB can be comprised of various combinations of five different subunits

(p50, p65 (RelA), c-Rel, p52, and Rel B). Moreover, each combination of subunits (i.e.,

p50-p50, p50-p65, RelB-p52, RelB-p50) functions differently (Baldwin, Jr., 1996;Ghosh

et al., 1998). For example, because p65 has a DNA trans-activating domain but p50 does

not, the p65-p50 heterodimer can activate transcription. A p50-p50 homodimer cannot

do so however without the help of a co-activator.

NFκB (p50-p65) contains classical nuclear translocation sequences that bind to importins

α3 and α4. The complex is then cargo for importin β for nuclear-cytoplasmic shuttling

through the nuclear pore complex (Fagerlund et al., 2005). However, the nuclear translocation sequences of p65 are typically sequestered by IκB in the cytoplasm. Hence,

NFκB translocation to the nucleus occurs subsequent to its liberation from its cytosolic restraints (IκB) and subsequent exposure of its nuclear translocation sequences.

NFκB itself is yet another potential regulatory site for glutathionylation. For example, glutathionylation of recombinant p50 protein diminishes its ability to bind DNA in vitro

(Pineda-Molina et al., 2001;Qanungo et al., 2007). Moreover, exogenous GRx1 restored

DNA binding of the p65 subunit in nuclear extracts from hypoxic pancreatic cancer cells that had been treated with N-acetylcysteine (Qanungo et al., 2007). Similar effects on

DNA binding were observed for hypoxic pancreatic cancer cells treated with N- acetylcysteine and decreased in Grx1 content by shRNA-knockdown. These findings suggest that NFκB was inhibited via glutathionylation of p50, p65, or a transcriptional co-activator/repressor (Qanungo et al., 2007).

Summary of potential regulation of NFκB signaling by reversible glutathionylation- S-

glutathionylation has been reported to inhibit at least 13 proteins involved in signaling

events leading to activation of NFκB, including NFκB itself (Fig. 1.3, pg. 44). The

complexity and cross-talk of signaling pathways provide multiple check points that

support proper cell function. However, these same characteristics challenge the

elucidation of specific regulatory mechanisms. To add to the complexity, the control

points may differ among cell types and cellular conditions. Nevertheless,

glutathionylation of the NFκB subunits would be the most straightforward mechanism of

regulation. Furthermore, IKK is the most logical target of regulation by glutathionylation

in the cytoplasm because it is a major convergent point of signaling to NFκB activation.

In addition, the IKKβ subunit specifically has a redox sensitive sulfhydryl group

(Cys179) in the active site of the enzyme that mediates pro-inflammatory responses of

NFκB. Furthermore, glutathionylation of IKKβ on Cys179 in airway epithelial cells has

been associated with the inhibition of NFκB activity in IKKβ over-expressing cells, and

chemokine secretion was decreased substantially in primary tracheal epithelial cells from

Grx1 knockout mice.

Experimental approach to identifying key proteins regulated by glutathionylation in the

NFκB signaling pathway- For such a complicated network of signaling proteins, utilizing

inhibitors to block specific components is often helpful in identifying important regulatory sites. An example of such an inhibitor is sn50. Sn50 inhibits nuclear

translocation of NFκB by competing for binding sites on importin proteins. Hence it and

others are useful tools for distinguishing regulatory events of NFκB within the cytosol

from those in the nucleus.

1.6 Cytokines and glutaredoxin

TNFα and IL-1β are the prototypic pro-inflammatory cytokines. The anti-inflammatory

cytokines (e.g., IL-10, IL-13) inhibit pro-inflammatory cytokine production by

macrophages. Chemokines are chemotactic cytokines, responsible for the recruitment of

circulating leukocytes into sites of injury in tissue, including RANTES, MCP-1, MIP-1α,

and MIP-1β, and IL-8.

ROS are known to induce cytokine production and have been reported to induce Grx1

expression in human coronary artery smooth muscle cells (Okuda et al., 2001). Recently, several studies have reported effects of purified cytokines on Grx1 mRNA, protein content, and protein activity, but the induction of Grx1 in cells by an endogenous stimulus (i.e., a cytokine secreted from a cell) has not been documented (Table 1.1, pg.

45). For example, in murine monocytic leukemia-derived M1 cells pure IL-6 has been reported to substantially elevate Grx1 mRNA, but Grx1 protein content was only slightly increased (Takashima et al., 1999). Treatment with pure IL-13 led to increases in Grx1 activity whereas IL-4 led to decreases in Grx1 activity in mouse airway epithelial cells

(Reynaert et al., 2007). IFNγ led to an increase in Grx activity, protein and mRNA with a corresponding decrease in protein glutathionylation in mouse airway epithelial cells

(Reynaert et al., 2007). As would be expected, the effects of cytokines on glutaredoxin

seem to be cytokine-specific because TGFβ1 led to a decrease in Grx activity and mRNA

with a corresponding increase in total protein glutathionylation in mouse airway

epithelium (Reynaert et al., 2007) and to a decrease in Grx1 protein in A549 cells

(Peltoniemi et al., 2004). In addition, in bovine aortic endothelial cells TNFα leads to an increase in activity of glutaredoxin and a corresponding decrease in glutathionylation of pro-Caspase 3 without a change in Grx1 protein content (Pan and Berk, 2007). In contrast, no change in Grx1 activity in mouse airway epithelium and no change in Grx1 protein in A549 cells were found in response to TNFα (Reynaert et al., 2007;Peltoniemi et al., 2004). Increased glutathionylation of ICAM-1 and IKKα in TNFα− treated human pulmonary aortic endothelial cells suggests a decrease in glutaredoxin activity in that context, adding to the complexity (Mukherjee et al., 2007). Therefore, TNFα seems to induce differential changes in glutaredoxin and glutathionylation likely contingent on species, cell type, and treatment conditions.

The mechanism of regulation of Grx1 is currently unknown. As described above, studies using purified cytokines show that they can induce cellular Grx1, but whether this induction occurs in response to cytokines secreted endogenously has not been demonstrated previously. If the cytokines secreted by the cells (autocrine response) or by neighboring cells (paracrine response) are not of sufficient concentration to evoke a response, then this induction would not have physiological relevance. In turn, Grx1 has been shown to regulate the cellular secretion of two cytokines (keratinocyte-derived chemokine and macrophage inflammatory protein 2) in primary airway epithelial cells

(Reynaert et al., 2006).

1.5 Glutaredoxin in diabetes

Grx1 is a ubiquitously expressed protein, and its expression is increased in inflammatory diseases associated with many diverse types of cells and tissues. For example, Grx1 is increased in the brain of post-mortem patients with Alzheimer’s disease, in pancreatic and colon cancer cells, and in the lung of allergic airway disease (Fig. 4.2, pg. 195).

Therefore, Grx1 is expected to have important roles in diverse tissues afflicted with diabetes. However, few studies have analyzed changes in Grx1 in animal or cell culture models or in patients with diabetes.

Di Simplicio et al. (Di Simplicio et al., 1995) was the first study to report changes in

Grx1 in diabetes. The authors found a decrease in Grx1 in platelets of diabetic patients, whereas an increase in Grx1 has been reported in the heart cells of diabetic rats (Li et al.,

2005). The apparently divergent changes in different tissues suggest an intriguing area of study, and analysis of Grx1 in diabetic retinopathy specially has been unexplored previously. Further analysis of Grx1 in various tissue and cell types from diabetic models is needed to elucidate the cell specificity associated with both the mechanisms of regulation of Grx1 itself and its effects on various cellular functions.

1.7 Diabetic retinopathy

Diabetes is a vascular disease that affects most organs including the heart, kidney, and eyes. Lesions indicative of complications of diabetes in the retina include acellular capillaries, microaneurysms, hemorrhages, and pericyte ghosts (Kern and Engerman,

2001). These lesions can ultimately lead to blindness. The effects of diabetes on the retina are consistent with the overriding characteristics of the disease emanating from uncontrolled glucose concentrations, namely oxidative stress and inflammation. For example, decreased concentrations of glutathione (Kowluru et al., 2001), increased superoxide production from the mitochondria (Du et al., 2003b), and increased nitric oxide (Du et al., 2002;Du et al., 2004) have all been associated with diabetic retinopathy, indicating perturbations in redox homeostasis.

An increase in NFκB activity has been implicated in the complications of diabetic retinopathy (Romeo et al., 2002;Zheng et al., 2004;Zheng et al., 2007). Consistent with their anti-diabetic effects discussed above, the inhibition of IKKβ by aspirin and salicylate-derivatives also has beneficial effects in the retina. These drugs have been shown to inhibit endothelial and neuronal cell death, occurrence of acellular capillaries, retinal hemorrhages, and the upregulation of VCAM, ICAM-1, iNOS, and COX-2 in diabetic animals (Zheng et al., 2007;Kern and Engerman, 2001).

1.8 Retinal Müller glial cells and diabetic retinopathy

Müller cells are the primary glial cells in the retina, spanning up to 70% of the depth of the retina. These cells are considered to be essential for metabolic and structural support

to other cells within the retina. The radial spanning of the Müller cells allows it to

surround capillaries in the vascular portion of the retina and neuronal cells throughout the retina, such as amacrine, bipolar, horizontal, ganglion, and photoreceptor cells (Fig 1.4,

pg. 46) (Newman and Reichenbach, 1996).

Changes in the expression of proteins such as glial fibrillary acidic protein in Müller cells

are observed in the early stages of diabetic retinopathy, and this response to injury or

inflammation is called reactive gliosis or glial reactivity (Rungger-Brandle et al., 2000).

Reactive glial cells are have been found in retinal Müller cells of diabetic mice and rats

and in cultured rat Müller cells (rMC-1) (Gerhardinger et al., 2005;Rungger-Brandle et

al., 2000;Lu et al., 1999;Feit-Leichman et al., 2005). In addition, high glucose has been

reported to increase nuclear translocation of GAPDH, caspase activation, and Annexin V

staining in rMC-1 cells (Mohr et al., 1999;Mohr et al., 2002;Vincent and Mohr, 2007).

The role of Müller cells in the inflammatory processes in diabetic retinopathy has not

been fully elucidated, but changes in cytokine expression and secretion certainly suggest

that these cells are likely to be involved. For example, TNFα, IL-1β, and MCP-1 have

been reported to be upregulated within the Müller cell (Walker and Steinle,

2007;Nakazawa et al., 2007). Interesting, TNFα and IL-1β protein expression was

induced within the cell in response to high glucose (Walker and Steinle, 2007).

Furthermore, IL-8, MCP-1, and TGFβ-2 have been reported to be secreted from Müller

cells under various conditions (Malgorzata Goczalik et al., 2005;Nakazawa et al.,

2007;Walker and Steinle, 2007;Eichler et al., 2004). Studies utilizing a neutralizing

37 antibody suggest that IL-1β is released from Müller cells, and the secretion seems to increase in response to high glucose (Vincent and Mohr, 2007). These data support a pro-inflammatory role of Müller cells in diabetic retinopathy and provide a basis for further studies to evaluate cytokine secretion and regulation in retinal cells in response to high glucose medium.

1.9 Introduction to thesis

Several recent studies have shed light on the roles of Grx1 in cellular regulation and signal transduction (e.g., (Adachi et al., 2004a;Wang et al., 2003b)). Many more have reported observations or correlations without any mechanistic insight (Fig. 1.1, pg. 42).

For glutathionylation to serve as a regulatory mechanism, it must fulfill several criteria

(Shelton et al., 2005): (1) it must change the function of the modified protein, (2) it must occur in intact cells in response to a physiological stimulus and elicit a physiological response, (3) it must occur under physiological conditions, (4) it must be rapidly and efficiently formed, and (5) it must be rapidly and efficiently reversed (catalyzed by

Grx1). Despite the explosion of literature on Grx1 or glutathionylation, these criteria are typically not addressed. For example, few studies have documented changes in glutathionylation in response to a physiological stimulus under physiological conditions, and even fewer have tested its reversibility by Grx1. Thus, further analysis of the Grx1 redox system in cells or animals is needed to advance our understanding of the overall physiological significance.

Disruptions in redox balance and inflammatory responses routinely surface as important

contributors in disease. Diabetic retinopathy is a disease characterized by perturbations

in the cellular redox status and inflammatory responses. Müller cells are glial cells

specific to the retina and provide critical support to surrounding cells. Perturbations in

these cells would likely have a widespread impact on surrounding cells. Furthermore, the

redox-regulated signal transduction pathway leading to activation of Nuclear Factor

κappa B (NFκB) is known to mediate inflammatory cytokines such as interleukin 6 (IL-

6) and adhesion molecules such as intercellular adhesion molecule 1 (ICAM-1).

Potential sites for regulation in the NFκB pathway by glutathionylation are abundant

(Fig. 1.3, pg. 44) but largely unexplored in terms of reversible regulation by Grx1 and

physiological impact. This thesis addresses the role of Grx1 in a model of diabetic

retinopathy.

We hypothesized that changes in Grx1 would lead to changes in pro-inflammatory

markers via regulation of mediators in the NFκB pathway. We tested the hypothesis in a

model of diabetic retinopathy, namely retinal Müller glial cells cultured in high glucose

medium. The aims of this study were to (1) determine whether Grx1 protein and activity

are changed in Müller cells cultured in high glucose medium and in retinae from diabetic

animals compared to non-diabetic animals, (2) elucidate the impact of the changes in

Grx1 on pro-inflammatory markers in diabetic-like conditions compared to non-diabetic- like conditions, (3) discover the molecular mechanism(s) by which changes in Grx1 lead to physiological changes in cells, and (4) learn whether changes in Grx1 in one cell

(Müller cell) influence neighboring cells such as the endothelial cell.

In this study, we have discovered that the deglutathionylase activity of Grx1 is increased in retinae from diabetic rats and in a Müller cell line (rMC-1 cells) in response to high glucose conditions that simulate diabetic retinopathy (i.e., 25 mM glucose) (Chapter 2).

This increase in activity correlates to an increase in Grx1 protein content and to increased activation of NFκB and production of ICAM-1 in rMC-1 cells cultured in high glucose medium (Chapter 2). Grx1 was over-expressed in rMC-1 cells via adenoviral infections to an extent which mimics the level of expression induced in high glucose medium. This provides a model in which the effects of Grx1 are isolated from other changes that occur in response to high glucose. Grx1 over-expression leads to increased Grx1 activity, activation of NFκB, and production of ICAM-1 in rMC-1 cells, similar to those observed with high glucose medium (Chapter 2). An exciting and unprecedented finding is that rMC-1 cells over-expressing Grx1 secrete more IL-6 into the culture medium than control cells (Chapter 3). This medium (conditioned medium) induces the expression of GRx1 and ICAM-1 proteins in freshly cultured rMC-1 cells (autocrine regulation) (Chapter 3).

Similarly, these same proteins are induced in endothelial cells in response to the conditioned medium from rMC-1 cells over-expressing Grx1 (paracrine regulation)

(Chapter 3). These findings suggest that upregulation of Grx1 leads to ICAM-1 production and IL-6 secretion via activation of the NFκB pathway. Furthermore, it was found that the site of regulation of the NFκB pathway by Grx1 seems to be primarily located in cytosolic compartment of the cell (Chapter 2). Specifically, IKK is an important cytosolic regulator of NFκB activation, and its glutathionylation in cells in the absence of a stimulus or over-expression of proteins is an important and novel finding

(Chapter 3). Overall, these data suggest that upregulation of Grx1 in diabetic-like conditions leads to deglutathionylation and activation of IKKβ-SSG which in turn dissociates the IkB-NFκB complex resulting in the activation of NFκB. This NFκB activation subsequently leads to increased production of the adhesion molecule ICAM-1 and secretion of the cytokine IL-6. It appears as though IL-6 mediates the autocrine and paracrine pro-inflammatory responses in retinal cells which is indicated by altered cell adhesion properties (i.e. increased protein expression of ICAM-1). Overall, an increase in Grx1 seems to have pro-inflammatory effects on retinal cells and likely contributes to the overt hyper-inflammatory responses in diabetic retinopathy. Inhibition of Grx1 leads to a decrease in the inflammatory marker, ICAM-1. Thus, targeting Grx1 therapeutically may have beneficial effects in vivo in diabetic retinopathy and other inflammatory diseases.

Figure 1.1. Number of yearly publications involving studies on glutaredoxin, glutathionylation, or glutaredoxin and glutathionylation as measured in PubMed entries.

Figure 1.2. Leukostasis and signal transduction pathways; glutathionylation of proteins that are involved in inflammatory responses. Illustrated here are leukostasis, diapedesis, and production of pro-inflammatory cytokines in response to HMGB1 that is secreted from resident macrophages. Inset: Ligands (HMGB1, oxLDL) binding to the

RAGE and TLR2/TLR4 receptors (highlighted in small black boxes) induce an intracellular signaling cascade. (Modified from Shelton and Mieyal 2008)

Figure 1.3. The NFκB signaling pathway and proteins that are potential sites for regulation by glutathionylation and Grx1. Cytosolic signaling cascades converge on

IKK which phosphorylates the NFκB inhibitory protein, IκB. IκB is subsequently degraded, exposing the nuclear translocation sequences of NFκB. Proteins that have been implicated as potentially regulated by glutathionylation and Grx1 are numbered 1-

13. The citations for each are as follows: (1) (Adachi et al., 2004a), (2) (Cross and

Templeton, 2004), (3) (Hirota et al., 2000), (4) (Murata et al., 2003), (5) (Cruz et al.,

2007), (6) (Rao and Clayton, 2002), (7) (Reynaert et al., 2006), (8) (Pineda-Molina et al.,

2001), (9) (Qanungo et al., 2007), (10) (Jahngen-Hodge et al., 1997), (11) (Jahngen-

Hodge et al., 1997), (12) (Demasi et al., 2003), (13) (Mukherjee et al., 2007).

Table 1.1: The effects of cytokine treatments on glutaredoxin and/or glutathionylation of proteins. *TGFβ1 (2 ng/ml) had no affect on Grx1 at 24 and 48 hr (Peltoniemi et al.,

2004). Abbreviations: CHX, cycloheximide; PSSG, glutathionylated proteins. (Modified from Shelton and Mieyal 2008).

Figure 1.4. The retinal Müller glial cell. A, an amacrine cell; B, a bipolar cell; C, a cone photoreceptor; CAP, a retinal capillary; EF, the end foot of the Müller cell; G, a ganglion cell; H, a horizontal cell; M, a Müller cell; MV, microvilli; R, a rod photoreceptor. Reprinted from Trends in Neuroscience, 19, Newman E and Reichenbach

A, The Müller cell: a functional element of the retina, 307-312, Copyright (1996), with permission from Elsevier.

Chapter 2: Grx1 regulated activation of Nuclear Factor-κappa B (NFκB) and

production of adhesion molecule-1 (ICAM-1) in retinal Müller (rMC-1) cells in

diabetic-like conditions

Part of the work presented in Chapter 2 was published in the Journal of Biological

Chemistry, 282, 12467–12474, April 27, 2007

2.1 Abstract

Reversible S-glutathionylation of proteins (protein-SSG) is a critical regulatory mechanism in redox signaling and cellular defense against oxidative stress. This post-

translational modification alters protein function, and its reversal (de-glutathionylation) is

catalyzed specifically and efficiently by glutaredoxin (Grx1, thioltransferase), a thiol-

disulfide oxidoreductase (TDOR). A change in Grx1 was hypothesized to be important

in the pathogenesis of diabetic retinopathy, a condition characterized by oxidative stress.

Grx activity was found to be increased in retinal homogenates from streptozotocin-

diabetic rats. Also, incubation of rat retinal Müller cells (rMC-1) in diabetic-like glucose

(25 mM) medium instead of normal glucose (5 mM) led to upregulation of Grx1 but not

thioredoxin, the other TDOR enzyme. Under the same conditions, NFκB translocated to

the nucleus, and expression of ICAM-1 (intercellular adhesion molecule-1), a

transcriptional product of NFκB, increased. Pro-inflammatory ICAM-1 is increased in

diabetic retinae, and it is implicated in pathogenesis of retinopathy. To evaluate the role

of Grx1 in mediating these changes, intracellular Grx1 content and activity were

47 upregulated in rMC-1 cells in normal glucose via infection with an adenoviral Grx1 construct (Ad-Grx1). The infected rMC-1 cells exhibited adenovirus concentration- dependent increases in Grx1, and corresponding increases in NFκB nuclear translocation,

NFκB luciferase gene reporter activity, and ICAM-1 expression. Blocking the increase in Grx1 via siRNA in rMC-1 cells in high glucose (25 mM) prevented the increase in

ICAM-1 expression. These data suggest that redox regulation by glutaredoxin in retinal glial cells is perturbed by hyperglycemia, leading to NFκB activation and a pro- inflammatory response. Thus, inhibition of Grx1 may represent a novel therapeutic intervention for diabetic retinopathy.

2.2 Introduction

Reactive oxygen species (ROS) are redox second messengers essential to physiological processes, but in excess they can disrupt normal redox signaling, damage cell components, and irreversibly oxidize cellular proteins. Thus, oxidative signals promote protein modifications on redox sensitive cysteine sulfhydryls in a range of oxidative states from redox-activated signal transduction to oxidative stress-induced molecular damage (Shelton et al., 2005). Reversible post-translational modifications such as protein-sulfenic acids (protein-SOH), S-nitrosylated proteins (protein-SNO), and S- glutathionylated proteins (protein-SSG) are thought to protect against irreversible oxidation (Shelton et al., 2005;Mieyal et al., 1995), and S-glutathionylation is likely the predominant physiological sulfhydryl modification due to the abundance of cellular glutathione (Meister and Anderson, 1983) and the ready conversion of cys-SNO and cys-

SOH moieties to cys-SSG (Thomas and Mallis, 2003). S-glutathionylation results in protein-specific functional changes (activation or de-activation), important in regulation of signaling mediators involved in cellular processes. For example, S-glutathionylation activates Ras and leads to downstream phosphorylation of Akt and p38, increased protein synthesis, and cell proliferation (Adachi et al., 2004a). In contrast glutathionylation inactivates PTP-1B phosphatase which consequently amplifies the effects of kinase activation in associated signaling pathways (Barrett et al., 1999b;Barrett et al., 1999a).

Reversal of S-glutathionylation, i.e., deglutathionylation, is catalyzed specifically and efficiently by the thiol-disulfide oxidoreductase enzyme glutaredoxin (Grx1) (Gravina and Mieyal, 1993;Yang et al., 1998;Apetoh et al., 2007;Chrestensen et al., 2000). This characteristic has led to the use of Grx1 as a diagnostic tool, whereby reversal of protein functional changes by Grx1, both in biochemical analyses and in cell culture studies, is interpreted as regulation by S-glutathionylation (Hamnell-Pamment et al.,

2005;Humphries et al., 2002;Nulton-Persson et al., 2003;Wang et al., 2003b).

Accordingly, alterations in the activity of Grx1 in cells in the context of diseases that involve oxidative stress would be expected to perturb sulfhydryl homeostasis and redox signaling. With this in mind, it seemed like Grx1 might be involved in the redox regulation in Diabetes, where the hyperglycemia produces a state of chronic oxidative stress leading to a variety of complications such as retinopathy.

Retinas from diabetic animals show numerous abnormalities consistent with oxidative stress (Du et al., 2002;Kowluru et al., 2001). Much evidence supports the interpretation of the retinopathy as a progressive inflammatory disease (Joussen et al., 2004;Zheng et

49 al., 2004;Gerhardinger et al., 2005). Increased adhesion of leukocytes to the wall of retinal vessels has been linked to increased vascular permeability and capillary cell death, and each of these has been linked to increased expression of intercellular adhesion molecule-1 (ICAM-1) in retinas of diabetic animals (Joussen et al., 2004;Joussen et al.,

2004). Retinal glial (Müller) cells from diabetic rats also display increased ICAM gene expression (Gerhardinger et al., 2005), suggesting a contribution of ICAM-1 to the inflammatory response. Müller cells play an essential support role in the retina, interacting with nearly all the other retinal cells, spanning 70% of the width of the retina, acting as metabolic regulators, and storing most of the retinal glutathione content (Pow and Crook, 1995;Schutte and Werner, 1998;Newman and Reichenbach, 1996). Hence an immortalized rat retinal glial cell line (rMC-1) was chosen as the model system for the current study.

Adhesion molecules such as ICAM-1 are central to inflammatory processes such as leukostasis and their expression is regulated by NFκB. NFκB is activated in retinal glial cells, pericytes, and endothelial cells in diabetes (the current study, (Romeo et al.,

2002;Zheng et al., 2004) respectively), suggesting that oxidative signals within cells affect transcriptional activity of NFκB. In this regard, many proteins that are implicated in the pathway of regulation of NFκB activity have been reported to have their functions altered by S-glutathionylation in different contexts, as discussed in Chapter 1 of this thesis.

The current study was designed to test the hypothesis that the oxidative stress associated with high glucose alters glutaredoxin-regulated redox signaling in retinal glial cells. Here high glucose is shown to induce glutaredoxin in retinal Müller cells, with concomitant

NFκB activation, and increased ICAM-1 expression. Over-expression of glutaredoxin in these cells in normal glucose leads to analogous increases in NFκB activation, and

ICAM-1 expression. Conversely, knockdown of Grx1 in cells in high glucose prevents the induction of ICAM-1. These data suggest that redox regulation by glutaredoxin in retinal glial cells is perturbed by hyperglycemia, leading to NFκB activation and a pro-

inflammatory response.

2.3 Results

Grx is induced in diabetic rat retinae. Diabetes was induced in rats via a streptozotocin

injection by Dr. Timothy Kern’s research group. Ten weeks after the injection, retinae

from non-diabetic rats and rats diabetic were homogenized and assayed for Grx activity to assess whether the diabetic condition altered glutaredoxin activity. In fact, the Grx

activity of diabetic rat retinae was increased approximately 2.5-fold relative to control

(Fig. 2.1, pg. 91). These data reflect the collective change in Grx activity for all cells

comprising the retina, although the extent of change in any of the individual cell types is

not known (see Chapter 3). Because retinal Müller cells comprise a large portion of the

total retina and influence the vitality of neighboring cells, further studies were conducted

in this well known in vitro model, namely immortalized rat retinal (rMC-1) glial cells.

Glucose depletion of cell culture medium was carefully avoided. To characterize cell culture conditions in order for rMC-1 cells to serve as a cellular model for the diabetic retina, glucose concentrations in the medium were monitored daily for 5 days. Cells were cultured either in normal glucose medium (5 mM) or high glucose medium (25 mM).

Glucose depletion was observed when greater than 0.5 million cells were cultured in normal glucose (5 mM) medium for an incubation time of 4-5 days. Therefore, in order to avoid glucose depletion, standard procedures included daily replacements of both high and normal glucose media, and the recommended plating numbers for a 4-5 day incubation are 0.3-0.5 million cells with at least 10 ml medium/100 mm dish or 0.5 million cells with 25 ml medium/100 mm dish. Plating less than 0.25 million cells leads to unevenly dispersed cell growth (i.e., clusters) due to a lack of adequate intercellular contact.

High glucose selectively induces Grx1 in rMC-1 cells. To elucidate changes in sulfhydryl homeostasis in the retinal glial (Müller) cells in response to high glucose, the rat Müller cell line (rMC-1) was used, and the cells were cultured under conditions simulating conditions of diabetes (i.e., 25 mM glucose in the medium). Cells were assayed for total disulfide reducing capacity with two different disulfide substrates to distinguish the relative contributions of the glutaredoxin and thioredoxin (Trx) systems in the intact cells

(Biaglow et al., 2000). The disulfide reducing capacity of cells is attributable to the two cytosolic TDOR enzyme systems, i.e., Grx and Trx and their corresponding reductase systems (GSH, GRase, NADPH) and (TRase, NADPH), respectively. Reduction of

HEDS is attributable to total reducing capacity (Grx1 and Trx systems), and lipoate

reduction is attributable to the Trx system alone. Therefore changes in the capacity of the

respective systems can be distinguished. After three to five days of high glucose treatment, the activity of the Trx system (rate of lipoate reduction) of rMC-1 cells was not significantly changed (Fig. 2.2, pg. 92), and Trx protein was unchanged in Western blot analyzed by David Starke in our lab (data not shown). However, the total reducing capacity (rate of HEDS reduction) was increased by nearly 2-fold (Fig. 2.2, pg. 92), and together, these data indicate that the change in total disulfide reducing capacity is due to a selective increase in activity of the Grx system. Unlike Grx, high glucose led to a decrease of approximately 40-60% in TRase activity, GRase activity, and total cellular thiol content in rMC-1 cells (Fig. 2.3, pg. 93). These results suggest that Grx is selectively induced by high glucose in rMC-1 cells. Furthermore, the activities of these enzymes were reflected in the total sulfhydryl reducing capacity assay of intact cells where the Trx system did not show a net change, and the remaining activity attributed to the Grx system was increased, suggesting that the selective induction of Grx has the greatest impact.

To test whether the increase in Grx1 activity led to corresponding increases in protein content, lysates were analyzed via Western blot for anti-Grx1. Grx1 protein expression was increased in lysates from high glucose treated rMC-1 cells by more than 2-fold according to Western blot analysis (Fig. 2.4, pg. 94). In separate experiments, it was confirmed that Grx1 content did not change when cells were incubated in 25 mM L- glucose as a control for increased osmolarity.

High glucose led to increases in Grx1 protein to a similar extent as Grx activity in intact cells, suggesting the increase in activity was due to the induction of the Grx enzyme.

Studies were conducted on Grx activity in lysate of rMC-1 via two different standard

enzymatic assays. A trend of increased Grx activity was observed in pilot studies of

lysates of cells cultured in high glucose medium assayed via the standard plate reader

assay (see Methods). Grx activity in cellular lysate was then assayed via the rate of release of the GS-moiety [35S] from [35S] BSA- SSG as the soluble product [35S] GSSG.

Grx activity of cells grown in normal and high glucose medium was 0.37 nmol/min/mg

(+/-0.08) and 0.62 nmol/min/mg (+/-0.18) respectively (Fig. 2.5, pg. 95) (n=14) (p=0.07).

This standard assay was used to measure Grx activity in 293, A549, CHO, CV1, H9,

Jurkat, K562, MT-2, Q7, and U937 cell lysates previously, and the lowest Grx activity was found in the CV1 epithelial adherent cell line derived from the kidney of African green monkeys (0.3 nmol/min/mg) and a K562 non-adherent cell line isolated from a patient with Chronic Myelogenous Leukemia (0.7 nmol/min/mg) (Chrestensen et al.,

2000). The other cell lines all contained Grx activity in the range of 1.1-9.9 nmol/min/mg. Compared to these data, Grx activity in rMC-1 cells is low. Nevertheless, the observed 1.7-fold increase in Grx activity in rMC-1 cells grown in high glucose medium is similar to the increase in Grx activity indicated in the total disulfide reducing capacity assay above.

High glucose leads to upregulation of ICAM-1 in rMC-1cells. The increase in Grx in rMC-1 in high glucose agrees with the increase observed in the whole retinae from diabetic animals. ICAM-1 is a pro-inflammatory marker also increased in the retinae of

diabetic animals (Zheng et al., 2004;Joussen et al., 2002). To determine whether the increase in Grx correlated to changes in production ICAM-1 in rMC-1 cells, cells cultured in normal or high glucose medium were processed for Western blot analysis.

ICAM-1 production was increased by 3-fold in lysates of rMC-1 cells treated with high glucose for five days (Fig. 2.6, pg. 96). Soluble forms of ICAM-1 (sICAM-1) were undetectable in Western blot analysis of normal or high glucose medium of rMC-1 cells.

Unaltered medium and medium concentrated by precipitation with 20% cold TCA followed by neutralization with sodium hydroxide were both analyzed, and no protein bands were immunoreactive with anti-ICAM-1 antibody. Furthermore, ICAM-1 was not detected in conditioned medium of rMC-1 cells cultured in normal glucose medium or high glucose medium or adenoviral-Grx1 over-expressing rMC-1 cells in normal glucose according to blinded ELISA assays with a limit of detectibility of <31 pg/ml performed by the CFAR core facility. These data suggest that intracellular and/or membrane bound

ICAM-1 is the predominant form of ICAM-1 induced by high glucose in rMC-1 cells.

High glucose leads to increased nuclear translocation of NFκB (p50 and p65) in rMC-1 cells. To test whether the observed increase in ICAM-1 production in rMC-1 cells is mediated by NFκB, we measured changes in nuclear NFκB after incubation in high glucose. Protein expression of the p50 and p65 subunits of NFκB was increased in the nucleus about 2-3-fold, while cytoplasmic contents were essentially unchanged (Fig. 2.7,

A-C, and pg. 97). Ying Yang 1 protein (YY1) was used as a nuclear marker and loading control. Initially, actin was probed exclusively as a cytosolic loading control but was found to occur to a similar extent in the nuclear fractions (Fig. 2.8, A, pg. 98).

Normalization of nuclear p50 and p65 to YY1 typically yielded the same result as when

they were normalized to actin. Documentation of actin in the nucleus of rMC-1 cells

specifically has not been found in the literature, and actin was once thought to be

localized exclusively to the cytoplasm. Western blot analysis of nuclear fractions of

rMC-1 cells of an alternative cytosolic protein control such as lactate dehydrogenase

would be necessary to eliminate cytosolic contamination as the source of actin in the

nuclear fractions. However, actin is now well established for presence in the nucleus in cells and involvement in most nuclear processes such as assembly and maintenance of the nuclear envelope, nuclear transport, chromatin remodeling, activity of RNA polymerase, infrastructure of the nucleus, and intranuclear trafficking, (Jockusch et al., 2006;Ondrej

et al., 2008;Pederson and Aebi, 2005). Nevertheless, to test whether the nuclear presence

of actin in rMC-1 cells was due to cytosolic contamination, blots were immunoprobed for

GAPDH. As expected, detection of GAPDH was robust for the cytoplasm, but minimal

GAPH was found in the nuclear fractions from cells grown in normal glucose medium

(Fig. 2.8, A-B, pg.98). Interestingly, high glucose medium led to an increase in GAPDH

in the nuclear fractions of rMC-1 cells (Fig. 2.8, pg.98). These data are consistent with

previous reports of high glucose-induced nuclear translocation of GAPDH in rMC-1 cells

(Kusner et al., 2004). Unlike actin and GAPDH, YY1 was only found in the nucleus and

was therefore the nuclear marker used in these studies. The concomitant increase in

Grx1, NFκB translocation, and ICAM-1 expression in response to high glucose,

suggested that Grx1 might be directly responsible for regulating NFκB activity and

ICAM-1 expression in Müller cells. Therefore we tested this hypothesis directly.

Infection of rMC-1 cells in normal glucose with Adenovirus containing cDNA for Grx1

leads to increased Grx content and activity, and concomitant increase in ICAM-1

production. We selectively increased Grx activity in rMC-1 cells grown in normal

glucose conditions (5 mM) by over-expressing Grx1 using adenovirus containing Grx1

cDNA (Ad-Grx1). Infection of the cells with empty vector (Ad-Empty) served as

control. Ad-Grx1 increased cellular Grx activity and content in an MOI (multiplicity of

infection)-dependent fashion (Fig. 2.9, pg. 99 and 2.10 A-B, pg. 100). Grx activity

correlated well to Grx1 protein content at most MOIs, but cells infected with Ad-Grx1 at

MOI 40 showed an unexplained high amount of Grx1 protein content. Ad-Empty had no effect on either Grx activity or content (Fig. 2.9, pg. 99 and 2.10 A-B, pg. 100, respectively). Western blot analysis of lysates from the Ad-Grx1 infected rMC-1 cells in normal glucose, and not cells infected with empty vector, showed MOI-dependent increases in production of ICAM-1 (Fig. 2.10 C-D, pg. 100). Expression of ICAM-1 was shown to be elevated more than 2-fold (+/-0.3) in Grx1 over-expressing rMC-1 cells as confirmed via blinded ELISA assays performed by Center For Aids Research (CFAR)

(n=4) (Fig. 2.11, pg. 101). Basal amounts of ICAM-1 were detected at approximately

650 ng/ml and induced up to nearly 1,500 ng/ml. Due to variation, four replicates of

ICAM-1 measurements in cells over-expressing Grx1 at MOI 10 did not quite reach

statistical significance (p = 0.051). ICAM-1 was also elevated in cells over-expressing

Grx1 at MOIs of 2.5 and 160 compared to empty vector control (n=2-3). Overall, these

results indicate that Grx1 regulates ICAM-1 production in rMC-1 cells.

Over-expression of Grx1 in rMC-1 cells in normal glucose medium increases nuclear

localization and gene transcription activity of NFκB. To test whether Grx1 regulates

NFκB activity, rMC-1 cells were transfected with Ad-Grx1 or Ad-Empty, incubated in normal glucose concentrations, and analyzed for NFκB nuclear translocation and activity.

Over-expression of Grx1 increased nuclear p50 and p65 proteins by about 3 to 6-fold

(Fig. 2.12, A-B, pg. 102 and 2.13 A-B, pg. 103). As expected, Ad-Empty did not increase nuclear p50 (Fig. 2.12, A-B, pg. 102) or nuclear p65 (Fig. 2.13, A-B, pg.103).

Neither Ad-Grx1 nor Ad-Empty had a significant effect on abundance of cytoplasmic p50

(Fig. 2.12, C-D, pg. 102) or cytoplasmic p65 (Fig. 2.13, C-D, pg.103).

To confirm that increased NFκB in the nucleus corresponded to increased NFκB transcriptional activity, we assayed for NFκB luciferase (NFκB-luc) activity in rMC-1 cells after infections with Ad-Grx1 or Ad-Empty. Whereas empty vector had no significant effect, over-expression of Grx1 (Ad-Grx1) resulted in up to 3.5-fold increase in NFκB-luc activity (Fig. 2.14, pg. 104). Grx activity of Grx1 over-expressed cells that were transfected with NFκB-luc was measured to determine whether the plateau of

NFκB-luc activity at higher concentrations of Ad-Grx1 was due to an interference of

NFκB-luc transfection on elevated Grx activity. However, Grx activity demonstrated a dose-response in cells that were transfected with NFκB-luc and over-expressing Grx1, and Grx activity was elevated up to nearly 4.5-fold at MOI 40 (n=1). These results support the conclusion that Grx1 regulates NFκB activity in rMC-1 cells, and

consequently the transcription of ICAM-1.

Sn50 blocks the Grx1-induced increase in ICAM-1 production in Ad-Grx1 infected rMC-

1 cells. NFκB is the most common regulator of ICAM-1 transcription and appears to be

the key mediator in Grx1-induced ICAM-1 expression in rMC-1 cells. We used an

inhibitor of NFκB nuclear translocation (sn50) to further test whether signaling to NFκB

is the major pathway regulating ICAM-1, and whether regulation by Grx1 is exerted on

the NFκB pathway in the cytosol or the nucleus. rMC-1 cells were transfected with Ad-

Grx1 at MOI 10 in the absence or presence of sn50 inhibitor in normal glucose

concentrations, and the lysates were analyzed for ICAM-1 production. Treatment with

12 µM and 20µM sn50 resulted in a dose-dependent decrease in the Grx1-induced

ICAM-1 production (Fig. 2.15, pg. 105). Cells over-expressing Grx1 in the presence of

20 μM sn50 contained amounts of ICAM-1 indistinguishable from control cells (no virus

(0) and Ad-Empty at MOI 10) (Fig. 2.15, A-B, pg. 105). Lower concentrations of sn50 did not have a significant impact on ICAM-1 expression in rMC-1 cells over-expressing

Grx1. No inhibition of ICAM-1 production was observed with 3µM sn50 (n=4, p=0.9), and similar results were observed with 6 µM sn50 (n=5, p=0.8). Several concentrations

(3, 6, and 12 µM) of sn50 were tested for inhibition on nuclear translocation of the p50 and p65 NFκB subunits and exhibited some inhibition but not in a dose-dependent

fashion (n=1). Based on the effects of sn50 on ICAM-1 production, 12 µM and 20 µM

sn50 would be expected to give a dose-dependent inhibition of nuclear translocation of

p50 and p65.

Knockdown of Grx1 in rMC-1 cells in high glucose prevents induction of ICAM-1

expression. To test the effects of decreasing intracellular Grx1 in high glucose

59 conditions, rMC-1 cells in diabetic-like concentrations of glucose (25 mM), were transfected with siRNA directed against Grx1, or with siRNA-control. Grx1 was knocked down by about 50% in the cells in high-glucose, i.e., to an amount similar to that in cells in normal glucose (Fig. 2.16, A and C, pg. 106). This knockdown of Grx1 was associated with a concomitant decrease in ICAM-1 expression (Fig. 2.16, B-C, pg. 106).

This result shows that a targeted decrease in Grx activity can prevent increased production of pro-inflammatory ICAM-1 under hyperglycemic conditions, indicating that the increase in ICAM-1 production in the cells under high glucose conditions is primarily due to upregulation of Grx1.

Attempts to detect membrane-bound ICAM-1 in rMC-1 cells. To further explore the

ICAM-1 induction in rMC-1 cells in high glucose, or from over-expression with Grx1, pilot studies of four different experimental methods were initiated to detect ICAM-1 expressed on the extracellular membrane of these cells. The techniques developed here were expected to determine whether the ICAM-1 induction seen on Western blots of cell lysates corresponded to an increase in surface expression of ICAM-1, and provide a foundation for subsequent studies (see Ch. 2 Discussion and Ch. 4; Future Directions).

The first experimental approach entailed a 96-well plate of cells incubated with anti-

ICAM-1 containing a fluorescein tag and analyzed via a fluorescent plate reader. No fluorescence values for cells incubated with anti-ICAM-FITC were detected above those for cells incubated with IgG-FITC control. To test the sensitivity of this system, a fluorescein labeled antibody (anti-mouse fluorescein conjugated antibody, GE Healthcare

0.5 mg/ml) that had an equivalent concentration to the ICAM-1-FITC antibody was diluted 1:5 to roughly 1:200,000 and used to generate a standard curve in the Novostar spectrofluorometer. Detection of the antibody (0.5 mg/ml) occurred between 0.1 mg/ml

(1:5 dilution) and 12.5 µg/ml (1:40 dilution), and 0.1 mg/ml gave the strongest signal.

Therefore, it is reasonable that no appreciable signal was detected on cells with 5 µg/ml

(1:100 dilution) anti-ICAM-1-FITC. Repeating these experiments with anti-ICAM-1-

FITC at a dilution of 1:5 might increase the signal from cells to a measurable value.

However, it is unlikely that the amount of ICAM-1 expressed on the surface of this small number of cells is 5 µg (0.1 mg/ml x 50 µl anti-ICAM-1 antibody) and that 100% of the antibody would adhere to ICAM-1 on the cell surface. Thus, incubating cells with a more concentrated solution of anti-ICAM-1-FITC (50 µl 0.1 mg/ml) is unlikely to give a measurable value; hence a more sensitive technique would be necessary. Furthermore, cells that were less than about 80% confluent did not remain well adhered to the 96-well plate in the absence of fixative. This presents a substantial problem with the initial goal of comparing ICAM-1 expressed on the surface of cells infected with Ad-Grx1 or Ad-

Empty because adenoviral infection are most efficient on cells that have not reached confluency.

Next, cells were plated in a 6-well dish, incubated with anti-ICAM-1-FITC, anti-ICAM-1 followed by anti-mouse-FITC, or anti-ICAM-1 followed by anti-mouse-HRP and TSA amplification. Cells were analyzed via a fluorescent microscope but no immunostaining was detected, likely due to insufficient sensitivity.

Thirdly, intact cells were plated in a 100 mm dish, labeled with biotin, lysed for

immunoprecipitation of ICAM-1 and proteins in the lysate were separated by SDS PAGE

for detection with streptavidin. A strong band corresponding to approximately 100 kDa

was detected via streptavidin from ICAM-1 immunoprecipitated from lysed cells that

were labeled with biotin after they were detached from the culture dish with trypsin (Fig.

2.17, A, pg. 107) (n=1). As expected, this band was not present when anti-YY1 was

substituted for anti-ICAM-1 in the immunoprecipitation of lysed cells that were labeled

with biotin when adherent to the dish or detached from the dish using trypsin.

Interestingly, cells that were labeled with biotin while adherent to the culture dish did not

yield a streptavidin band subsequent to immunoprecipitation with anti-ICAM-1 antibody.

This observation is likely due to the fact that detachment from the dish gives rise to

increased surface exposure of cells to biotin, increasing the number of ICAM-1 molecules that become biotin-labeled. This blot was immunoprobed with anti-ICAM-1 to verify that the observed band has immunoreactivity expected from ICAM-1. Indeed, an anti-ICAM-1 immunoreactive band was detected from ICAM-1 immunoprecipitations from lysed cells that were labeled with biotin as adherent or trypsinized cells, and this band was not detected in the immunoprecipitation of anti-YY1 (negative control) (Fig.

2.17,B, pg. 107) (n=1). The molecular weight of ICAM-1 is approximately 90 kDa, and typically, this band shows up between the 75 kDa and 100 kDa bands of these particular molecular weight markers (BioRad, Kaledioscope markers). The ICAM-1 immunoreactive bands detected in Figure 2.17, B (pg. 107) migrate slightly higher at 100 kDa, and this is likely due to the presence of many biotin molecules (Sulfo-NHS-Biotin, mw = 443.43) bound to ICAM-1. An anomaly of this experiment was that the whole cell

lysate did not show a clear band in response anti-ICAM-1, but this could be due to the

omission of methanol in the transfer buffer because previous Western blots without

methanol have given erratic results for detecting ICAM-1 in whole cell lysates.

Lastly, biotin labeling of intact cells (100 mm dish), followed by streptavidin pull down

and Western blot for ICAM-1 was carried out simultaneously to the previous experiment.

A band at approximately 100 kDa was immunoreactive with anti-ICAM-1 antibody after

a streptavidin pull down from cells that were trypsinized from the culture dish and subsequently labeled with biotin, but this band was less pronounced than that in the experiment presented above (Fig. 2.17, C, pg. 107) (n=1). A corresponding band was not observed from cells that had been labeled with biotin while adherent to the culture dish.

Taken together, these experiments support the conclusion that biotin labels the ICAM-1 on cell surface membranes more efficiently when the cells are in suspension compared to when they are adherent to a culture dish. Overall, these last two methods give rise to promising data that could be pursued to test whether the increased ICAM-1 protein in high glucose and Ad-Grx1 over-expressing cells corresponds to increased surface expression of ICAM-1 (see Ch. 2 Discussion and Ch. 4 Future Directions).

2.4 Discussion

Glutaredoxin in oxidative stress and disease

The regulation of cellular sulfhydryl homeostasis by glutaredoxin typically is considered

a protective mechanism, preventing irreversible oxidation of proteins and damage in cells

under oxidative stress (Shelton et al., 2005). Hence, changes in glutaredoxin activity could disrupt normal redox signaling and lead to pathological consequences. Indeed, several recent studies have identified changes in glutaredoxin in diseases involving redox perturbations. For example, glutaredoxin is increased in the brains of post-mortem patients with Alzheimer’s disease (AD), and it was proposed to mediate Aβ toxicity

(Akterin et al., 2006). Patients with chronic obstructive pulmonary disease (COPD) were reported to have decreased glutaredoxin with disease progression in alveolar macrophages and lung homogenates, but increased glutaredoxin in sputum supernatants

(Peltoniemi et al., 2006). In addition, glutaredoxin is elevated in animal models of

Parkinson’s disease, where inactivation of mitochondrial complex I is characteristic of

the disease (Kenchappa and Ravindranath, 2003). In isolated mitochondria, complex I

has been reported to be S-glutathionylated, and this modification was correlated to a loss

of activity (Taylor et al., 2003;Hurd et al., 2008). Taken together, these studies implicate

glutaredoxin in alterations of redox regulation associated with a variety of diseases

involving oxidative stress. For further review, see (Mieyal et al., 2008).

The importance of glutaredoxin in regulation of oxidative signal transduction is underscored by the identification of an increasing number of proteins whose cellular functions are modulated by S-glutathionylation, reversible by glutaredoxin (Shelton et al., 2005;Klatt and Lamas, 2000;Cotgreave and Gerdes, 1998;Giustarini et al., 2004).

The findings in the current study support the interpretation that glutaredoxin regulates the

NFκB signaling pathway and subsequent ICAM-1 expression in retinal Müller cells under diabetes-like conditions, suggesting that glutaredoxin might play an important role

in the hyperglycemia-induced inflammatory response known to occur in the retina in

diabetes.

Impact of TDORs in retinal Müller cells under diabetes-like conditions

The induction of glutaredoxin activity in rMC-1 cells by high glucose was indicated by a

disulfide reducing capacity assay of intact cells (Fig. 2.2, pg. 92) and a standard

radiolabel release assay of lysate (Fig. 2.5, pg. 95). Nearly a 2-fold increase in

glutaredoxin activity was observed in both the enzymatic radiolabel release assay and the disulfide reducing capacity. After accounting for the contributions of lipoate to the reduction of HEDS, the values of glutaredoxin activity according to the disulfide reducing capacity assay are greater than that of the radiolabel assay. This difference could be due to the experimental conditions. For example, the radiolabel assay reflects the de-glutathionylation of [35S] BSA-SSG catalyzed by glutaredoxin in lysate for 5

minutes. However, the reducing capacity assay monitors the glutaredoxin activity in

viable, intact cells over the course of 60 minutes, and the readout is the reduction of

DTNB by the newly deglutathionylated substrate (i.e., β−ME; see Ch. 2 Methods).

The induction of glutaredoxin by high glucose is underscored by the lack of induction of related enzymes. As mentioned above, glutaredoxin and thioredoxin are the two

cytosolic thiol disulfide oxidoreductases (TDOR), but differ in the mechanism of

regulation of thiol homeostasis (i.e., Grx1 specifically reduces Protein-SSG, and Trx

reduces intramolecular disulfides, intermolecular disulfides, and sulfenic acids). The

change attributable to the Grx system in the total disulfide reducing capacity assay

corresponded to independent measures of Grx1 protein content (Fig. 2.4, pg. 94) and

activity (Fig. 2.5, pg. 95). Therefore, the selective upregulation of glutaredoxin by high

glucose were chosen for the focus of this thesis work.

ICAM-1 in diabetic retinopathy and Müller cells - regulation by glutaredoxin

ICAM-1 protein expression is increased 2 to 3-fold in the whole retinae of diabetic rats

(Zheng et al., 2004;Joussen et al., 2002). Also, diabetic mice in which ICAM-1 has been

knocked out have decreased adherent leukocytes in the retina and less cell death (Joussen

et al., 2004), indicating that ICAM-1 contributes to disease progression in diabetic

retinopathy. In Müller cells in particular, isolated from diabetic rats, ICAM-1 gene

expression has been reported to be increased about 3-fold (Gerhardinger et al., 2005).

The role of ICAM-1 in retinal Müller glial cell function is yet to be fully elucidated, but

the contribution of these cells to retinal inflammation is considered likely by analogy to

glial cells in other contexts. Glial cells of the central nervous system (e.g. astrocytes)

contribute to neuronal inflammation in spinal cord injury, Parkinson’s Disease, and AIDS

via ICAM-1 production (Brambilla et al., 2005;Shrikant et al., 1995;Shrikant et al.,

1996;Sawada et al., 2006;Miklossy et al., 2006). In the studies of recovery after spinal

cord injury, selective inhibition of the NFκB signaling pathway provided neurological

protection (Brambilla et al., 2005).

Analogously, the rMC-1 cell culture model involving exposure to high glucose showed

increased ICAM-1 expression concomitant with increased expression of glutaredoxin

(Fig. 2.6, pg. 96). Furthermore, the increase in ICAM-1 production was elicited in

normal glucose by adenoviral expression of increased amounts of glutaredoxin (Fig. 2.10,

pg. 100 and Fig. 2.11, pg. 101). The magnitude of increase in ICAM-1 corresponds well with that of glutaredoxin except at the higher MOIs where glutaredoxin content exceeds the amount induced by high glucose. This observation suggests a maximal effect of glutaredoxin on ICAM-1 production. However, only the lower MOIs correspond to increases in Grx1 content comparable to those in rMC-1 cells cultured in high glucose or in the diabetic rat retinae.

To determine whether the observed increases in ICAM-1 expression in these model systems corresponded to an increase in ICAM-1 exposed on the extracellular membrane

surface, we labeled proteins on the extracellular membrane of rMC-1 cells with a biotin

derivative (see Ch. 2 Methods) and used streptavidin and an ICAM-1 antibody to detect

ICAM-1 (Fig. 2.17, pg. 107). The rationale is that if there is an increase in ICAM-1

expression in lysate and on the extracellular membrane, then ICAM-1 likely plays a pro-

inflammatory role in rMC-1 cells that is related to the adherence of leukocytes in the

retina. Alternatively, ICAM-1 could play an important role in cell signaling. Ligand

binding to ICAM-1 has been shown to trigger the endocytosis of the ligand-ICAM-1

complex in endothelial cells, and ICAM-1 can be subsequently returned to the membrane

for another cycle of shuttling (Muro et al., 2005).

Results consistent with the detection of membrane-bound ICAM-1 on rMC-1 cells cultured in normal glucose medium are shown in Figure 2.17 (pg. 107). However, the finding that ICAM-1 can undergo endocytosis in a ligand-ICAM-1 complex (Muro et al.,

2005) suggests that the techniques developed in this thesis would need to be applied to enriched membrane fractions isolated from cells to be more conclusive (see also Ch. 4;

Future Direction). Confirmation of the apparent detection of membrane-bound ICAM-1 would be interesting because it would indicate that rMC-1 cells express ICAM-1 on the membrane surface at a detectable level under non-stimulated conditions. Subsequently, the quantitative differences would need to be determined in membrane-bound ICAM-1 on rMC-1 cells cultured in normal glucose medium, high glucose medium, or over- expressing with glutaredoxin in normal glucose medium.

Utilization of sn50 as an inhibitor of NFκB nuclear translocation

Nuclear import is regulated by importin α/β heterodimers, and specifically, importin α directly interacts with nuclear localization signal (NLS) of proteins while importin β docks the importin-cargo complex to the nuclear pore complex (NPC). The p50 and p65 subunits each contain classical monopartite NLS consisting of basic amino acids, arginine and lysine residues, and undergo TNFα-induced nuclear import as a heterodimer predominantly via importin α3 and α4 (Fagerlund et al., 2005;Torgerson et al., 1998).

Additionally, importin α3 was reported to mediate nuclear translocation of p50 homodimers under basal conditions. Interestingly, importin α3 binds to p50 and p65 via its N-terminal and C-terminal NLS binding sites respectively (Fagerlund et al., 2005).

The cell permeable sn50 peptide contains the NLS of NFκB p50 and was designed to prevent NFκB nuclear translocation by blocking its binding to the importin α/β heterodimer (Torgerson et al., 1998). However, promiscuous effects of sn50 on other

transcriptional factors have become controversial. Unlike NFκB, c-Jun and c-Fos of the

AP-1 transcription factor contain bipartite NLSs, NFAT predominantly uses a unique sequence for transport, and STAT has a unique but unidentified NLS (Torgerson et al.,

1998). Despite seemingly diverse transport signals, sn50 at higher concentrations (27-75

µM) has been reported to inhibit the nuclear import of these transcription factors

(Torgerson et al., 1998;Kolenko et al., 1999;Boothby, 2001;Ray, 2001). The specificity of sn50 was addressed in the context of the suspension cells, lymphocytes and Jurkat cells, and could be cell specific, especially in regards to adherent cells. Nevertheless, selective inhibition of NFκB was reported at 13 µM sn50 (Kolenko et al., 1999;Ray,

2001), and these data served as the basis of selecting inhibitor concentrations in the current studies. The data generated under sn50 inhibition suggested that (1) the site of regulation by glutaredoxin within the NFκB pathway primarily occurs in the cytoplasm, not the nucleus and (2) NFκB is the predominant transcriptional factor responsible for

changes in ICAM-1 in glutaredoxin over-expressing rMC-1 cells (Fig. 2.15, pg. 105).

These key observations set the stage for the next phase of these studies (see Chapter 3).

NFκB signaling and potential targets for regulation by S-glutathionylation and

glutaredoxin.

NFκB nuclear translocation of p50 was more pronounced than that of p65. This could be due to the nuclear translocation of abundant p50 homodimers. Also, unlike p50, p65 has a

strong nuclear export signal (Fagerlund et al., 2005), suggesting that p50 spends an

extended time in the nucleus compared to p65. Comparing p50 and p65 in the nucleus at

an earlier time point than we studied may reveal identical expression levels.

Furthermore, IκBα moves in and out of the nucleus by its own NLS and NES, and can be exported but not imported in complex with NFκB (Fagerlund et al., 2005). IκBα primarily associates with p65, and therefore may facilitate nuclear export of p65. p50 homodimers are thought to act as DNA repressors, and if more p50 is present in the nucleus, enhanced NFκB-driven expression of ICAM-1 would not be expected.

Therefore, several possibilities exist: (1) the difference in nuclear translocation observed between p50 and p65 is not substantial enough to make a functional difference; (2) more p50 exists via homodimers but does not act as a repressor in Müller glial cells; (3) more p50 exists and acts as a repressor, but a co-factor or additional protein that also responds to Grx1 over-expression prevents repression by p50; or (4) more p50 exists but does not act as a repressor in this system.

The current study is unique in showing the increased activation of NFκB in Müller cells linked to corresponding changes in glutaredoxin activity, implicating regulation via reversible glutathionylation of one or more components of the NFκB signaling pathway.

Inhibiting the nuclear translocation of NFκB in rMC-1 cells over-expressing Grx1 blocks the corresponding increase in ICAM-1 production (Fig. 2.15; pg. 105), suggesting that the target for Grx1-regulated S-glutathionylation is a cytoplasmic signaling protein in the

NFκB pathway, upstream of nuclear p50-p65. Glutaredoxin could regulate NFκB activity via the glutathionylation status of upstream mediators in the cytoplasmic NFκB signaling pathway or via the glutathionylation status of the NFκB subunits (p50 and p65) in the nucleus. The site of Grx1 regulation may be cell type and signal dependent. P50 has been shown to lose DNA binding activity upon S-glutathionylation in vitro (Pineda-

Molina et al., 2001), and this is likely predictive of modulation of p50 activity in a physiological setting. With pancreatic cancer cells, hypoxia and N-acetylcysteine treatment led to inactivation of p65, and glutaredoxin was shown to restore the p65 transcriptional activity, indicative of p65-SSG formation in situ (Qanungo et al., 2007).

Typically, inactive NFκB is sequestered in the cytoplasm by its inhibitory protein, inhibitor of NFκB (IκB). The phosphorylation of IκB by a complex of IκB kinases

(IKKα, IKKβ, and IKKγ) precedes ubiquitination and degradation of IκB, releasing

NFκB for nuclear translocation where it binds to DNA and activates transcription.

Whereas IKK regulates the phosphorylation of IκB, mediators in the ubiquitin-protease pathway (UPP) regulate the degradation of IκB, both processes contributing to NFκB activation.

IKKβ is the IKK subunit with a primary role in inflammation, and it has redox sensitive cysteines (Jeon et al., 2003;Rossi et al., 2000). NFκB activity was recently shown to be regulated by S-glutathionylation of IKKβ in alveolar epithelial cells (Reynaert et al.,

2006). S-glutathionylation has also been reported to inhibit the ubiquitin-activating (E1) and ubiquitin-carrier (E2) enzymes (Obin et al., 1998;Jahngen-Hodge et al., 1997), and the 20S proteasome in S. cerevisiae (Demasi et al., 2003). The 20S proteasome constitutes part of the 26S proteasome that degrades IκB, and cleaves p50 from its p105 precursor (Moorthy et al., 2006). The multitude of regulatory sites mediated by S- glutathionylation presents a complex picture, and further studies are needed to distinguish which mediators are most pertinent to regulation of the NFκB pathway by glutaredoxin

within the context of diabetic retinopathy and the retinal Müller cells (See Chapters 3 and

4).

Regardless of which specific components of the NFκB pathway are modulated, the

current studies show that increases in glutaredoxin activity lead to increased nuclear p50

and p65 subunits of NFκB in rMC-1 cells, both in response to high glucose and in response to over-expression of glutaredoxin in normal glucose; and the increases in Grx1

were similar under these different situations (Figs. 2.7, pg. 97; 2.12, pg. 102; 2.13, pg.

103). Collectively these data support the conclusion that glutaredoxin regulates NFκB

activity and concurrently the production of ICAM-1. This regulation is altered under

high glucose conditions mimicking diabetes.

Glutaredoxin; a potential therapeutic target in diabetic retinopathy.

When glutaredoxin is over-expressed to mimic induction of the enzyme by high glucose

(2 to 4-fold), the extent of the changes in ICAM-1 and NFκB are comparable with

changes induced by high glucose in cell culture, or physiologically in the diabetic

animals. These studies suggest that glutaredoxin plays an inflammatory role in the

response to diabetes in the retina, most likely through regulation of S-glutathionylation

status of redox sensitive cysteine-containing proteins. Remarkably, knocking down

glutaredoxin is an effective means of dampening ICAM-1 production under high glucose

(Fig. 2.16, pg.106). This finding identifies glutaredoxin as a potential target for

pharmacological intervention in diabetic retinopathy. In addition, induction of

glutaredoxin in hearts of diabetic rats (Li et al., 2005) suggests that inhibition of the

enzyme in diabetes might have benefits to other tissues besides the retina that also suffer

from complications of diabetes. Additional work is needed to determine the efficacy of

glutaredoxin inhibition in combating the tissue pathologies associated with diabetes, and

possibly other inflammatory diseases.

Overall, these results directly identify a role for glutaredoxin in regulation of ICAM-1

expression in Müller cells, and implicate changes in glutaredoxin in the inflammatory

response characteristic of diabetic retinopathy.

2.5 Materials and Methods

Cell Culture- Cell culture supplies were obtained from Invitrogen except where indicated.

Rat retinal glial (Müller) cells (rMC-1) were a kind gift from Dr. Vijay Sarthy

(Northwestern University, Chicago, IL). Cells were cultured for up to five days in high glucose (25 mM, cat # 11995-065) or normal glucose (5 mM, cat # 11885-084) DMEM

with 2% heat inactivated FBS (Fisher, Cellgro cat # MT35-011-CV) and 2 mM glutamine

with daily replacements in a humidified 37°C incubator with 5% CO2. HEK 293 cells

were cultured in high glucose DMEM with 10% FBS and 1% Pen-Strep in a humidified

37°C incubator with 5% CO2. Both cell lines were stored frozen in 10% DMSO in 10%

FBS containing DMEM in a liquid nitrogen cell storage tank.

Measurement of Glucose in Cell Culture Medium- Glucose concentrations in the medium were monitored with a glucose oxidase kit (POINTE Scientific #G7521-120) to ensure

73 that glucose consumption by the cells did not deplete the medium. Glucose oxidase oxidizes glucose to gluconate and hydrogen peroxide, which reacts with phenol, peroxidase, and an aminophenazone to produce a red quinoneimine dye. 200 µl glucose oxidase reagent was added to a 96 well plate and pre-incubated at 37°C for 5 min. 2 µl of medium from cells cultured in normal or high glucose medium were added to each well, incubated an additional 10 min at 37°C, and read at an absorbance of 490 nm with a

Molecular Devices THERMOmaxTM microplate reader. To ensure that the medium being applied to the cells was of the expected concentrations, stock medium (normal or high glucose) was assayed in each experiment. The functional coefficient (0.0388 A mM-1) was determined from a standard curve for glucose (0-27 mM) glucose standards. Data were analyzed with the Molecular Devices SOFTmax® version 2.3.

Animal Retinae- Treatment of animals was in accordance with the ARVO Resolution on

Treatment of Animals in Research and Case Western Reserve University guidelines.

Animals were treated with streptozotocin (STZ) to induce diabetes and with insulin to prevent wasting, as described previously (Du et al., 2002). Retinas were excised from rats ten weeks after induction of STZ-induced diabetes or from non-diabetic control rats, and homogenized in 50 mM Tris-HCl pH=7.4, 10% NP40, 0.25% sodium deoxycholate, and 150 mM NaCl.

Grx Activity of Rat Retinal Homogenates- Rat retinal homogenates (0.1-0.2 mg) were assayed for Grx activity via GSH-dependent release of radiolabel (as [35S] GSSG) from

the prototype substrate [35S] BSA-SSG as previously described (Chrestensen et al.,

2000;Srinivasan et al., 1997).

Grx Activity of rMC-1 Cells in Normal Glucose (5 mM) or High Glucose (25 mM)

Medium- Grx activity in lysates from cells cultured in either normal glucose (5 mM) or

high glucose (25 mM) medium was measured via a plate reader assay monitoring

changes in absorbance due to the oxidation of NADPH by GSSG reductase that is

coupled to Grx activity in lysates; or by the catalyzed release of radiolabeled GSSG (i.e.,

de-glutathionylation of [35S] BSA- SSG) by the Grx1 in lysates. Alternatively, Grx

activity of intact cells was measured by the changes in the reduction of the cell permeable

substrate, bis (2-hydroxyethyl) disulfide (HEDS) that becomes S-glutathionylated within

the cell. All three methods are described below.

Cellular disulfide reducing capacity of intact rMC-1 cells: rMC-1 cells (50,000-100,000 cells per 60 mm dish) were cultured in normal or high glucose medium for 3-5 days and

assayed for total disulfide reducing capacity (Biaglow et al., 2000) with two different cell permeable disulfides. Reduction of bis (2-hydroxyethyl) disulfide (HEDS) is attributable to total disulfide reducing capacity (Trx and Grx systems), and lipoate reduction is selective to the Trx system (Biaglow et al., 2000). Intracellular HEDS ((β−ME)2) can be

directly reduced by the Trx system (NAPDH and thioredoxin reductase) or converted by

glutathione to generate beta-mercaptoethanol (β−ME) and hydroxyethyl-glutathione

disulfide which then acts as a substrate for Grx1 coupled to GRase and NADPH. Lipoate is selectively reduced to dihydrolipoic acid by the Trx system. Cells were incubated in 5

ml of medium containing 5 mM HEDS or 5 mM lipoate. Aliquots of the medium were

taken at 0, 5, 10, 20, 30, 45, and 60 min, added to separate wells of a 96-well plate

containing dithio-bis (2-nitrobenzoic acid) (DTNB) (1 mM final) in each well, and

absorbance change at 405 nm for each well was monitored in a plate reader. The

functional coefficient (6.1 A mM-1) was determined from a standard curve for GSH using

the DTNB assay with 0.2 ml total volume in each well, and reading absorbance values

with a Molecular Devices THERMOmaxTM microplate reader. Data were analyzed with

the Molecular Devices SOFTmax® version 2.3.

Grx activity in lysates of rMC-1 cells via de-glutathionylation of [35S] BSA-SSG- 0.3-5 mg of protein from lysate of cells grown in either normal or high glucose medium was tested for Grx deglutathionylation activity via the standard radiolabel release assay using

[35S] BSA-SSG as the substrate, as previously described (Chrestensen et al., 2000).

Briefly, the lysate was pre-incubated with 0.5 mM GSH in 0.1 M sodium potassium

phosphate buffer pH 7.5 at 30°C for 5 min. The reaction was initiated by the addition of

[35S] BSA-SSG (0.1 mM final concentration) and the total volume of the mixture was

500 µl. Aliquots (100µl) were removed at 10 sec, 1 min, 2 min, and 3 min of each

reaction and mixed with 100 µl ice cold 20% trichloroacetic acid (TCA). Proteins

(cellular proteins, BSA-SH product, and un-reacted [35S] BSA-SSG substrate) were

precipitated, and newly released [35S] GSSG product is left in the supernatant, which is then counted in a liquid scintillation counter.

Grx activity in lysates of rMC-1 cells via the GRase-coupled assay(Chrestensen et al.,

2000): rMC-1 cells were cultured in normal or high glucose medium for 5 days, and then

1-3 million cells were sonicated in 0.75% triton X lysis buffer (0.75% triton X-100 (v/v),

3.8% Sigma Protease Inhibitor (SPI) and 50 mM Tris-HCl pH 7.5 was diluted with 50%

Dulbecco’s PBS (Invitrogen) for a final concentration of 4 mM sodium phosphate, 0.7 mM potassium phosphate, 69 mM sodium chloride, 1.3 mM potassium chloride).

Alternatively, cells were lysed in assay mix (0.2 mM NAPDH, 0.5 mM GSH, 0.1 M sodium potassium phosphate buffer pH = 7.5, and 2 U/ml GSSG reductase (GRase) in

Millipore water pre-treated with chelex resin). The lysate was pre-incubated in 180 µl assay mix at 37°C for 5 min. The enzymatic reaction was initiated by addition of 20 µl of the prototype substrate cysteinyl-glutathione disulfide (CSSG, 1 mM) with a multi- channel pipette. The absorbance was read at 340 nm over the course of 5 min in a

Molecular Devices THERMOmaxTM microplate reader. Data were analyzed with the

Molecular Devices SOFTmax® version 2.3.

Grx was barely detectable in cells that were lysed by sonication in Triton X-100 lysis buffer (0-0.01 U/ml) (n=2). Cells were then lysed by sonication in assay mix, and gave rise to Grx activities in the range of 0.1-1.9 U/ml. Cells cultured in high glucose medium gave rise to about 0.9 U/ml +/- 0.4, and that of normal glucose medium was nearly 0.2

U/ml +/- 0.07 (n=6, p = 0.08) (n=6). The high glucose-grown cells displayed greater variability than those in normal glucose but it is likely that more repetitions would yield a significant difference at the p < 0.05 level in Grx activity between cells grown in normal and high glucose medium. Protein concentration of the lysates is needed to normalize the

77 samples to one another and to compare these data to those of the radiolabel release assay described above. However, concerns with the interpretation of these values due to potential interferences with the spectrophotometric assay when cell lysates or tissue homogenates are used led to the use of the standard radiolabel release assay to determine

Grx activity. Namely, NADPH oxidase in cell lysate could contribute to the NAPDH oxidation at 340 nm in the coupled reaction in the plate reader, and light scattering from fine particulate matter in the lysate could interfere with the absorbance readings.

Glutathione Reductase (GRase) Activity in Lysates of rMC-1 Cells Cultured in Either

Normal Glucose (5 mM) or High Glucose 25 mM) Medium- 10 million rMC-1 cells were sonicated in 300 µl PBS. 5-10 µl of lysates (equivalent to 1.7-3.3 million cells) was pre- incubated at 30°C for 5 min. with 180 µl 0.2 mM NADPH in potassium phosphate buffer in wells of a 96-well plate. The reaction was initiated with 20 µl 1 mM GSSG, and the rate of oxidation of NADPH (A340) was monitored in a plate reader for 5 min.

Thioredoxin Reductase (TRase) Activity in Lysates of rMC-1 Cells Cultured in Either

Normal Glucose (5 mM) or High Glucose (25 mM) Medium- 10 million rMC-1 cells were sonicated in 300 µl PBS. 10-20 µl of lysates (equivalent to 3.3-6.6 million cells) was pre-incubated at 30°C for 5 min. with 10 µl 100 mM EDTA and 10 µl 20 mM DTNB in

1% NaHCO3. 160 µl 0.2 mM NADPH in 0.1 mM sodium potassium phosphate buffer pH

7.5 was added, and the rate of reduction of DTNB (A405/min) was monitored in a 96- well plate reader for 5 min.

Thiol Content of Lysates of rMC-1 Cells Cultured in either Normal Glucose (5 mM) or

High Glucose (25 mM) Medium- 10 million rMC-1 cells were sonicated in 300 µl PBS.

10-20 µl of lysates (equivalent to 3.3-6.6 million cells) was pre-incubated at 30°C for 5

min. with 10 µl 100 mM EDTA and 10 µl 20 mM DTNB in 1% NaHCO3. 160 µl 0.2

mM NADPH in potassium phosphate buffer was added, and the end point absorbance

reading was measured at 405 nm in a 96-well plate reader.

Propagation and Titration of Adenoviral Constructs in HEK 293 Cells - Adenoviral

vector containing the Grx1 cDNA construct (Ad-Grx1) and empty vector control

construct (Ad-Empty) were created with the CRE-Lox recombination system in

collaboration with Dr. Yong Lee (University of Pittsburgh, PA) (Song et al., 2002).

These vectors have been made deficient for E1 and E3 genes. While E3 is dispensable,

E1 is necessary for the assembly of infective virus particles. Therefore, in order for the virus to propagate, E1 must be provided in trans from a cell type such as HEK 293 that has been transformed by sheared human adenovirus type 5 (Ad 5). For propagation,

HEK 293 cells were infected with 5 plaque forming units (pfu)/cell of adenovirus (Ad-

Grx1 or Ad-Empty). Cells and corresponding medium were collected together at the time when the cells lifted off the plate (usually after 3-6 days). The cells were lysed via freeze-thaw three times and then virus was collected by centrifugation at 2,300 x g for 10 min at 4°C. For adenoviral titration, HEK 293 cells were infected with serial dilutions

(down to a dilution of 10-14) of stock virus, overlayed with 0.9% low melting point

agarose, and incubated until plaques stopped forming (usually 5-7 days). A plaque is

formed by the death of HEK 293 cells, and appears as a hole in an otherwise confluent

79 dish of HEK 293 cells. In order to determine the concentration of the virus (pfu/ml), the final number of plaques was counted and divided by the volume of the adenovirus used to infect the cells. Common mistakes to avoid are application of agarose to the cells that is too warm and utilization of HEK 293 cells that are not at 90-95% confluency at the time of overlay.

Adenoviral Expression of Grx1 in rMC-1 Cells – rMC-1 cells (500,000 cells in a 100 mm dishes) were grown in normal glucose medium for two days, infected with various multiplicities of infection (MOI 0 to 80) of Ad-Grx1 or Ad-Empty in 1 ml serum-free

DMEM for one hour. The viral medium was removed, and cells were cultured for two days in normal glucose medium and collected in 1% NP40 lysis buffer (50 mM Tris pH=8, 1% NP40, and 150 mM NaCl),

Inhibition of Nuclear Translocation of NFκB via sn50 in Adenoviral Over-expressing rMC-1 Cells - rMC-1 cells (approx. 1,000,000 cells in a 60-100 mm dish) were grown in normal glucose medium for two days, infected with MOI 10 of Ad-Grx1 in 0.5-1 ml serum-free DMEM for one hour in the absence or presence of 3-24 µM sn50 inhibitor

(Biomol). Cells were subsequently cultured in normal glucose medium in the absence or presence of sn50 inhibitor and collected in 50-100 µl 1% NP40 lysis buffer the following day, yielding roughly 0.5-1 mg of total protein. 100 µg of lysate was loaded to an SDS

PAGE gel and immunoprobed for ICAM-1 expression and normalized to actin or

GAPDH. Control cells (uninfected, and MOI 10 Ad-Empty) were incubated in parallel in

the absence of sn50. Nuclear extracts were generated from Ad-Grx1 cells treated with 3,

6, or 12 µM sn50, and immunoprobed for p65, p50, actin, and YY1.

Immunoblotting - Müller cells were collected, lysed in 1% NP40 lysis buffer and centrifuged at 1,500 x g for 5 min. Cleared supernatants were assayed for protein content with the Micro-bicinchroninic acid method (BCA) (Pierce, Rockford, IL), according to the manufacturers protocol. Samples were mixed 4:1 with 4X SDS sample buffer (0.5 M

Tris-HCL pH=6.8, 20% glycerol, 10% SDS (w/v), 1% bromophenol blue and 20 mM dithiolthreitol (DTT)), heated for 15 minutes at 95°C. Proteins were separated by 12%

SDS-PAGE and transferred to Immobilon P membranes (Millipore). Membranes were immunoprobed with the appropriate antibodies and corresponding dilutions indicated in

parenthesis: anti-p50 (1:1,000) (ab7971) (AbCam, Cambridge, MA), anti-p65 (1:3,000)

(cat # sc372) (Santa Cruz Biotechnology, Santa Cruz, CA), anti-ICAM-1 (1:500) (R & D

Systems, Minneapolis, MN cat # MAB5832), anti-IL1β (1:500) (R & D Systems cat #

MAB501), anti-GAPDH (1:10,000) (Chemicon International Inc., Temecula, CA cat #

MAB374); anti-actin (1:30,000) (Sigma, St. Louis, MO cat # A1978), anti-YY1 (1:1,000)

(Santa Cruz Biotechnology cat # sc-7341) and anti-Grx1 (1:1,000) (prep1, generated and

purified via an adaptation of the McKinney & Parkinson caprylic acid method (Gravina

and Mieyal, 1993). Peroxidase-conjugated secondary goat anti-rabbit (cat # 111-035-

144) or anti-mouse antibodies (cat # 115-035-166) (1:10,000) (Jackson ImmunoResearch

Laboratories, West Grove, PA).were used, and Western Lightning Chemiluminescence

Reagent Plus (Perkin Elmer Life Sciences, Boston, MA) was used according to the manufacturer’s protocol (Perkin Elmer Life Sciences). Band intensities were quantified

using a BioRad Calibrated Imaging Densitometer GS-710 with BioRad Quantity One

software version 4.1.1. Changes in band intensity are reported as ratios relative to

loading controls.

Nuclear Extraction - Müller (rMC-1) cells were collected in 1 ml PBS and centrifuged for 3 min. at 800 x g and lysed in 300 μl low salt buffer (20 mM HEPES pH=7.6, 20% glycerol, 10 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, and 0.1% Triton X-100) for 20

min. Centrifugation at 800 x g for 3 min yielded a cytosolic supernatant. The nuclear

pellet was washed twice in PBS, incubated in 80 μl of high salt buffer (10 mM HEPES pH=7.6, 10% glycerol, 0.5 M NaCl, 0.7 mM MgCl2, 0.1 mM EDTA, and 0.05% Triton-X

100) for 30 min 4°C, and centrifuged at 16,000 x g for 15 min in 4°C. Protein content

was determined via BCA assay. Equal amounts of protein from nuclear and cytosolic

fractions were run on a gel for Western blot analysis of p50, p65, YY1, actin, and

GAPDH.

Plasmids- NFκB-luciferase plasmid (pNFκB-luc, 5x NFκB enhancer, 5.7 kb, Ampicillin

resistant, cat # 219078, Stratagene) was used to analyze NFκB activity. The binding

element for the NFκB-luciferase plasmid is derived from the consensus NFκB binding

sequence and contains five repeats of (TGGGACTTTCCGC). Empty-luc (Stratagene, cat

# 219087) was used as a negative control, and the Renilla plasmid (phRG-TK, 4.8kb,

Ampicillin resistant, Promega, Madison, WI) was co-transfected as a transfection

efficiency control with the plasmid of interest. All plasmids (approx. 1 µg) were

transformed in DH5α bacterial cells. Briefly, the plasmid was incubated with the cells

for 30 min on ice, heat shocked for 90 sec at 42°C, incubated on ice for an additional 1 min, added to 200 µl LB and incubated at 37°C in a shaker for 2 hrs. Thereafter, 20 µl of cells were spread onto an ampicillin-resistant (50-100 µg/ml) agar plate and incubated overnight at 37°C. A colony from the plate (or stick from glycerol stock) was used to initiate a 2 ml starter culture containing 50-100 µg/ml ampicillin. After 4 hrs at 37°C shaking, this culture was added to an overnight culture of 250 ml for subsequent DNA purification via maxi-prep kits (Qiagen cat # 12362 or BioRad Quantum prep cat #732-

6130). Sometimes the starter culture step was omitted. The newly generated NFκB-

Firefly-luciferase and Renilla-luciferase plasmids were digested with the restriction

endonucleases PVU1 and HINDIII respectively, and looked consistent with the original

stock of plasmids on a 1% agarose gel. Nicked, linear, and supercoiled forms of DNA

have distinct migration patterns on a gel, and the individual bands are labeled in Figure

2.18, (pg. 108). A Traf plasmid (provided by Dr. Jonathan Mosley, laboratory of Dr.

Ruth Keri) was also utilized briefly during the optimization period as a positive control

for activation of NFκB.

NFκB-Luciferase Reporter Assay - Müller (rMC-1) cells (50,000 cells/well of a 6-well

dish) were grown for two days, and co-transfected for 10-12 hr with 1 μg of NFκB-

Firefly-luciferase plasmid (Stratagene) and 0.1 μg Renilla-luciferase plasmid (Promega)

as a transfection efficiency control reporter according to the lipofectamineTM reagent protocol (Invitrogen). 2-4 hrs after the end of the transfection, cells were infected with

Ad-Grx1 or Ad-Empty for 1 hr and collected 8hrs later in 1x passive lysis buffer

(Promega). NFκB activity was assayed via the Dual-Luciferase® Reporter Assay System

(Promega) with the Molecular Devices Lmax Luminometer and SOFTmax PRO software. Assay readouts were reported as ratios of luminescence readings for firefly- luciferase and renilla-luciferase.

Initially these experiments were designed to analyze NFκB-luc activity of cells that had been transfected with NFκB one day subsequent to Ad-Grx1 infection (0.2-0.3 ml Ad-

Grx1 infection medium). This method gave rise to approximately 1.5-3-fold increased

NFκB-luc activity in Ad-Grx1 over-expressing cells at 8-36 hr from the end of the NFκB transfection (n=3). The increase in NFκB-luc in response to Ad-Grx1 infection did not seem to be well maintained at 48 hr after the end of transfection (i.e., 72 hr post-Ad-Grx1

infection) (n=2). In addition, as expected co-transfection of an Empty-luc with Renilla

gave minimal values (n=3). Nevertheless, transfection of the cells with NFκB-luc prior

to Ad-Grx1 infection somewhat enhanced the observed increase in NFκB-luc activity to 2

to 4-fold and was therefore the preferred protocol. Co-transfection of Renilla and NFκB

plasmids with the positive control Traf plasmid (0.1, 0.5, and 1 µg), an upstream activator

of NFκB, led to a 3 to 6-fold increase in NFκB-luc activity, but an unusually low value

for NFκB-luc activity with 0.5 µg Traf limited the experiment to a two-point dose-

response relationship (n=1). Unfortunately, for unknown reasons, the increase in NFκB-

luc activity in response to Traf was not reproducible (n=3).

To determine whether the plateau of NFκB-luc activity at high MOIs of Ad-Grx1 was

due to transfection interference with Grx activity, the Grx activity of NFκB-luc

transfected cells was measured via the [35S] BSA-SSG assay. One 6-well dish was

transfected with NFκB-luc and Renilla followed by Ad-Grx1 at MOI 5, 10, 20, or 40.

Each well of cells in a 6-well plate was collected in 80 µl NP40 lysis buffer and 50 µl

(0.2-0.3 mg) of lysate was assayed. Grx activity demonstrated an Ad-Grx concentration

dependence (see Results).

Grx1 Knockdown via rat siRNA - rMC-1 cells (30,000-40,000 cells/well of a 6-well dish)

were grown in high glucose (25 mM) for one day, transfected with Dharmacon ON-

TARGETplus SMARTpool siRNA targeted to rat Grx1 or ON-TARGETplus

siCONTROL non-targeting pool siRNA according to manufactures instruction for

oligofectamine (Invitrogen), and grown in high glucose for three subsequent days. Cells

were lysed in NP40 lysis buffer for immunoblotting with antibodies directed towards

Grx1, ICAM-1, actin, and GAPDH.

Alternative methods attempted to knockdown Grx1 - The human sequence for Grx1 (Glrx

gene) has about an 84% homology with the rat homolog, and we hypothesized that the

human oligonucleotides might give a partial knock down of Grx1 in rat cells.

Knockdown of Grx1 was initially attempted using human siRNA-Grx1 and seemed to decrease protein expression of Grx1 by about 40% and ICAM-1 by approximately 20%

(n=2). However, control siRNA-scrm was not analyzed in parallel in these experiments,

and it knocked down both Grx1 and ICAM-1 more than siRNA-Grx1 (n=1). Therefore,

an alternative approach of stable Grx1-knockdown was pursued. Dr. Harish Pai in our lab ligated four different oligonucleotide sequences (rat/mouse Grx1 94-112, rat/mouse

Grx1 231-249, rat/mouse Grx1 300-318, and human/rat/mouse Grx1 240-258) or a

scrambled (scrm) sequence for control into p.SUPER.retro.puro vector (OligoEngine,

Seattle WA) and subsequently transformed in DH5α cells. Plasmids were purified with

Qiagen plasmid kits, and 2-4 µg of each plasmid (total amount of plasmid was 8-16 µg

for Grx1 and 2-4 µg for scrm) was transfected with liopfectamine and PLUS reagent into

retroviral producing cells (BOSC cells). Medium (5 ml) from the BOSC cells was added

to 5 ml standard rMC-1 DMEM medium and filtered onto rMC-1 target cells in the

presence of 4 µg/ml solution of polybrene. rMC-1 cells received medium from BOSC

cells plus polybrene for three consecutive days at which time puromycin selection was

begun. A kill curve on untransfected rMC-1 cells revealed that 2 µg/ml puromycin was

the minimal dose that killed the cells, and therefore this concentration was used on the

transfected cells. shRNA viral infected rMC-1 cells were assayed for Grx activity via

[35S] BSA-SSG assay. Grx activity was found to be decreased by approximately 50% in

cells infected with shRNA-Grx1 over shRNA-control (scrm) when grown in either

normal or high glucose medium for four days (n=9). However, Western blot analysis revealed that sh-scrm infected rMC-1 cells in high glucose contained about 20% more of

Grx1 and 60% more ICAM-1 than uninfected cells in either normal or high glucose

medium (n=4). Therefore, these knockdown systems were not serving as valid models,

and siRNA directed towards rat Grx1 was pursued.

Approaches to detection of ICAM-1 Bound to the Surface of Membranes of rMC-1 Cells-

Four methods were employed to analyze the membrane-bound expression of ICAM-1 in rMC-1 cells.

Intact cells immunolabeled with an anti-ICAM-1-FITC antibody and detected with a fluorescent plate reader: To optimize cell growth in a 96-well plate, different numbers of cells were plated and monitored for growth over the course of three days. Plating 5000 cells/well led to 85% confluency, 2000 cells led to 50% confluency, 1000 led to 30% confluency, and 500 cells led to 10% confluency after 3 days in culture. Subsequently,

5000 cells were seeded and infected with Ad-Grx1 or Ad-Empty after two days of growth. The following day, cells were incubated for 30min with 50-100 µl 5 µg/ml mouse anti-ICAM-FITC (BD Pharmingen cat # 554969) or mouse IgG-FITC (BD

Pharmingen cat # 550616) diluted in medium. A fixative step with 1% paraformaldehyde was either omitted, or performed subsequent to the primary antibody incubation, or done prior to the primary antibody incubation. Fixation can increase the permeability of the cell membrane, and therefore fixative prior to the primary antibody was expected to represent total cellular expression of ICAM-1 (i.e., intracellular and surface expressed).

ICAM-1 has been identified in endosomal compartments (Muro et al., 2005), and according to the rationale that the paraformaldehyde permeabilizes all membranes equally, this compartment of ICAM-1 would be detected by anti-ICAM-1-FITC as well.

After three washes in PBS, cells were read in the 96-well plate by the Storm 840

Fluorescent Plate Reader at a gain of 1000. Gains of 500, 800, and 1200 were also tested.

Intact cells immunolabeled with an anti-ICAM-1-fluorescein antibody and visualized with a fluorescent microscope: 25,000-50,000 cells/well were plated in a 6-well dish and after three days in culture, a section of the plate containing cells was outlined with a pap pen (Research Products International) and incubated with the anti-ICAM-1-FITC

described above at 1:10, 1:00, and 1:1000 dilutions in 300 µl medium. Cells were rinsed twice with PBS, and a cover slip was placed over the cells. Cells were visualized using a

LEICA DMI6000 inverted microscope with an XCite series 120 lamp source. The 1:10 dilution (50 µg/ml) of anti-ICAM-1-FITC was determined to be the optimal dilution among those tested, and the antibody was then incubated for 10, 20, and 30 min. Several approaches were undertaken to increase the signal, including incubation with mouse anti-

ICAM-1 followed by a fluorescein conjugated anti-mouse antibody and a mouse anti-

ICAM-1 antibody followed by a horseradish peroxidase (HRP)-conjugated anti-mouse

antibody detected with a tyramide signal amplification (TSA) kit (Perkin Elmer Life

Sciences cat # NEL701A).

Intact cells labeled with biotin, immunoprecipitated for ICAM-1, and detected with a

Western blot for streptavidin: 100 mm confluent dishes of cells were washed in Hanks

Buffered Salt Solution (HBBS), and cells either adherent to the dish or trypsinized and placed in a 0.5 ml eppendorf tube were incubated with 0.6 ml solution of HBSS with 2

mM Biotin reagent (Pierce cat # 21217, EZ-link Sulfo-NHS-Biotin, 100 µl 10 mM) on a

rotating platform for 30 min RT. The activated biotin reagent reacts with primary amines

including the ones found on adhesion molecules on the cell surface. Cells were washed

in HBSS containing 100 mM glycine three times to stop the biotinylation reaction. Cells

were lysed in 0.5 ml RIPA buffer (150 mM NaCl, 10 mM EDTA, 1% NP40, 0.5%

deoxycholic acid (DOC), 0.1% SDS, 50 mM Tris pH 7.5, 1.4 mM phenylmethylsulfonyl

fluoride (PMSF), 14µg/ml aprotinin, and 14 µg/ml leupeptin) on ice for 20 min. Lysate

was incubated with 5 µl anti-ICAM-1 antibody (0.5 mg/ml) or 12.5 µl anti-YYl (0.2

mg/ml, non-specific control antibody). Protein A/G beads (40 µl) were added to the

mixture and incubated overnight at 4°C on a rotating platform. The suspension was

centrifuged at 4,000 x g for 4 min, and the pelleted beads were washed once with RIPA

buffer and twice with PBS. The sample was eluted with boiling for 15 min in 40 µl 4X

sample buffer and then proteins were separated in an SDS PAGE gel for Western blot

detection with HRP-conjugated streptavidin at a dilution of 1:4,000. Methanol was omitted from the transfer step of the Western blot as was done for immunoblotting of other antibodies during that time, and this omission was necessary in order to reduce the non-specific background of the other antibodies. All blocking and antibody solutions were made in 1% BSA TBS with 0.1% Tween 20. Milk was not used in these experiments because it contains biotin, which could compete for the streptavidin-HRP while in solution, reducing sensitivity of the assay. Alternatively, streptavidin-HRP could bind to biotin adhered to the membrane, increasing non-specific background. The membrane was re-probed with anti-ICAM-1 antibody.

Intact cells labeled with biotin, lysed for a streptavidin pull down, and run in a Western

blot for ICAM-1: Biotin labeling and subsequent lysis of cells was carried out as described above. Lysate was incubated with 40µl of immobilized streptavidin (Pierce, cat # 20347) overnight at 4°C on a rotating platform. The mixture was centrifuged at

4,000 x g for 4 min and washed once with RIPA buffer and twice with PBS. The sample was eluted with boiling for 15 min in 40 µl 4X sample buffer and then it was run on the same gel as the samples above. After the samples were transferred to PVDF, the membrane was cut along the molecular weight marker. The half of the membrane

89 containing the above samples was as described. The remaining membrane was processed in a Western blot with anti-ICAM-1 antibody (1:500 dilution) followed by an anti-mouse-

HRP antibody and ECL incubation.

Statistical Analysis - All values and graphs report mean +/- standard error (S.E.).

Statistical analysis of differences between control values and sample values were determined via the Student’s T-test. Differences displaying p values < 0.05 were considered statistically significant.

Figure 2.1. Glutaredoxin activity is increased in retinal homogenates from hyperglycemic rats, according to a standard enzymatic radiolabel assay. Using the

[35S] BSA- SSG substrate, Grx activity of diabetic retinal homogenate was found to be

2.6 +/- 0.1 nmol/min/mg cellular protein compared to 1.1 +/- 0.1 nmol/min/mg cellular

protein in non-diabetic control rat retinae (n=3). *p < 0.001.

Figure 2.2. Grx activity is increased in rMC-1 cells cultured in high glucose (25 mM) medium, according to the disulfide reducing capacity. The cellular reduction rate of hydroxyl ethyl disulfide (HEDS) was 11.6 +/- 1.2 nmol thiol/mg protein/min in normal glucose (5 mM) treated cells and 19.0 +/-2.5 nmol thiol/mg protein/min in high glucose (25 mM) treated cells. The cellular reduction rate of lipoate gave rise to 4.2+/-

0.5 nmol thiol/mg protein/min in normal glucose and 4.7+/- 0.8 nmol thiol/mg protein/min in high glucose medium. (n=10) *p < 0.02.

Figure 2.3 Glutathione reductase (GRase) activity, Thioredoxin reductase (TRase) activity, and thiol content are decreased in rMC-1 cells cultured in high glucose (25 mM) medium. The rate of reduction of DTNB by TRase in lysates of rMC-1 cells cultured in either normal glucose (5 mM) or high glucose (25 mM) medium was monitored via A405 for 5 min. TRase activity in cells cultured in normal glucose medium (0.024 A405/µl/min) was decreased to 0.013 A405/µl/min in high glucose medium. The total thiol content in similar lysates was determined by the reduction of DTNB in an end point absorbance reading at A405. Thiol content of cells cultured in normal and high glucose medium was 0.051 A405/µl and 0.019 A405/µl, respectively. The rate of GSSG reduction by GRase in lysate was measured at A340 for 5 min. GRase activity in cells was 0.092 U/ml/min when cultured in normal glucose medium and 0.062 U/ml/min when cultured in high glucose medium. All experiments were carried out in a 96-well plate reader (n=4).

Figure 2.4. Grx1 protein content is increased in rMC-1 cells cultured in high glucose (25 mM) medium, according to Western blot analysis. After five days in normal (5 mM) or high (25 mM) glucose medium, rMC-1 cells were lysed and immunoprobed with anti-Grx1 (1:1,000 dilution). Anti-GAPDH (1:10,000 dilution) was used as a loading control. High glucose induced Grx1 expression by 2.3-fold (+/- 0.3)

(n=5). *p < 0.05.

Figure 2.5. Glutaredoxin activity is increased in rMC-1 cells cultured in high glucose (25 mM) medium, according to a standard enzymatic radiolabel assay. Grx activity in cellular lysate from either normal glucose (5 mM) or high glucose (25 mM) medium was assayed via the rate of removal of [35S] GSH from [35S] BSA- SSG. Grx

activity of cells grown in normal and high glucose medium was 0.37 nmol/min/mg (+/-

0.08) and 0.62 nmol/min/mg (+/-0.18) respectively (n=14) ∆p = 0.07.

Figure 2.6. ICAM-1 protein content is increased in rMC-1 cells cultured in high glucose (25 mM) medium, according to Western blot analysis. After five days in normal (5 mM) or high (25 mM) glucose medium, rMC-1 cells were lysed and immunoprobed with anti-ICAM-1 (1:500 dilution). Anti-GAPDH (1:10,000 dilution) was used as a loading control. High glucose induced Grx1 expression by 2.3-fold (+/-

0.3) (n=5). *p < 0.05.

Figure 2.7. Nuclear localization of the NFκB subunits p50 and p65 is increased in rMC-1 cells cultured in high glucose (25 mM) medium. After five days in normal (5 mM) or high (25 mM) glucose medium, rMC-1 cells, were separated into cytoplasmic and nuclear fractions and immunoprobed for anti-p50 (1:1,000 dilution) (A and B), anti- p65 (1:3,000 dilution) (A and C). Loading controls were actin (1:30,000 dilution) for the cytoplasm and YYl (1:1,000 dilution) for the nucleus. Nuclear p50 and p65 were increased 2.8-fold (+/- 0.6) and 1.8-fold (+/- 0.2) respectively, and cytoplasmic p50 and p65 were not significantly changed (n=5).

*p < 0.05.

Figure 2.8. Actin is expressed in the cytoplasm and nucleus of rMC-1 cells, and high glucose (25 mM) medium increases the nuclear content of GAPDH in rMC-1 cells. rMC-1 cells were cultured in normal (5 mM) or high (25 mM) glucose medium, fractionated into nuclear and cytoplasmic fractions, and probed on a Western blot with anti-GAPDH and anti-actin. Actin was found to similar extents in both the nucleus and cytoplasm of rMC-1 cells (A). GAPDH was primarily found in cytoplasmic fractions; however high glucose conditions led to apparent nuclear translocation of GAPDH in rMC-1 cells (A and B).

Figure 2.9. Glutaredoxin activity is increased in rMC-1 cells over-expressing Grx1, according to a standard enzymatic radiolabel assay. rMC-1 cells grown in normal glucose medium (5 mM) were infected with either Ad-Grx1 or Ad-Empty and assayed for Grx activity using the radiolabeled substrate [35S] BSA- SSG, Grx activity was

increased up to 13 nmol/min/mg (+/- 2.4) (n=10) after infection with Ad-Grx1, and was

unaffected by infection with Ad-Empty (n=5). *p < 0.01.

Figure 2.10. Grx1 and ICAM-1 are increased in rMC-1 cells over-expressing Grx1, according to Western blot analysis. rMC-1 cells grown in normal glucose (5 mM) medium and infected with Ad-Grx1 or Ad-Empty were lysed and immunoprobed with anti-Grx1 (1:1,000 dilution) (A and B) and anti-ICAM-1 (1:500 dilution) (C and D) antibodies. GAPDH was probed for a loading control, and quantification was based on the ratio of band intensities of Grx1 or ICAM-1 to GAPDH. Ad-Grx1 increased Grx1 content up to 15-fold (+/- 4.7) (n=4) with no effect for Ad-Empty (n=3). Ad-Grx1 increased ICAM-1 up to 6.8-fold (+/- 2.1) (n=6) with no effect observed after infection with Ad-Empty (n=4).*p < 0.05.

100

Figure 2.11. Expession of ICAM-1 is increased in rMC-1 cells over-expressing

Grx1. rMC-1 cells were transfected with Ad-Grx1 in normal glucose medium, and lysate was analyzed with an ICAM-1 ELISA. ICAM-1 was increased over 2-fold (+/- 0.3) in

Grx1 over expressing rMC-1 cells (n=4). Ad-Empty had no effect on ICAM-1 expression (n=4). *p < 0.04.

101

Figure 2.12. Nuclear localization of the NFκB subunit p50 is increased in rMC-1 cells over-expressing Grx1. rMC-1 cells grown in normal glucose medium (5 mM) and infected with Ad-Grx1 or Ad-Empty were separated into nuclear and cytoplasmic fractions and immunoprobed for anti-p50 (1:1,000 dilution). Actin (1:30,000 dilution) and YY1 (1:1,000 dilution) were probed as loading controls in the cytoplasm and nucleus, respectively, and quantification was based on the ratio of p50 to actin for cytosolic fractions and p50 to YY1 for nuclear fractions. Ad-Grx1 increased nuclear p50 by 5.5-fold (+/- 1.1), and Ad-empty had no effect (A and B) (n=6). Neither Ad-Grx1 nor

Ad-Empty had a significant effect on abundance of cytoplasmic p50 or p65 (C and D)

(n=5). *p < 0.05.

102

Figure 2.13. Nuclear localization of the NFκB subunit p65 is increased in rMC-1 cells over-expressing Grx1. rMC-1 cells grown in normal glucose medium (5 mM) and infected with Ad-Grx1 or Ad-Empty were separated into nuclear and cytoplasmic fractions, and immunoblotted for anti-p65 (1:3,000 dilution). Actin (1:30,000 dilution) and YY1 (1:1000 dilution) were probed as loading controls in the cytoplasm and nucleus, respectively, and quantification was based on the ratio as described in Figure 2.15. Ad-

Grx1 increased nuclear p65 by 3.9-fold (+/- 0.9), and Ad-Empty had no effect (A and B)

(n=6). Neither Ad-Grx1 nor Ad-Empty induced significant changes in cytoplasmic expression of p65 (C and D). *p < 0.05.

103

Figure 2.14. NFκB activity is increased in rMC-1 cells over-expressing Grx1, according to luciferase assays. rMC-1 cells co-transfected with NFκB-firefly-

luciferase and Renilla-luciferase plasmids were infected with Ad-Grx1 or Ad-Empty at

MOI 0-40 and assayed for NFκB activity with the Dual-Luciferase Reporter Assay

(Promega). NFκB luciferase activity was increased by 3.4 (+/- 0.4) and was unaffected

by Ad-Empty (n=5). *p < 0.01.

104

Figure 2.15. Inhibition of NFκB nuclear translocation with sn50 prevents ICAM-1 induction in rMC-1 cells over-expressing Grx1. rMC-1 cells infected with either Ad-

Grx1 (G10) or Ad-Empty (E10) at an MOI of 10 in the presence of sn50, in normal glucose (5 mM) medium were lysed and collected for immunoblotting of ICAM-1

(1:1000 dilution) and actin (1:30,000 dilution). ICAM-1 was increased by 2.5-fold (+/-

0.2) in cells infected with Ad-Grx1 (n=8). sn50 decreased ICAM-1 production in a dose dependent fashion in cells infected with Ad-Grx1. Cells infected with Ad-Grx1 in the presence of 20 μM sn50 had similar amounts of ICAM-1 as control cells (no adenovirus and Ad-Empty), and the amount of ICAM-1 was significantly different from cells infected with Ad-Grx1 in the absence of sn50 (n=8). *p < 0.0003 #p < 0.0006.

105

Figure 2.16. The induction of Grx1 and ICAM-1 is prevented in rMC-1 cells cultured in high glucose (25 mM) medium and knocked down for Grx1 using siRNA-Grx1. rMC-1 cells transfected with rat Grx1 siRNA or non-targeted siRNA in high glucose (25 mM) medium were lysed and immunoprobed for Grx1 (1:1,000 dilution) (A and C) and ICAM-1 (1:500 dilution) (B and C). Actin (1:30,000 dilution) and GAPDH (1:10,000 dilution) were probed as loading controls. Grx1 was decreased by 47% (+/- 4.9%) (n=9) and ICAM-1 was decreased by 33% (+/- 4.5%) (n=11). *p <

0.00002.

106

Figure 2.17. Detection of extracellular membrane-bound ICAM-1 in rMC-1 cells cells. The surface of rMC-1 cells, either adherent or trypsinized to detach them, were labeled with biotin, and the lysed cells were immunoprecipitated for ICAM-1 and run on an SDS PAGE gel. The proteins were transferred to a PVDF membrane and detected with streptavidin (A) or immunoblotted for ICAM-1 (B) (n=1). The arrows indicate a band at approximately 100 kDa that seems to be consistent with streptavidin-biotin-

ICAM-1. Alternatlively, the lysed cells were incubated with immobilized streptavidin.

Eluted proteins were separated and processed in a Western blot for ICAM-1. (n=1). The arrow indicates a band at approximately 100 kDa that seems to be consistent with streptavidin-biotin-ICAM-1.

107

Figure 2.18. Restriction digests of newly propagated plasmids were confirmed to run similar on an agarose gel to plasmids obtained originally. The newly propagated

NFκB-Firefly-luciferase plasmid was digested with PVU1 in parallel with the original

plasmid, and both were run on a 1% agarose gel (A). The newly propagated Renilla-

luciferase plasmid was digested with HINDIII in parallel with the original plasmid, and

both were run on a 1% agarose gel next to the undigested forms of each (B). All bands of

the new plasmids were consistent with the original plasmids.

108

Chapter 3: Glutaredoxin regulates ICAM-1 production and Interleukin-6 secretion

via IKK-SSG in retinal Müller cells, mediating autocrine and paracrine pro-

inflammatory responses

3.1 Abstract

Glutaredoxin (Grx1) plays a key role in redox regulation because it is a specific and

efficient catalyst of reversible glutathionylation. In Chapter 2, Grx1 was shown to be increased in retinae of diabetic rats and in rat retinal Müller glial cells (rMC-1) cultured

in diabetic-like conditions i.e., high glucose (25 mM) medium. This upregulation of

Grx1 was concomitant with NFκB activation and induction of ICAM-1 (intercellular

adhesion molecule-1). Moreover, an analogous pro-inflammatory response was observed

in rMC-1 cells over-expressing Grx1 via adenoviral-directed upregulation of Grx1 (Ad-

Grx1) in normal glucose (5 mM) medium. The site of regulation of NFκB was localized

to the cytoplasm, where IκB kinase (IKK) is an important regulator of NFκB activation.

In the current study, inhibition of IKK activity abrogated the increase in ICAM-1 induced

by high glucose or by Ad-Grx1. Consistent with this model, inhibition of IKK prevented

both the nuclear translocation of NFκB and increased ICAM-1 expression in Grx1 over-

expressing cells. Conditioned medium from the rMC-1 cells over-expressing Grx1 was

added to fresh cultures of rMC-1 cells and elicited increases in the Grx1 and ICAM-1

proteins (autocrine regulation). A similar induction of these proteins was observed when

conditioned medium from the rMC-1 cells over-expressing Grx1 was added to

endothelial cells (paracrine regulation). These effects correlate with a novel finding that

109

secretion of IL-6 was elevated in the cultures of Grx1 over-expressing cells. Thus, Grx1

appears to play an important role in pro-inflammatory responses. Furthermore,

IKKβ isolated from rMC-1 cells in normal glucose medium was found to be

glutathionylated on Cys179. Hence Grx1-mediated activation of IKK via

deglutathionylation may play a central role in diabetic complications in vivo where Grx1

is increased.

3.2 Introduction

The enzymatic process of reversing and thus regulating steady state concentrations of

protein-SSG has been well documented and is attributed to Grx1 (Chrestensen et al.,

2000;Gravina and Mieyal, 1993;Gallogly and Mieyal, 2007).

Redox sensitivity of cysteine residues on proteins in the upstream pathway leading to

NFκB-mediated transcription suggests that S-glutathionylation is an important regulatory

mechanism in this pathway (Shelton et al., 2005;Mieyal et al., 2008;Shelton and Mieyal,

2008). In Chapter 2 of this thesis, induction of Grx1 was shown in retinal homogenates

from streptozotocin-diabetic rats. Grx1 was also upregulated in retinal glial Müller

(rMC-1) cells cultured in diabetic-like high glucose (25 mM) conditions (Shelton et al.,

2007). Increased Grx1 expression in high glucose medium or Grx1 over-expression via

adenoviral-mediated infection (Ad-Grx1) in normal glucose medium corresponded with increased nuclear translocation of the redox sensitive transcription factor NFκB (p50-

p65) and expression of ICAM-1 (intercellular adhesion molecule-1), a transcriptional

110 product of NFκB. Furthermore, the site of NFκB regulation by Grx1 was indicated to reside in the cytoplasmic signaling pathway, prompting us to elucidate further the molecular mechanism of Grx1 regulation of NFκB activation. IκB kinase (IKK) is a central regulator of NFκB signaling in the cytoplasm, and the pro-inflammatory effects of

NFκB activation usually have been attributed to the IKKβ subunit specifically (Rossi et al., 2000;Jeon et al., 2003). Here we demonstrate that IKKβ is glutathionylated site- specifically (Cys179) in retinal rMC-1 cells, implicating it as a Grx1-regulated control point in NFκB-mediated ICAM-1 expression. Accordingly, inhibition of IKK abolished the induction of ICAM-1 expression by Ad-Grx1. Furthermore, conditioned medium from rMC-1 glial cells over-expressing Grx1 elicited an induction of ICAM-1 in freshly cultured rMC-1 glial and in endothelial TRiBRB cells, implicating Grx1 in autocrine and paracrine pro-inflammatory responses, respectively. Likewise conditioned medium from rMC-1 glial cells over-expressing Grx1 elicited an induction of Grx1 in freshly cultured rMC-1 glial and in endothelial TRiBRB cells, suggesting an auto-regulatory feedback mechanism for Grx1. Moreover, it was found that upregulation of Grx1 leads to increased secretion of the IL-6 cytokine, implicating IL-6 as a key contributor to the induction of Grx1 and ICAM-1. Overall, these findings implicate Grx1 and corresponding glutathionylation events in the pro-inflammatory progression of diabetic retinopathy.

3.3 Results

111

Inhibition of IKK prevents increases in ICAM-1 expression in rMC-1 cells cultured in

high glucose (25 mM) medium. The potential sites for glutathionylation in the cytosolic

pathway of NFκB are extensive as exemplified in Figure 1.3 (pg. 42). We utilized

inhibition of IKK, a point of signaling convergence, to determine whether the protein(s)

regulated by Grx1 are upstream, downstream, or directly at IKK. Specifically, we tested

changes in ICAM-1 expression in cells cultured in normal or high glucose medium in the

absence of presence of an inhibitor of IKK activity (Bay 11-7085). High glucose increased ICAM-1 expression nearly 2-fold, and this induction was blocked by increasing concentrations of Bay 11-7085 (Fig. 3.1, pg. 137), suggesting that IKK and/or upstream mediators are critical in the high glucose induction of NFκB-regulated transcription in

rMC-1 cells.

Inhibition of IKK prevents increases in ICAM-1 in rMC-1 cells over-expressing Grx1 in

normal glucose medium. To test the effects of Grx1 in the absence of other factors that

could be changed by high glucose, ICAM-1 production was assayed in cells with

inhibited IKK activity and over-expressed Grx1. Adenoviral Grx1 (Ad-Grx1)-mediated

over-expression of Grx1 in normal glucose medium increased ICAM-1 expression about

2.5-fold in rMC-1 cells. This extent of induction at this particular amount of infection

(MOI 10) agrees well with the observations in the earlier studies (Chapter 2, (Shelton et

al., 2007)). Since MOI 10 of Ad-Grx1 is the viral concentration that mimics the effect of

high glucose both with respect to induction of Grx1 and ICAM-1 (Shelton et al., 2007), it was used regularly in this phase of the studies. Next, the effect of IKK inhibition on

ICAM-1 production in the Grx1 over-expressing cells in normal glucose medium was

112

tested to determine whether this effect would mimic that seen in high glucose treated

cells. A concentration dependent inhibition with Bay 11-7085 was observed, indicating that inhibition of IKK activity prevented the ICAM-1 increase (Fig. 3.2, pg. 138). These data indicate that IKK is a critical control point in Grx1 over-expressing cells in normal glucose medium and also in cells cultured in high glucose medium.

Inhibition of IKK prevents the increased nuclear translocation of NFκB in rMC-1 cells

over-expressing Grx1 in normal glucose medium. To determine whether inhibition of

IKK corresponds to an inhibition of NFκB, we tested effects of Bay 11-7085 on the

nuclear translocation of the NFκB subunits, p50 and p65, in Grx1 over-expressing cells

in normal glucose medium. Over-expression of Grx1 induced nuclear translocation of

both NFκB subunits by 2 to 4-fold. These increases in NFκB in the nucleus were

prevented by Bay 11-7085 in a concentration-dependent manner (Fig. 3.3, A and B, pg.

139). The empty vector control had no effect on ICAM-1 expression (Fig. 3.3, A and B,

pg. 139). These results are consistent with NFκB being the predominant transcriptional

regulator of ICAM-1 in this cell system and with the site of regulation by Grx1 taking

place at and/or above IKK.

Analysis of the S-glutathionylation status of IKKβ immunoprecipitated from lysate of

rMC-1 cells. In order to isolate IKKβ from lysate of rMC-1 cells for subsequent analysis

of its glutathionylation status with mass spectrometry, an immunoprecipitation procedure

for IKKβ was developed, and immunoprecipitated IKKβ was analyzed by mass

spectrometry.

113

To determine whether IKKβ is glutathionylated in rMC-1 cells and to identify the specific residue(s) that undergoes this modification, cellular IKKβ was co- immunoprecipitated with IKKγ and analyzed via mass spectrometry. Western blot analysis of IKKβ isolated from rMC-1 cells cultured in normal glucose conditions confirmed successful pull-down of IKKβ (Fig. 3.4, A, pg. 140). Four individual immunoprecipitations (one 100 mm dish of cells per IP) were eluted, pooled, and digested with trypsin and then with proteinase K for analysis by mass spectrometry. A mass corresponding to the glutathionylated form of IKKβ was identified by peptide mass fingerprinting from MALDI-TOF mass spectral (MS) analysis. Subsequently, glutathionylation of the cysteine was confirmed by peptide sequencing with LC-MS-MS and data analysis by the Bioworks program. Of the 20 cysteine residues within

IKKβ, only Cys179 is located within the activation loop, and it was the site of S- glutathionylation detected in IKKβ immunoprecipitated from lysate of rMC-1 cells grown in normal glucose medium. The peptide sequences obtained from LC-MS-MS were searched against a rat database of protein sequences, and the search criteria included cysteines that were modified by IAM or a glutathionyl moiety, and serines, threonines, and tyrosines that were phosphorylated. Figure 3.4, B, (pg. 140) shows the MS/MS spectrum with the fragment ions of the peptide sequence labeled according to Biemann nomenclature. The presence of a glutathionylated peptide is documented by the loss of the cysteinyl-glutathione from the peptide to form the ionized fragment corresponding to the y3 peak, and this peptide contained the Cys179 residue (Fig. 3.4, B, pg. 140). This result suggests that Cys179 of endogenous IKKβ is modified by glutathionylation in cells under physiological conditions (i.e., cells that are unstimulated and not over-expressing

114

proteins). Hence, Grx1-mediated deglutathionylation and concomitant activation of

IKKβ is likely an important regulatory mechanism of NFκB-driven upregulation of

ICAM-1 in Müller cells under diabetic-like conditions.

Phosphorylation is well established for regulating the activity of IKK, and

glutathionylation has also been implicated in its regulation (Reynaert et al., 2006).

Glutathione has a net negative charge, suggesting that a phosphorylated protein may be resistant to simultaneous modification by glutathionylation on a nearby residue due to electrostatic repulsion, or vice versa. In addition to the peptide shown in the spectrum above, the simultaneous phosphorylation and glutathionylation of several other IKKβ peptide sequences containing Cys179 was indicated by the peptide matching in MS fingerprinting (Fig. 3.5, pg. 141). The peptides in this list were not detected in LC-MS-

MS, likely because this system often does not detect sequences of larger and/or less abundant peptides. The mass spectrometry analysis was conducted by Anne Distler,

Ph.D.

Conditioned medium from Grx1 over-expressing rMC-1 cells led to an increase in the expression of ICAM-1 and Grx1 in TRiBRB endothelial cells. To test whether Grx1- activation of the NFκB pathway upregulates expression of soluble mediators (e.g. cytokines) to an extent sufficient to induce an inflammatory response in other cells, we tested for effects of conditioned medium obtained from rMC-1 cells over-expressing

Grx1 on TRiBRB endothelial cells. ICAM-1 was upregulated in endothelial cells in a dose-dependent fashion in response to conditioned medium from rMC-1 cells over-

115

expressing Grx1 at two different concentrations (Ad-Grx1 at MOI 10 and MOI 20) (Fig.

3.6, pg. 142). The empty control vector construct did not induce expression of ICAM-1

(Fig. 3.6, pg.142). These data indicate that increased expression of Grx1 in Müller cells

leads to an increase in release of a signaling mediator capable of inducing a pro-

inflammatory paracrine response in endothelial cells.

To test whether the conditioned medium led to a corresponding increase in Grx1,

Western blot analysis was run on TRiBRB endothelial cells that had been cultured in the conditioned medium of rMC-1 cells over-expressing Grx1. Application of conditioned medium upregulated Grx1 in a similar dose-dependent manner in response to conditioned medium from rMC-1 cells over-expressing Grx1 at MOI 10 or MOI 20 (Fig. 3.6, pg.142).

The empty vector control did not induce expression of Grx1 (Fig. 3.6, pg.142). These data suggest that increased expression of Grx1 in rMC-1 cells induces NFκB-driven

production of cytokines that activate Grx1 in endothelial cells, leading to increased NFκB

activation and subsequent expression of ICAM-1.

Conditioned medium from Grx1 over-expressing rMC-1 cells led to increased expression

of ICAM-1 and Grx1 in rMC-1 cells. To test whether the signaling mediator(s) in the

conditioned medium could act in a positive feedback mechanism in Müller cells

analogous to the paracrine event observed in the TRiBRB endothelial cells, Western blot

analysis was run on ICAM-1 and Grx1 in rMC-1 cells treated for 24 hr with conditioned

medium from rMC-1 cells over-expressing Grx1. ICAM-1 and Grx1 were both increased

in rMC-1 cells in response to conditioned medium from rMC-1 cells over-expressing

116

Grx1 at two different concentrations of Grx1 (Ad-Grx1 at MOI 10 and MOI 20),

consistent with autocrine regulation (Fig. 3.7, pg. 143). Conditioned medium from rMC-

1 cells over-expressing Ad-Empty did not cause an upregulation of ICAM-1 or Grx1

(Fig. 3.7, pg. 143).

Cytokine analysis in conditioned medium from rMC-1 cells in which Grx1 was elevated;

IL-6 secretion is induced by Grx1. We set out to identify the soluble mediator(s) in the conditioned medium of rMC-1 cells that leads to upregulation of ICAM-1 and Grx1 in fresh cultures of Müller rMC-1 or endothelial cells. Target proteins to be screened were chosen according to their potential pro- or anti-inflammatory roles in diabetic retinopathy and/or retinal Müller cells. TNF-α, IL1-β, IFNγ, and IL-10 were undetected in the conditioned medium of rMC-1 cells over-expressing Grx1 (Table 3.1, pg. 144) or in the conditioned medium of the cells in high glucose medium. While VEGF could be detected in the medium, upregulation of Grx1 (Table 3.1, pg. 144) or incubation in high glucose medium did not elicit an increase in VEGF secretion. In contrast, a 3.7-fold increase in IL-6 expression was observed in medium from rMC-1 cells over-expressing

Grx1 (Ad-Grx1, MOI 10) (Fig. 3.8, A, pg. 145).

Having identified IL-6 in the medium, we tested directly the effect of pure IL-6 on rMC-1 cells. At concentrations as low as 60 pg/ml, approximately a 2-fold induction of Grx1 and a 2.5-fold induction of ICAM-1 were observed. In addition, in separate experiments it was confirmed that IL-6 was increased also in the medium of rMC-1 cells cultured in high glucose (Fig. 3.8, B, pg. 145); however the apparent fold-increase in IL-6 was

117

substantially less that of the Grx1 over-expressing cells due at least in part to incomplete

recovery of IL-6 during the more extensive processing of the medium that was required

(see below). Collectively, these data suggest that Grx1 leads to activation of NFκB via

deglutathionylation of proteins in the cytosolic signaling pathway (e.g. IKK), leading to increased expression and subsequent secretion of IL-6 in Müller cells.

Recovery of recombinant IL-6 in concentrated medium- The IL-6 in the medium of Grx1

over-expressing cells was more readily detected than that from cells cultured in high

glucose. However, there were differences in the experimental designs. For example, normal and high glucose cultured cells were grown for an extended time in culture and required a larger volume of medium to avoid glucose depletion. Also, the normal and high glucose medium contained 2% FBS whereas 0.67% FBS was present in the normal glucose medium used for Grx1 over-expressing cells contained. Therefore, the medium in the extended experiments comparing the effects of normal and high glucose medium

(2% FBS) had to be concentrated to a greater extent (20-fold more) than medium (0.67%

FBS) from Grx1 over-expressing cells. To test whether these differences interfere with the ability to detect IL-6, media containing 2% or 0.67% FBS were spiked with 75 pg recombinant IL-6 and concentrated 3-fold or 60-fold to a final volume of 0.5 ml. In medium containing 2% FBS or 0.67% FBS, a 3-fold concentration gave rise to greater yield than that from a 60-fold concentration (approx. 23% recovery) (Fig. 3.9, pg. 146).

Medium containing 2% FBS and concentrated 3-fold gave rise to a 42% recovery, and medium containing 0.67% FBS and concentrated 3-fold gave rise to 68% recovery.

118

Intracellular and extracellular IL-1β in rMC-1 cells cultured normal or high glucose-.

Consistent with the above results on the ELISAs of conditioned medium from rMC-1 cells, a band at 15-18 kDa corresponding to IL-1β was not found in conditioned medium or in cell lysate in Western blot analysis of rMC-1 cells with an anti-IL-1β antibody.

However, the conditioned medium did give rise to an immunoreactive band between 50 kDa and 75 kDa (n=3), and this band was detected whether or not actin could be detected in the same samples. The presence of actin distinguishes between cellular lysis and protein secretion as the source of proteins found in the medium. The band at this unexpected molecular weight could either correspond to a dimer of the IL-1β precursor

(mw 30-35 kDa) or a non-specific protein. It is unlikely that a protein that was undetected in ELISA assays would be discovered with Western blot analysis because the

ELISA is about 30,000-fold more sensitive (see Ch. 3 Methods). However, different sensitivities in the two systems is possible if the antibodies are designed to interact with different epitopes, but the likelihood of this is diminished by the fact that they are both purchased from the same company. Nevertheless, the active form of IL-1β was not detected in the conditioned medium via Western blot or ELISA or in the lysate itself.

Therefore, if IL-1β is a key player in the pathology of Müller cells in high glucose, it does so at concentrations below the sensitivity threshold of available assays (detection limit: < 0.035 pg/µl). Alternatively, IL-1β could be lost in the processing of the sample.

Intracellular and extracellular TNFα in rMC-1 cells cultured in normal or high glucose,

or over-expressing Grx1- TNFα was analyzed in a blinded fashion by the Inflammatory

Core (CWRU) via a Fluorokine MultiAnalyte Profiling (FMAP) cytokine multiplex kit

119

(see above), in a blinded fashion by the Center For Aids Research (CFAR; CWRU) via

ELISA assays, and in an un-blinded fashion via Western blot analysis. TNFα was not detected in lysates or medium of Grx1 over-expressing rMC-1 cells (MOI 0-80 Ad-Grx1)

assayed with ELISA (detection limit: <12.5 pg/ml). TNFα was not detected also in

medium of Grx1 over-expressing rMC-1 cells (MOI 0-20 Ad-Grx1) assayed with FMAP

which has a detection limit of 244 pg/ml. Anti-TNFα Western blots gave robust signals

for recombinant TNFα protein standards (1-32 ng), but did not detect TNFα in lysates.

Collectively, these data document that TNFα is not produced and/or secreted from rMC-1

cells at detectable levels (>0.01 pg/ µl).

3.4 Discussion

Grx1 regulation of the production of proteins that are transcriptionally regulated by NFκB (e.g.

cytokines)- Grx1 is an important cytosolic thiol redox enzyme that regulates reversible S- glutathionylation and corresponding activity of cellular proteins (Chrestensen et al.,

2000;Shelton et al., 2005;Mieyal et al., 2008;Shelton and Mieyal, 2008). Changes in

Grx1 have been reported previously to affect the amounts of pro-inflammatory chemokines secreted from primary tracheal epithelium cells, namely macrophage inflammatory protein-2 and keratinocyte-derived chemokine (Reynaert et al., 2006). The current study shows that upregulation of Grx1 correspondingly leads to upregulation of

ICAM-1 (Fig. 3.2, pg. 138, Chapter 2, (Shelton et al., 2007)) and IL-6 in rMC-1 cells

(Fig. 3.8, A, pg. 145 and Table 3.1, pg. 144). Interestingly, the induction of NFκB transcriptional gene products appears not to be pleiotropic in rMC-1 cells over-expressing

120

Grx1. Detectable amounts of TNFα, IL-10, IL1-β, or IFN-γ were not found in the

medium of rMC-1 cells cultured in normal or high glucose medium, or infected with Ad-

Grx1 (Table 3.1, pg. 144). Moreover, VEGF was released at substantial amounts by

rMC-1 cells, but its secretion was not induced by over-expression of Grx1 (Table 3.1, pg.

144) or by high glucose. Therefore IL-6 and ICAM-1 appear to be selective targets for

Grx1-mediated NFκB activation in these cells. However, the possibility that Grx1 induces the expression and secretion of additional NFκB-regulated inflammatory

mediators (e.g., IL-8) in Müller cells has not been ruled out (see Future Directions). The

following is a discussion on the rationale for choosing the particular cytokines assayed in

the current study and the relationship of the current data with the literature.

Effects of high glucose or viral-mediated elevations in Grx1 on TNFα in Müller cells -

TNFα is increased in the whole retina after 1 week of hyperglycemia (Joussen et al.,

2002) but not detected at six months (Gerhardinger et al., 2005). Despite the apparent limitation in the time course of TNFα expression indicated by these studies, TNFα is a prototypical pro-inflammatory cytokine and a likely candidate for regulation in the retinal

Müller cells. The mRNA and protein content of TNFα was previously reported to be increased in rMC-1 cells in response to high glucose, but changes in secretion of TNFα in retinal Müller cells in response to either high glucose or over-expression of Grx1 had not been addressed previously. Our findings suggest that if TNFα is a key player in the regulatory mechanisms of Ad-Grx1 and high glucose, it must do so at concentrations below the level of detectibility in these assays used here (see Methods).

121

Effects of high glucose or viral-mediated elevations in Grx1 on VEGF in retinal Müller

cells- VEGF is a well known angiogenic factor that promotes undesirable

neovascularization in diabetic retinopathy. It has been reported to have both protective

and harmful effects on the neuronal retina, and it seems to prevent osmotic swelling of

retinal Müller cells which contributes to undesirable edema (Wurm et al., 2008). Müller

cells from diabetic rats have enhanced gene expression of VEGF (Gerhardinger et al.,

2005). Furthermore, expression of VEGF has been found previously within retinal

Müller cells (Famiglietti et al., 2003;Gerhardinger et al., 2005), and its secretion from

these cells has been documented as well (Eichler et al., 2004). However, to date, a

change in secretion from these cells in response to high glucose or diabetes has not been

documented. Consistent with previous findings, our studies show that Müller cells

express and release substantial amounts of VEGF, but this basal secretion was not altered

in response to Ad-Grx1 or high glucose (Table 3.1, pg. 144). VEGF clearly has

important roles within the retina, and other cells are likely to have increased secretion of

VEGF in diabetes or high glucose conditions. However, it does not seem to mediate the effects of Müller cells on surrounding cells in response to high glucose and elevated

Grx1.

Effects of high glucose or viral-mediated elevations in Grx1 on IL-1β in retinal Müller

cells- IL-1β is a prototypical pro-inflammatory cytokine. It was predicted to be a central

signaling molecule in rMC-1 cells in response to high glucose medium or Ad-Grx1

because inhibition of IL-1β signaling with a neutralizing antibody diminishes the high

glucose-induced activation of caspase-1 and caspase-3 in rMC-1 cells (Vincent and

122

Mohr, 2007), suggesting that high glucose increased the secretion of IL-1β which activated the caspase proteins. That study showed that the caspase-dependent inhibition of IL-1β blocked retinal capillary degeneration in diabetic mice (Vincent and Mohr,

2007). In addition, both the mRNA and protein expression of IL-1β have been reported to be increased in rMC-1 cells in response to high glucose (Walker and Steinle, 2007).

However, IL-1β in the medium of rMC-1 cells was not tested. Similar to the previous studies evaluating TNFα, the study of IL-1β by Walker et al. reported serum starvation of cells for up to 24 hr, and differences such as this may account for their findings being different from ours. Vincent et al. did not report a serum starvation step in their methods, but IL-1β was not measured directly. Therefore, the concentration of secreted IL-1β that led to caspase activation in that study may be lower than the limit of detectibility of the assays used in the present study.

Effects of Grx1 high glucose or viral-mediated elevations in Grx1 on IL-6- IL-6 has both

pro-inflammatory and anti-inflammatory roles. IL-6 has been reported to be increased in

animal models of diabetes and in diabetic patients (Abu el Asrar et al.,

1992;Murugeswari et al., 2008;Kauffmann et al., 1994;Gustavsson et al.,

2008;Srinivasan et al., 2004;Mysliwiec et al., 2008). In addition, purified IL-6 protein applied to cultured cells has been reported to slightly increase Grx1 protein expression within the cell (Takashima et al., 1999). Our study found that upregulation of Grx1

substantially elevated the secretion of IL-6 in rMC-1 cells. Additional analysis is needed to confirm a similar extent of enhancement of IL-6 secretion from rMC-1 cells cultured in

high glucose. However, Grx1 is upregulated in high glucose of rMC-1 cells to about the

123 same extent as that elicited by MOI 10 Ad-Grx. Therefore, it is unlikely that the cells in high glucose would fail to replicate the extent of IL-6 induction seen with Ad-Grx1 infections. Overall, this is a novel finding, and it suggests that IL-6 is secreted from the

Müller cells in diabetic retinopathy, and that Grx1 regulates this secretion.

Effects of Grx1 over-expression or high glucose on IFNγ and IL-10 –IFNγ and IL-10 were the cytokines least expected to be upregulated by Grx1 and high glucose medium.

Interferons are cytokines that regulate various cellular functions. IFNγ is involved in anti-viral responses and anti-angiogenesis. In addition, it has been reported to be upregulated by NFκB in T-cells (Shin et al., 2006). IL-10 is a prototypical anti- inflammatory cytokine, and has an inhibitory effect on the production of many pro- inflammatory cytokines and chemokines (Couper et al., 2008). If IFNγ or IL-10 had been upregulated by Ad-Grx1 or high glucose, potential beneficial effects could have been attributed to the upregulation of Grx1. Neither of these cytokines was detected, however.

Autocrine and paracrine mechanisms of Grx1 regulation - Conditioned medium (IL-6) from glial (rMC-1) cells over-expressing Grx1 led to upregulation of Grx1 in other rMC-

1 cells (autocrine regulation) and in endothelial (TRiBRB) cells (paracrine regulation)

(Figs. 3.6, pg. 142 and 3.7, pg. 143). These findings are distinguished from previous reports on cytokine regulation of Grx1 because Grx1 in this case is shown to be induced by an endogenous protein secreted from cells. The previous reports involved exogenous application of purified proteins to cultured cells. For example, pure IL-6 led to a robust increase in Grx1 mRNA, but was reported to elicit only a slight increase in Grx1 protein

124

in a monocytic cell line (Takashima et al., 1999). In another study, treatment with

IFNγ was reported to increase mRNA and activity as well as Grx1 protein according to

immunohistochemical staining in airway epithelial cells (Reynaert et al., 2007). Grx1

protein expression was reported to be decreased in response to TGFβ1 in A549 cells

(Peltoniemi et al., 2004), but a change was not detected in mouse airway epithelium cells

(Reynaert et al., 2007). Other cytokines such as IL-4, IL-13, and TNFα have been

reported to change Grx-like activity and/or protein glutathionylation, but a change in

Grx1 protein either was not tested or not detected (Reynaert et al., 2007;Pan and Berk,

2007;Mukherjee et al., 2007). Clearly further studies are necessary to determine the

scope and the mechanism(s) of cytokine-mediated regulation of Grx1.

Grx1 regulation of the NFκB pathway and S-glutathionylation of IKKβ - A series of

cytoplasmic proteins initiate nuclear NFκB activation through the canonical NFκB signaling pathway. While these components represent a plethora of potential target sites for regulation by glutathionylation in the cytoplasm, signals from multiple upstream mediators converge in the cytoplasm on IKK.

IKK is a 700-900 kDa heteromeric complex comprised of a core enzyme of two catalytic subunits (IKKα and IKKβ) and at least one regulatory subunit (IKKγ). The additional composition of the IKK complex is not clear and is likely to be cell type dependent. For example, the oligomeric state of IKKγ in the holocomplex has been proposed to be a monomer, trimer, and tetramer (Scheidereit, 2006). Furthermore, IKK has been reported

to interact with at least 65 individual proteins in different contexts (Scheidereit, 2006).

125

The redox sensitive IKKβ is best characterized for mediating NFκB activation from pro- inflammatory stimuli (Rossi et al., 2000;Jeon et al., 2003), and has the highest catalytic activity towards IκB (Scheidereit, 2006). Glutathionylation of IKKβ has been shown to regulate NFκB activity and subsequent expression of keratinocyte-derived chemokine and macrophage inflammatory protein 2 in airway epithelial cells (Reynaert et al., 2006).

Moreover, cells knocked down in Grx1 have decreased expression of these chemokines.

These data suggest that Grx1 catalysis of IKKβ deglutathionylation leads to activation of

NFκB and subsequent expression of cytokines.

Regulation of the NFκB pathway by Grx1 has been implicated in several contexts including diabetes (Pineda-Molina et al., 2001;Qanungo et al., 2007;Shelton et al.,

2007;Reynaert et al., 2006). The current study indicates that IKKβ is a critical point of regulation of the NFκB pathway in rMC-1 cells exposed to high glucose. Since the rat sequence database for the IKKα subunit is not available, analysis could not be conducted for the glutathionylation on IKKα. It is possible that the IKKα subunit is glutathionylated but since IKKβ seems to play a dominant role in the activation of the

NFκB pathway it is likely a critical site of regulation. Moreover, this is a novel documentation of the S-glutathionylation of endogenous IKK under physiological conditions, i.e, IKK isolated from untreated and non-over-expressing cells (Fig. 3.4, pg.

140). These data suggest that S-glutathionylation of IKKβ regulates the activity and downstream signaling of IKK and is involved in the inflammatory response of Müller cells in diabetic retinopathy.

126

Impact of IL-6 and Grx1 in Diabetic Retinopathy- Previous studies have suggested a

neuroprotective role in retinal microglial cells for IL-6 in response to hydrostatic pressure

and ischemia (Sappington and Calkins, 2006;Sanchez et al., 2003) and specifically in rat

Müller cells in response to pituitary adenylate cyclase-activating peptide (Seki et al.,

2006). Despite the potential for beneficial effects of IL-6 in Müller cells in response to

these stimuli, IL-6 has been associated with deleterious pro-inflammatory effects in

diabetic retinopathy. IL-6 is increased in the vitreous humor of patients with diabetic

retinopathy (Abu el Asrar et al., 1992;Murugeswari et al., 2008;Kauffmann et al., 1994),

in retinas from diabetic mice and rats (Gustavsson et al., 2008;Srinivasan et al., 2004), and in serum from children with diabetic retinopathy (Mysliwiec et al., 2008). The current study implicates IL-6 for induction of ICAM-1 expression in Müller cells and endothelial cells, and these findings are consistent with an overall pro-inflammatory effect. IL-6 is likely a key contributor in the initiation of the inflammatory response in the diabetic retina. We show here that Grx1 regulates expression of IL-6 and is thus an important component of this response.

Grx1 has been reported to be increased in the hearts of diabetic rats previously (Li et al.,

2005), and the current study shows that Grx1 is upregulated in retinae of diabetic rats as well as retinal Müller cells in response to high glucose medium (Chapter 2, (Shelton et al., 2007)). Our findings are consistent with Grx1 regulation of NFκB activation and subsequent expression of ICAM-1 and IL-6 in retinal Müller cells via IKKβ-SSG. The

NFκB pathway has been scrutinized for therapeutic potential in diabetes for many years

(Cameron and Cotter, 2008;Shoelson et al., 2003), but effective intervention have yet to

127 be developed. In this context, Grx1 may emerge as a new therapeutic target for diabetic retinopathy.

3.5 Materials and Methods

Cell Culture - Cell culture supplies were obtained from Invitrogen except where indicated. Rat retinal glial (Müller) cells (rMC-1) were a kind gift from Dr. Vijay Sarthy

(Northwestern University, IL). Cells were cultured for up to five days in high glucose

(25 mM) or normal glucose (5 mM) in DMEM with 2% heat-inactivated FBS (Fisher,

Cellgro MT) with daily medium replacement in a humidified 37°C incubator with 5%

CO2 as described in Chapter 2. Glucose concentrations in the medium were monitored via the glucose oxidase kit as instructed by the manufacturer (Pointe Scientific). A transformed rat endothelial cell line isolated from the blood retinal barrier of transgenic mice (TRiBRB) was a kind gift from Dr. Tetsuya Terasaki (Hosoya et al., 2001). The use of this cell line was possible via a transfer agreement between Dr. Timothy Kern and

Dr. Tetsuya Terasaki. These cells were cultured in 5 mM glucose DMEM with 2% heat inactivated FBS and 15 µg/ml endothelial cell growth supplement (ECGS) (Sigma) on

0.1% gelatin coated plates in a humidified incubator with 5% CO2 at 33°C.

Adenoviral expression of Grx1 in rMC-1 Cells - rMC-1 cells (500,000 cells in a 100 mm dish) were grown in normal glucose medium for two days, infected with adenoviral vector containing a construct for expressing Grx1 (Ad-Grx1) or an empty vector control

(Ad-Empty) at a multiplicity of infection of 10 (MOI 10), except where indicated

128

otherwise. The adenoviral infections were carried out in 1 ml serum-free DMEM for one hour, as described in Chapter 2. Cells were cultured for two subsequent days in normal glucose medium and collected in 1% NP40 lysis buffer 50 mM Tris pH 8, 1% Nonidet

P40 (NP40) detergent, and 150 mM NaCl).

Inhibition of IKK with Bay 11-7085 in Grx1 over-expressing rMC-1 cells - Cells treated with Bay 11-7085 inhibitor (Biomol Int.) were pre-incubated with the inhibitor for 30-40 min in normal glucose medium and subsequently infected with Ad-Grx1 or Ad-Empty in the absence of the inhibitor. These cells were then cultured for an additional 24-48 hr in normal glucose medium in the presence of inhibitor, and collected in 1% NP40 lysis buffer.

Inhibition of IKK with Bay 11-7085 in rMC-1 cells in high glucose - Cells were cultured in high glucose (25 mM) medium for 2-3 days, treated with Bay 11-7085 for 5-10 min, and grown for an additional 2-3 days in high glucose medium. These cells were treated for less time with the inhibitor than those over-expressing Grx1 because the cells had to be less confluent to grow over the course of five days in culture. An incubation of 30-40 min with the inhibitor seemed to be cytotoxic on cells that were less confluent. All cells were grown for a total of five days in high glucose medium and collected in 1% NP40 lysis buffer.

Nuclear Extraction - Samples were separated into nuclear and cytoplasmic fractions as described in Chapter 2. rMC-1 cells from one 100 mm dish were collected in 1 ml of

129

phosphate-buffered saline, centrifuged for 3 min at 800 x g, and lysed in 300 μl of low salt buffer (20 mM HEPES, pH 7.6, 20% glycerol (v/v), 10 mM NaCl, 1.5 mM MgCl2,

0.2 mM EDTA, and 0.1% (v/v) Triton X-100) for 20 min. Centrifugation at 800 x g for 3 min yielded a cytosolic supernatant. The nuclear pellet was washed twice in phosphate- buffered saline, incubated in 80 μl of high salt buffer (10 mM HEPES, pH 7.6, 10% glycerol (v/v), 0.5 M NaCl, 0.7 mM MgCl2, 0.1 mM EDTA, and 0.05% (v/v) Triton X-

100) for 30 min at 4°C, and centrifuged at 16,000 x g for 15 min in 4°C.

Co-Immunoprecipitation of IKKβ - rMC-1 cells (100 mm dish, confluent) were collected

and lysed in 500-600 μl 1% Triton lysis buffer (1% Triton X-100, 250 mM Tris-HCl pH

8, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 μg/ml aprotinin, and

10 μg/ml leupeptin) for 10-15 min. Cells were lysed in 5 mM iodoacetamide (IAM), a

thiol-blocking reagent, to prevent artifactual glutathione-protein oxidation or thiol

exchange during processing. Five μl rabbit anti-IKKγ (0.2 μg/μl, Santa Cruz

Biotechnology, cat # sc8330) antibody and 40 μl of a 1:2 slurry of Protein A/G PLUS

agarose beads (Santa Cruz Biotechnology, cat # sc-2003) were added to cell lysates

(approx. 1-2 mg), and incubated on a rotating platform at 4°C overnight. For a

nonspecific immunoprecipitation control, samples were incubated with 1 μg Rabbit IgG

(Santa Cruz Biotechnology). Beads were centrifuged for 3 min at 3,000 x g, washed

twice with PBS, and boiled in 1X SDS sample buffer (0.5 M Tris-HCL pH 6.8, 20%

glycerol (v/v), 10% SDS (w/v), 1% bromophenol blue) for 15 min. Alternatively, the

immobilized IKKβ was used for [14C]-IAM labeling experiments (see below), or eluted

with a solution of trifluoroacetic acid and acetonitrile for mass spectral analysis (see

130 below). Samples for immunoblotting were processed by 12% SDS-PAGE and immunoblotted for mouse anti-IKKβ (Upstate, 05-535; 1:1000 dilution) antibody, and detected with goat anti-mouse HRP (1:10,000 dilution) antibody and chemiluminescence

(Western Lightning Kit) (see below for a more details on Western blot analysis).

Immunoprecipitated IKKβ was compared to that in the cell lysate. Recombinant IKKβ

(His-tagged, Calbiochem cat # 481404) (10 ng, 50 ng, and 100 ng) was used as standards for semi-quantitative analysis of Western blot. Optimization of the immunoprecipitation of IKKβ is discussed in the Appendix.

Detection of S-glutathionylation of IKKβ by Mass Spectrometry - IKKβ was immunoprecipitated from 500-600 µl of lysate from a 100 mm dish of rMC-1 cells cultured in normal glucose medium as described above. The beads were eluted with 40

µl 5% (v/v) trifluoroacetic acid (TFA) and 47.5% (v/v) acetonitrile. The eluates from four separate immunoprecipitations were combined and lyophilized. The sample was reconstituted in 50 μl 100 mM ammonium bicarbonate and digested with approximately

0.1 μg trypsin/15 μl in 25 mM ammonium bicarbonate over night at room temperature.

Each sample was further digested with 5 μg proteinase K to produce smaller molecular weight peptides. The sample was subsequently processed on a ThermoElectron LTQ linear quadrupole ion trap mass spectrometer coupled to a GE Healthcare Ettan MDLC liquid chromatograph. The mobile phases for LC were A) 0.1% formic acid (v/v) in

HPLC grade water and B) 0.1% formic acid (v/v) in HPLC grade acetonitrile. The gradient elution was run from 5% to 65% B over 50 min, and that of A was run simultaneously from 95% to 35%. All mass spectral data files were searched against a rat

131 protein subset taken from the NR database using the ThermoElectron Bioworks 3.3 search program. In particular the fragmentation patterns for peptides containing cysteine residues were additionally searched for specific masses (m/z) corresponding to addition of the glutathionyl moiety.

Western Blotting - Protein content of samples from whole cell lysates, nuclear extracts, or immunoprecipitations was determined via the Micro- bicinchroninic acid method (BCA)

(Pierce), according to the manufacturer’s protocol. For Western blot analysis, lysates

(100 μg) were mixed 4:1 with 4X SDS sample buffer (0.5 M Tris-HCl pH 6.8, 20% glycerol (v/v), 10% SDS (w/v), 1% bromophenol blue and in all studies not directed towards detecting the S-glutathionylation of IKKβ, 20 mM dithiothreitol (DTT)), heated for 15 min at 95°C, separated by 12% SDS-PAGE, and transferred to Immobilon-P polyvinyldifluoride (PVDF) membranes (Millipore) for up to 16 hrs at 0.1 Amp at 4°C.

Membranes were immunoprobed overnight at 4°C with the respective antibody in 5% milk, and the dilutions of the antibodies are indicated in parenthesis: anti-Grx1 (1:1,000)

(generated and purified via an adaptation of the McKinney and Parkinson caprylic acid method (Gravina, 1993), anti-p50 (1:1,000) (ab7971 Ab Cam, Cambridge, MA), anti-p65

(1:3,000) (sc372, Santa Cruz Biotechnology), anti-ICAM-1 (1:500) (R & D Systems), anti-actin (1:30,000) (Sigma). Peroxidase conjugated secondary goat anti-rabbit or anti- mouse antibodies (1:10,000) (Jackson ImmunoResearch Laboratories) were used, and

Western Lightning Chemiluminescence Reagent Plus (Perkin Elmer Life Sciences) was used according to the manufacturer’s protocol (Perkin Elmer Life Sciences). Relative band intensities were quantified using a BioRad Calibrated Imaging Densitometer GS-

132

710 with BioRad Quantity One software version 4.1.1. Changes in band intensity are reported as ratios relative to loading controls.

Treatment of rMC-1 cells with conditioned medium from Grx1 over-expressing rMC-1 cells – 400,000-600,000 rMC-1 cells/well in a 6-well dish were infected with Ad-Empty or Ad-Grx1 at MOI 10 or MOI 20 or uninfected and cultured for 24 hr in 1.5 ml medium that contained 0.67% heat inactivated FBS and 5 mM glucose. The conditioned medium was supplemented with glucose to re-establish a concentration of 5 mM and an additional

0.5 ml medium that contained 2% heat inactivated FBS and 5 mM glucose. The medium was then passed through a 0.22 μm syringe filter to remove cells and maintain sterility.

The medium was then applied to freshly cultured rMC-1 cells (200,000-400,000 cells/well in a 6-well dish) and the cells were cultured in this medium for 24 hr. Cells were lysed in 1% NP40 buffer and processed for Western blot analysis.

Treatment of TRiBRB endothelial cells with conditioned medium from Grx1 over- expressing rMC-1 cells – 200,000-300,000 rMC-1 cells/well of a 6-well dish were infected with Ad-Empty or Ad-Grx1 at MOI 10 or MOI 20 or uninfected and cultured for

24 hr in 1.5 ml medium that contained 0.67% heat inactivated FBS, and 5 mM glucose.

The conditioned medium was supplemented with glucose to re-establish a concentration of 5 mM, 20 µl of 15 mg/ml ECGS, and 0.5 ml medium that contained 2% heat inactivated FBS and 5 mM glucose. The medium was then passed through a 0.22 μm syringe filter to remove cells and to maintain sterility. The conditioned medium was then placed on roughly 200,000-300,000 TRiBRB endothelial cells/well of a 6-well dish, and

133

the cells were cultured in this medium for 24 hr. Cells were lysed in 1% NP40 buffer,

and processed for Western blot analysis.

Analysis of cytokines in the medium of rMC-1 cells grown in normal or high glucose

medium, or over-expressing Grx1 in normal glucose medium - rMC-1 cells (750,000 cells

in a 150 mm dish) were plated for a 3-4 day culture in normal or high glucose medium

(30 ml per dish) containing 2% heat inactivated FBS. Alternatively, 6 million rMC-1

cells in a 100 mm dish were infected with Ad-Grx1 or Ad-Empty (both MOI 10) in

normal glucose medium (8 ml per dish) and grown for 16-24 hr in medium containing

0.67% heat inactivated FBS and 5 mM glucose. To clear whole cells and cell debris, the

medium from each type of experiment was passed through a 0.22 μm syringe filter.

According to their initial volumes and cell density, medium from normal or high glucose

was concentrated 60-fold, and medium from adenoviral infected cell cultures was concentrated 3-fold with Amicon Ultra centrifugal filter devices (mw cutoff 10,000 Da).

Concentrated medium was analyzed via a Fluorokine MultiAnalyte Profiling (FMAP)

cytokine multiplex kit, a rat IL-6 Quantikine Immunoassay, and a rat VEGF

Immunoassay according the manufacturer’s instructions (R & D Systems). Fifty micoliters of sample were assayed in the FMAP analysis, and limit of detectibility for

each cytokine is as follows: 35 pg/ml IL-1β, 244 pg/ml TNFα, 24pg/ml IL-10, 398

pg/ml IL-6, and 354 pg/ml IFNγ. The ELISAs detected IL-6 at concentrations of 62.5

pg/ml and VEGF at concentrations of 31.2 pg/ml.

134

Treatment of rMC-1 cells with pure IL-6 - 150,000-300,000 rMC-1 cells in wells of a 6-

well dish were cultured for 16-24 hr. Recombinant IL-6 (60-8000 pg/ml) (R & D

Systems) was added to 1-2 ml normal glucose medium, passed through a 0.22 μm syringe filter, and applied to the cells. After 24 hr of treatment, the cells were lysed in 1% NP40 buffer and processed for immunoblotting with anti-Grx1 and anti-ICAM-1 antibodies.

Western blot analysis of IL-1β in lysate and medium of rMC-1 cells - 0.075-0.1 million

rMC-1 cells were plated in wells of 6-well dishes, and cultured in 2 ml normal or high

glucose medium for 4-5 days. Lysate (63-200 µg) and unconcentrated conditioned

medium (40 µl, 50 µg, or 65 µg) were run on a 12% SDS PAGE gel and immunoprobed for IL-1β (R & D Systems, cat # MAB501; 1:500 dilution). In anticipation that the concentration of IL-1β in the medium might not be sufficient to detect, medium was treated with TCA to concentrate the proteins. 1.5 ml conditioned medium was treated with 1-1.5 ml cold 20% TCA for 30 min and centrifuged at 7,000 x g for 10 min. The

TCA pellet was washed with 1 ml 10% TCA, centrifuged again, dissolved with 10 mM

NaOH. The suspension was then added to sample buffer, and the total volume was added to a gel for SDS-PAGE separation of the proteins. No bands, including the one at 50-75 kDa, were detected in these samples. In one case, the concentrated proteins did not migrate on the gel possibly due to problems in pH and/or ionic strength, and in another case, they migrated but did not show immunoreactive bands. The sensitivity of the IL-1β antibody is 50 ng per lane. A 50 µl volume gives rise to a concentration of 1 ng/µl

(compared to nearly 0.00004 ng/µl in the ELISA, see above). Trouble shooting for this analysis would include spiking the medium with pure IL-1β to confirm recovery.

135

Anti-TNFα Western blot analysis of lysate of rMC-1 cells- rMC-1 cells were infected with Ad-Grx1 or Ad-Empty at MOI 0.5 to MOI 25, and lysed for immunoblotting with anti-TNFα (R & D Systems, cat # MAB510; 1:500 dilution). Recombinant TNFα standards were used as a positive control at 1-32 ng (R & D Systems, cat # 5100-RT).

The sensitivity of the antibody was 1 ng/lane. For a 10-well standard gel, 50 µl volumes are typically loaded per well. Thus, the limit of detectibility of this antibody in this system was 20 pg/µl. This technique was the least sensitive of the three assays used;

0.013 pg/µl for ELISA and 0.24 pg/µl for FMAP cytokine multiplex kit (see above).

Sample volumes were 50 µl in the ELISA and FMAP.

Recovery of concentrated recombinant IL-6 - Seventy five pg of recombinant rat IL-6 (R

& D Systems, cat # 506-RL) or rat recombinant IL-6 that accompanied the ELISA kit (R

& D Systems, see above) was added to 30 ml high glucose medium with 2% or 0.67%

FBS or to 1.5 ml high glucose medium with 2% or 0.67% FBS. Each sample was then concentrated to 0.5 ml with Amicon Ultra centrifugal filtration devices (Millipore, cat #

UFC801024, mw cutoff 10,000 Da). Fifty µl of the concentrated medium was assayed by

IL-6 ELISA discussed above.

Statistical Analysis - All values and graphs report mean +/- standard error (S.E.).

Statistical analysis of differences between control and experimental values were determined via the Student’s T-test. Differences displaying p values < 0.05 were considered statistically significant.

136

Figure 3.1. Inhibition of IKK prevents the increased ICAM-1 expression in lysates of rMC-1 cells cultured in normal or high glucose (25 mM) medium, according to

Western blot analysis. rMC-1 cells were cultured in high glucose (25 mM) medium for five days and treated with Bay 11-7085 for 5-10 min on day 2-3 of incubation. A representative Western blot for ICAM-1 and actin loading control and the quantification of blots is shown in A (n=12). Cells in high glucose expressed more ICAM-1, nearly 2- fold (+/-0.6) over cells grown in normal glucose medium, *p<0.002. Treatment with 10

μM and 20 μM Bay inhibitor 11-7085 decreased ICAM-1 expression in high glucose

treated cells. #p< 0.03.

137

Figure 3.2. Inhibition of IKK prevents the increased ICAM-1 expression in lysates of rMC-1 cells over-expressing Grx1 in normal glucose (5 mM) medium, according to Western blot analysis. rMC-1 cells were cultured in normal glucose (5 mM) medium, pre-treated with Bay 11-7085 for 30-40 min. The cells were then infected with MOI 10 of either Ad-Grx1 or Ad-Empty, cultured for 16 - 24 hr in normal glucose medium, and collected in NP40 lysis buffer for immunoblotting of ICAM-1 (1:1000 dilution) and actin (1:30,000 dilution) ( n=7). Ad-Grx1 increased ICAM-1 expression by 2.6-fold (+/-0.5), *p<0.01. ICAM-1 expression in Ad-Grx1 infected cells treated with 10 μM Bay 11-7085 was statistically increased, but cells treated with 20 μM Bay 11-7085 expressed ICAM-1 in similar amounts as control cells (no adenovirus and Ad-Empty MOI 10). In addition, ICAM-1 in cells infected with Ad-Grx1 and treated with 10 μM or 20 μM inhibitor were significantly decreased from Ad-Grx1 infected cells with no inhibitor, #p<0.01.

138

Figure 3.3. Inhibition of IKK prevents the increased nuclear translocation of NFκB in lysates of rMC-1 cells over-expressing Grx1 in normal glucose medium. rMC-1 cells cultured in normal glucose (5 mM) medium were pre-treated with Bay 11-7085 for 30 min, immediately infected with Ad-Grx1 or Ad-Empty (MOI 10), cultured for two days, and collected. Nuclear and cytoplasmic fractions were immunoprobed for p50 (1:1,000 dilution), p65 (1:3,000 dilution), actin (1:30,000 dilution) and YY1 (1:1,000 dilution). Ad-Grx1 increased p50 in the nucleus by 4-fold (+/-0.7) (n=9) (2A) *p<0.002. Cells over-expressing Grx1 treated with 10 μM and 20 μM Bay 11-7085 had nuclear p50 contents that were significantly decreased from those in Grx1-over-expressing cells with no inhibitor (n=9) ( A) #p<0.02. Ad-Grx1 increased p65 in the nucleus by nearly 2-fold (+/-0.7) (n=11) (B)*p<0.002. Cells over-expressing Grx1 treated with 10 μM or 20 μM Bay 11-7085 had nuclear contents of p65 that were significantly decreased from those in Grx1-over-expressing cells with no inhibitor (n=11) (B) #p<0.02.

139

Figure 3.4. Identification of S-glutathionylated IKKβ immunoprecipitated from lysates of rMC-1 cells cultured in normal glucose (5 mM) medium. IKKβ was immunoprecipitated in the presence of iodoacetamide (IAM) from rMC-1 cells cultured in normal glucose medium and immunoblotted under non-reducing conditions (A).

Immunoprecipitated IKKβ is S-glutathionylated at Cys179 as shown by the y3 peak in the LC-MS-MS spectrum corresponding to the loss of the cysteinyl-S-S-glutathione moiety of the peptide sequence containing Cys179 (B). Mass spectral analyses were generated and interpreted by Dr. Anne Distler.

140

Figure 3.5. Peptide sequences of IKKβ that have glutathionylated cysteines. The immunoprecipitated IKKβ was digested and analyzed via MALDI-TOF. The masses of the peptides correlated with the masses of several peptide sequences in the rat IKKβ database. Glutathione was found associated with only those peptides that contained

Cys179. Mass spectral analyses were conducted by Dr. Anne Distler.

141

Figure 3.6. Rat TRiBRB endothelial cells cultured in conditioned medium from rat rMC-1 cells over-expressing Grx1 produce increased Grx1 and ICAM-1 compared to normal medium, according to Western blot analysis. rMC-1 cells over-expressing

Grx1 with Ad-Grx1 at MOI 10 or MOI 20 were cultured for 24 hr, and the conditioned medium (C.M.) was collected. The C.M. was placed on TRiBRB endothelial cells for 24 hr, and the cell lysate was processed for Western blot analysis. C.M. from rMC-1 cells over-expressing two concentrations of Grx1 (Ad-Grx1 MOI 10 and MOI 20) led to an increase of 2.1-fold (+/-0.4) and 2.8-fold (+/- 0.6) in Grx1, respectively (n=4), *p<0.05.

ICAM-1 was increased by 1.5-fold (+/- 0.2) and 2.6-fold (+/- 0.4) in response to conditioned medium from rMC-1 cells over-expressing Grx1 at MOI 10 or MOI 20, respectively (n=7), *p<0.03.

142

Figure 3.7. rMC-1 cells cultured in conditioned medium from rMC-1 cells over- expressing Grx1 produce increased Grx1 and ICAM-1 compared to normal glucose medium, according to Western blot analysis. rMC-1 cells were transfected with Ad-

Grx1 or Ad-Empty at MOI 10 or MOI 20, and the conditioned medium (C.M.) was collected from these cells after 24 hr. C.M. was then placed on newly cultured rMC-1 cells for 24 hr, and the cell lysate was processed for Western blot analysis. C.M. from rMC-1 over-expressing two concentrations of Grx1 (Ad-Grx1 MOI 10 and MOI 20) led to an increase of 2.4-fold (+/-0.6) and 2.9-fold (+/- 0.7) in Grx1, respectively (n=6),

*p<0.03. ICAM-1 was increased by 1.7-fold (+/- 0.3) and 2.2-fold (+/- 0.5) in response

to conditioned medium from rMC-1 cells over-expressing Grx1 at MOI 10 or MOI 20,

respectively (n=6), *p<0.04.

143

Table 3.1. Expression and induction of cytokines and the growth factor, VEGF, in

the conditioned medium from rMC-1 cells over-expressing Grx1. The presence of cytokines in the culture medium of rMC-1 cells over-expressing Grx1 in normal glucose medium was measured via a multiplex cytokine luminex ELISA, and VEGF protein was analyzed from the same samples of medium using a VEGF ELISA. Release of TNFα,

IFNγ, IL-1β, and IL-10 into the medium was not detected. IL-6 and VEGF were both detected in the medium, but only IL-6 was induced by the over-expression of Grx1 in

Müller cells.

144

Figure 3.8. IL-6 is increased in medium from rMC-1 cells over-expressing Grx1 in normal glucose medium (A) or cultured in high glucose medium (B). rMC-1 cells were cultured in normal (5 mM) or high glucose (25 mM), or infected with Ad-Grx1 at

MOI 10 in normal glucose medium. The medium was collected after 24 hr and analyzed via IL-6 ELISA. IL-6 was increased in medium from cells over-expressing Grx1 by 3.7- fold +/- 0.1 (n=6) (*p < 0.003) (A) and cells cultured in high glucose medium by 1.2-fold

+/- 0.04 (n=19) (*p < 0.05) (B).

145

Figure 3.9. Recombinant IL-6 is not fully recovered in cell culture medium that is highly concentrated. Recombinant IL-6 was added to 30 ml high glucose medium with

2% FBS, 30 ml high glucose medium with 0.67% FBS, 1.5 ml high glucose medium with

2% FBS, or 1.5 ml high glucose medium with 0.67% FBS. The medium was concentrated via centrifugation against size exclusion filters by 3-fold (3X) or 60-fold

(60X). IL-6 in the concentrated medium was then assayed using an ELISA. Detection of

IL-6 was diminished with increased concentration (n=2).

146

Chapter 4 Discussion and Future Directions

4.1 Discussion

This thesis work was primarily focused on the role of Grx1 in diabetic retinopathy;

however the findings here coupled with other studies more broadly implicate Grx1 for

important roles in the inflammatory response of cells. The following is a discussion

focused primarily on Grx1 within the context of pro-inflammatory cytokines and

adhesion molecules, the signal transduction cascades involved in triggering an

inflammatory response, and inflammatory diseases. In addition, Grx1 may have

important roles in the complications of diabetes beyond those associated with

inflammation. An example is the potential regulation of potassium channels by Grx1 which is discussed further.

Regulation of pro-inflammatory cytokines and adhesion molecules by Grx1

ICAM-1 is a primary inflammatory marker best characterized for its role in the tight adherence of leukocyte integrins to endothelial cells, an essential step in leukocyte extravasation (Fig. 1.2, pg. 43). Throughout this thesis, an increase in the expression of

ICAM-1 in rMC-1 cells has been used as a marker of NFκB activation and interpreted as

a pro-inflammatory response. Increases in Grx1 consistently lead to corresponding

increases in protein expression of ICAM-1 in rMC-1 cells. In addition, in rMC-1 cells

upregulation of Grx1 induces the secretion of IL-6, a cytokine commonly categorized for

147

its pro-inflammatory effects. Conditioned medium from Grx1 over-expressing rMC-1

cells contains elevated IL-6 protein and induces ICAM-1 expression in freshly cultured rMC-1 or endothelial cells. Furthermore, recombinant IL-6 induces the production of

ICAM-1 in rMC-1 and endothelial cells. Thus, the pro-inflammatory effects exerted by

Grx1 can be observed in the changes in IL-6 and ICAM-1.

The classical definition of an inflammatory response is a reaction of the vascular tissue to a harmful stimulus which results in the recruitment of white blood cells to the site of injury. The response involves a series of adhesion molecules such as ICAM-1 whose function is to recruit monocytes and macrophages to the site of injury. Leukostasis and diapedesis are hallmarks of inflammation. Accordingly, the increased protein expression of ICAM-1 by the Müller cell is not necessarily a pro-inflammatory response unless the

ICAM-1 on the surface of the Müller cell bind to infiltrated monocytes and macrophages, promoting leukostasis. However, the conditioned medium containing IL-6 could be regarded as the harmful stimulus, and the increase in expression of ICAM-1 on the endothelial cell could be regarded as the vascular response. Furthermore, changes in pro- inflammatory markers are commonly accepted to be indicative of inflammation regardless of whether the function of that marker in the inflammatory response is clear in that particular cell type.

Alternatively, ICAM-1 could have pro-inflammatory effects or be a mediator whose primary function is to transduce signals from the outside of the cell to the inside but not necessarily within the context of classical inflammation. ICAM-1 expression is induced

148

by cytokines such as IFNγ, IL-1, TNFα, and IL-6 (Teppo et al., 2001). In addition to

inducing cytokines, ICAM-1 can itself be induced by cytokines. Anti-ICAM-1 antibody

cross-linking to ICAM-1 can lead to induction of RANTES and ICAM-1 in renal

fibroblasts (Blaber et al., 2003). Similarly, this cross-linking can lead to the induction of

ICAM-1 and VCAM in endothelial cells and fibroblasts (Clayton et al., 1998). In addition, the ligation of ICAM-1 can lead to induction and secretion of RANTES and IL-

8 from endothelial cells (Blaber et al., 2003;Clayton et al., 1998;Sano et al., 1998).

Ligand binding to ICAM-1 can also induce internalization of ICAM-1 and its binding

partner in a unique mechanism of CAM-mediated endocytosis in endothelial cells (Muro

et al., 2005). These studies support a model where cytokines induce ICAM-1, and

ICAM-1 internalization might subsequently induce cytokines.

Other studies have implicated ICAM-1 in intercellular uptake of extracellular particles.

For example, ICAM-1 was recently suggested to play a role in neuroinvasion by the West

Nile virus and the subsequent encephalitis (Dai et al., 2008). Consistent with this

interpretation, ICAM-1 is important in the entry of enteroviruses and rhinoviruses and

infectivity of HIV-1 (Dai et al., 2008;Xiao et al., 2001). Therefore the role of ICAM-1

may be multi-faceted. In addition, in the current study the particular role of ICAM-1 in

glial cells needs to be elucidated to understand why it is expressed and upregulated in the

Müller cell, and how Grx regulation impacts these regulatory processes (see future

directions).

149

S-glutathionylation of ICAM-1 was reported in human pulmonary aortic endothelial cells

(HPAEC), and this was increased by TNFα (Mukherjee et al., 2007). Increased glutathionylation of ICAM-1 correlated with enhanced surface expression of ICAM-1 and monocyte adhesion, implying that glutathionylation of ICAM-1 would be a pro- inflammatory signal. This is an intriguing area of study, but regulation of ICAM-1 glutathionylation by Grx1 needs to be examined, and documentation of functional changes in ICAM-1 corresponding to changes in its glutathionylation status will be critical. Also, the HPAEC model contrasts to the findings in retinal glial cells in the current study and in epithelial cells in another study (Reynaert et al., 2006) which imply that glutathionylation plays a anti-inflammatory role via inhibition of the IKK/NFκB signaling pathway, leading to attenuated cytokine production and intracellular ICAM-1 production. Therefore, it is important to determine whether the increased ICAM-1 observed in the current study corresponds to increased surface expression (see future directions). Additional regulation of ICAM-1 by direct glutathionylation would be an interesting field of study with implications for viral infectivity, endocytosis, ICAM-1 oligomerization, leukostasis, proteolytic cleavage to sICAM forms, and cellular signal transduction.

Grx1 regulation of NFκB signal transduction

IKK is the major point of convergence of many signaling cascades (Fig. 1.3, pg. 44).

Thus its regulation is of great importance. Regulation of IKKβ by Grx1 has been more

thoroughly documented than for other IKK subunits, but the overall impact of IKK

150

subunit glutathionylation on inflammatory responses will need to be evaluated with

respect to cell type, subcellular localization, and stimulus because glutathionylation of

IKKα also has been reported in endothelial cells (Mukherjee et al., 2007). In addition, it

will be interesting also to determine the catalytic efficiency (Vmax/Km) of Grx1 for

deglutathionylation of each IKK subunit. This would allow for the assessment of how

IKK activity is affected by changes in Grx1.

Studies aimed at learning whether the additional accessory proteins in the IKK complex influence the substrate specificity or accessibility of the core IKK subunits to Grx1 or whether glutathionylation of the other IKK proteins (α, γ) proteins impact signaling via the IKK complex will advance understanding of this complicated regulatory process.

Furthermore, comparative studies of the relative efficiency of Grx-mediated deglutathionylation of IKK, ICAM, and other proteins discussed throughout this treatise

(e.g., Akt, PKC, SERCA, and Ras) would be helpful in evaluating the potential intracellular substrates for catalysis by Grx1. These findings could then be used to predict the overall effects of Grx1 on cells in vivo.

S-glutathionylation of inflammatory signaling mediators

NFκB signaling and ICAM-1 are the central focus of this thesis because they are prominent inflammatory mediators. However, many other proteins involved in inflammation have been identified as potentially regulated by glutathionylation, expanding the potential impact of Grx1 in these processes (Fig. 1.2, pg. 43). Analogous

151 to ICAM-1 binding to its counter-receptor, very late antigen-4 (VLA-4, α4β1) is a leukocyte integrin that binds to vascular cell adhesion molecule-1 (VCAM-1), mucosal addressin cell adhesion molecule-1 (MadCAM-1), and fibronectin (Liu et al., 2008) (Fig.

1.2, pg. 43). All 24 cysteines of the α4 subunit of VLA-4 are found in its extracellular domain (Liu et al., 2008), and mutations of Cys278 and Cys717 inhibited VCAM-1 binding in K562 cells (Pujades et al., 1996). Recently, the α4 subunit was shown to be glutathionylated under basal conditions in HL-60 clone 15 cells, a cell line that is known to differentiate into eosinophil-like cells upon treatment with n-butyrate (Liu et al.,

2008). Increases in glutathionylation of α4 were correlated to increases in binding to recombinant VCAM-1, and decreases in cell rolling on VCAM-1-coated flow chambers

(Liu et al., 2008). These studies demonstrated redox regulation in a stimulus- and concentration-dependent manner. Thus, VCAM-1 and ICAM-1 are at least two adhesion molecules implicated to be regulated by glutathionylation.

High mobility group B1 (HMGB1) is a nuclear alarmin that is released into the extracellular milieu via cellular necrosis or lysosomal exocytosis after translocation to the cytoplasm in activated monocytes, macrophages, mature dendritic cells, natural killer cells, and endothelial cells (Harris and Raucci, 2006;van Beijnum et al., 2008). HMGB1 acts as a chemokine to recruit circulating monocytes, T cells and dendritic cells, induces

NFκB activation and subsequent production of pro-inflammatory cytokines, and elicits cross presentation and antigen specific T cell immunity of activated dendritic cells

(Hagemann et al., 2007;Harris and Andersson, 2004;van Beijnum et al., 2008). HMGB1 can initiate pro-inflammatory cytokine production through both toll-like receptors (TLRs)

152

and the receptor for advanced glycation end products (RAGE) (Fig. 1.2, inset, pg. 43).

HMGB1 has a 7-fold higher affinity for RAGE than does its classical ligands, advanced

glycation end products (AGEs), and independent of HMGB1, involvement of the RAGE

receptor is found to be prevalent in diabetes.

Glutathionylation of HMGB1 was reported in the nucleus of human and rat retinal pigmented epithelial (RPE) cell lines in response to a non-physiological oxidant (Hoppe

et al., 2006). However, cytosolic contamination controls and physiological stimuli are

needed to establish that the glutathionylated form of HMGB1 is indeed in the nucleus.

Nevertheless, evidence of protein glutathionylation in the nucleus is limited but not

unprecedented. S-glutathionylation of nuclear NFκB has been implicated previously in

pancreatic cancer cells (Qanungo et al., 2007). Overall, delineating the impact of changes in Grx1 and glutathionylation on HMGB1 localization, recruitment of T cells, dendritic cells, and monocytes, and binding to RAGE and TLR4 on dendritic cells and monocytes would uncover important new mechanisms in regulation of inflammation.

HS60 and HS70 are other examples of TLR2/4 ligands (Seong and Matzinger, 2004) that have been reported to be glutathionylated in T lymphocytes (Fratelli et al., 2002).

Furthermore, oxLDL can activate signal transduction pathways by binding to TLRs and

RAGE (Fig. 1.2; inset, pg. 43) (Harja et al., 2008;Seong and Matzinger, 2004). PON1, a paraoxonase found in complex with high density lipoprotein (HDL), prevents oxidation of low density lipoprotein (LDL), and hydrolyzes lipid peroxides in LDL (Rozenberg and

Aviram, 2006;Singh et al., 2007). Overall, PON1 activity is associated with

153

inflammatory diseases such as coronary artery disease, atherosclerosis, and diabetes

(Flekac et al., 2007;Singh et al., 2007); however, the reported activity of PON1 seems to

be dependent on the nature of the study, and displays inter-individual and racial variation.

For example, PON1 activity is decreased in coronary artery disease in plasma of North

West Indian Punjabis irrespective of diabetes (Singh et al., 2007). However, PON1 activity has been reported to be decreased in the serum of diabetic patients in a Japanese

population (Sakai et al., 1998) and in a group of diabetic patients of both European and

Asian Indian decent (Mackness et al., 1998). Activity of PON1 specifically in leukocytes of a non-specific population of people is decreased in both type 1 and type 2 diabetes

(Flekac et al., 2007). Modifications to Cys283 in PON1 inhibit the LDL oxidation protection, arylesterase activity, and paraoxonase activity (Aviram et al., 1999). PON1 that was evolved in E. coli via several rounds of DNA shuffling and screening and PON1 associated with serum HDL that was extracted from human volunteers were both reported to be inhibited by glutathionylation (Rozenberg and Aviram, 2006). These arylesterase, paraoxonase, and lactonase activities could be restored with DTT. However, more analysis is needed to determine whether glutathionylation of PON1 occurs in vivo in response to a physiological stimulus, and is reversible by an endogenous catalyst i.e.,

Grx1.

ROS are known to activate the PI3K-Akt pathway, and extracellular ATP leads to generation of ROS in macrophages via the purinergic receptors (P2X and P2Y) (Cruz et al., 2007;Foell et al., 2007). Akt activity is promoted by PI3K-catalyzed PIP3 formation, and inhibited by conversion of PIP3 back to PIP2, mediated by PTEN (phosphatase and

154 tensin homologue deleted from chromosome 10) (Fig. 1.2, inset, pg. 43 and Fig. 1.3, pg.

44). ATP-driven formation of ROS led to the activation of the PI3K/Akt pathway and production of IL-1β in alveolar macrophages, and this response corresponded to an increase in glutathionylation of PTEN (Cruz et al., 2007). Glutathionylation was interpreted to inhibit PTEN activity, thereby enhancing activation of the PI3K/Akt pathway. Akt itself was previously suggested to be regulated by Grx1 (Murata et al.,

2003;Wang et al., 2007a), and additional Akt interacting proteins PP2A and ASK-1 have been implicated in regulation by glutathionylation and binding to Grx1, respectively.

Learning the extent of regulation by glutathionylation of each signaling mediator is necessary to reveal individual contributions to the overall control of the activity of the

PI3K/Akt pathway. Nevertheless, PI3K/Akt signaling is a prevalent inflammatory pathway, and potential regulation by glutathionylation has significant implications.

Taken together, these studies exemplify the diversity of proteins that are implicated in regulation by glutathionylation. The widely different functions of these proteins include adhesion molecules, enzymes, ion transporters, cytokines, and transcription factors.

Further studies are encouraged that will evaluate whether reversible glutathionylation of these identified proteins actually plays an important role in regulation of cellular functions. In particular, does glutathionylation or deglutathionylation of these proteins occur within the cell under natural conditions in response to physiological stimuli and lead to functional changes that are reversible by Grx1?

155

Grx1 and glutathionylation in inflammatory diseases- Findings on changes in Grx1 in

various diseases have exploded in the past five years (Fig. 1.1, pg. 42). Much work is

still needed to fully determine the underlying meaning of most of these changes.

Nevertheless, many diseases that are already linked by the involvement of inflammatory

events are now also reported to share changes in Grx1. Studies implicating Grx1 in

diabetes, atherosclerosis, COPD, asthma, cancer and Alzheimer’s disease are discussed

below, with an emphasis on diabetes.

Diabetes- Grx1 has been reported previously to be decreased in platelets of diabetic patients (Di Simplicio et al., 1995) and increased in hearts of diabetic rats (Di Simplicio

et al., 1995;Li et al., 2005). The novel work presented in this thesis demonstrates

changes in Grx1 in the diabetic (rat) retinae. Moreover, mechanistic insight into the

regulation of the NFκB signaling pathway by Grx1, regulation of Grx, and regulation of

ICAM-1 and IL-6 by Grx has been gained.

Insulin is essential in facilitating cellular glucose uptake, and glucose is the major energy

source of most cells. However, excess glucose metabolism can have harmful effects

within the cell. Chronic exposure to high glucose can lead to insulin resistance, i.e., cells

no longer take up glucose in response to insulin, and without glucose, cells undergo

starvation. Regulation of insulin release or activity and glucose uptake is at the core of

diabetes. Figure 4.1 (pg. 194) depicts potential sites of regulation by glutathionylation

and Grx1 in the “triggering pathway’ of insulin secretion from pancreatic β-cells via

exocytosis in response to glucose metabolism (Ashcroft, 2005;Henquin, 2000). All the

156

regulatory points are shown within β-cells to illustrate the potential impact that Grx1

could have within a single cell type and to provide a conceptual framework for

discussion. However, each of the studies on glutathionylation and Grx1 has been conducted in different contexts, and only a few actually have taken place within the context of diabetes. Cell one shows the key events in insulin secretion, and cell two illustrates signal transduction pathways downstream of insulin. The mediators that have not been discussed elsewhere in this thesis will be discussed below.

(1) Glutathionylation of aldose reductase (AR) - Glucose is typically phosphorylated by

hexokinase prior to entering the glycolysis pathway. However, elevated glucose

concentrations saturate hexokinase and trigger a second metabolic pathway called the

polyol (sorbitol) pathway. The first step in this pathway depends on aldose reductase

(AR) to convert glucose to sorbitol, and serves as a ‘backup’ system since hexokinase has

a higher affinity (i.e., lower Km) for glucose than does aldose reductase (King et al.,

1994;Tomlinson and Gardiner, 2008). Mechanisms of sorbitol related complications

include sorbitol accumulation leading to osmotic swelling and cataract formation in

diabetes (Srivastava et al., 2005;Packer et al., 2001).

Inhibition of aldose reductase in diabetes has been a popular therapeutic goal for many

years. Cys298 is in the active site of AR, and S-glutathionylation at Cys298 inhibits its

activity in the presence of normal glucose concentrations (Srivastava et al., 2005),

suggesting basal glutathionylation similar to that observed with actin (Wang et al., 2001)

and IKK (this thesis). Inhibitors of AR thus far have proven more promiscuous than

157

effective (Srivastava et al., 2005), but since it is a regulatory target for Grx1, inhibition of

Grx1 may provide additional means for therapeutic intervention in diabetes. Since AR is

glutathionylated and inhibited at basal states, either Grx1 activity is low within the

cellular micro-domain of aldose reductase, the glutathionylation site of AR is

sequestered, or AR has a high Km for deglutathionylation by Grx1. Regardless, AR is

more active in diabetes, predictive of a decrease in glutathionylation of AR (↓ AR-SSG,

↑ AR-SH) corresponding to an increase in Grx1 activity, as observed for the retinae and

heart (Shelton et al., 2007;Li et al., 2005). Additionally, hyperglycemia may lead to

alterations in the structure of AR so that the glutathionyl motiety is exposed to Grx1.

(2) Grx1 regulation of potassium channels- Increased glucose metabolism leads to the

closing of KATP channels via increased ATP concentrations (Henquin, 2000;Ashcroft,

2005). The subsequent decrease in potassium efflux causes a depolarization in the membrane potential of pancreatic β-cells and triggers an influx of calcium ions via the

voltage-gated calcium ion channels (Henquin, 2000;Ashcroft, 2005). The activity of

KATP channels can be modulated by thiol and redox sensitive cysteines, suggestive of

potential regulation by S-glutathionylation (Matsuo et al., 1999;Trapp et al., 1998;Lee et

al., 1994;Coetzee et al., 1995;Weik and Neumcke, 1989). Sulfonylurea drugs increase

insulin release by closing KATP channels (Ashcroft, 2005), and similarly, if

glutathionylation leads to inhibition and downstream insulin release, inhibition of Grx1

would have therapeutic benefits for diabetic patients.

158

A decrease in calcium-independent transient outward K+ current contributes to the

extended action potential duration observed in hearts of diabetic rats (Li et al., 2005).

Decreased potassium current in myocytes from diabetic rats was correlated to an increase in Grx1 (Li et al., 2005), proposing a protective role of the Grx1 system in regulation of cardiac potassium ion channels. Unlike the beneficial effects of potassium channel inhibition on insulin release in β-cells discussed above, inhibition of these particular cardiac channels have undesirable consequences. The association of increased Grx1 with channel inhibition promotes Grx1 as a therapeutic target in diabetes. Further studies are needed to determine whether channel deglutathionylation is the mechanism of action of

Grx1 in these systems.

(3) S-glutathionylation of the calcium channels, RyR and SERCA- Ryanodine receptors

(RyR) release Ca2+ from the endoplasmic reticulum into the cytosol. Whether or not protein expression of RyR is decreased in diabetes is contentious (Yaras et al.,

2005;Bidasee et al., 2001), but a decrease in the function of the RyR in hearts of diabetic rats is generally accepted. Glutathionylation of RyR has been related to calcium release from sarcoplasmic reticulum vesicles (SRV) from rabbit muscle (Aracena et al., 2003).

Multiple cysteines can become glutathionylated under specific stimuli, including

Cys3635, the only cysteine known to be involved in RyR-regulated calcium release thus far (Aracena et al., 2003).

In direct contrast to the RyRs, SERCA pumps actively transport cytosolic Ca2+ into the sarcoplasmic reticulum (SER), quenching cytoplasmic Ca2+ signals and regulating

159

calcium oscillations in response to glucose (Arredouani et al., 2002). Each of the three

isoforms of SERCA (SERCA1, SERCA2, and SERCA3) has highly spliced tissue

dependent variants (Erkasap, 2007). Glutathionylation of SERCA1 from rabbit skeletal

muscle was reported for five different cysteine residues (Viner et al., 1999).

Nitric oxide (NO)-activated calcium uptake into the SER is mediated by increases in

SERCA2 activation via glutathionylation of SERCA2 at Cys674 (Adachi et al., 2004b),

presumably via stepwise formation of S-nitrosylated SERCA and then S-glutathionylated

SERCA. Subsequently, NO was shown to promote activation of the calcium pump,

SERCA2b, via Cys674-SSG formation in vascular smooth muscle cells (VSMC), and this activation counteracted the elevated cytosolic Ca2+ and VSMC migration under normal

glucose conditions (Tong et al., 2008). High glucose prevented NO-induced inhibition of

VSMC migration, and the inhibition could be circumvented by superoxide dismutase

(Tong et al., 2008) suggesting a role for peroxynitrite. These results suggest a scenario

involving increased Grx1 de-glutathionylation activity producing more reduced SERCA-

Cys674-SH that is then susceptible to irreversible oxidation to the sulfonic acid or direct

Grx1-mediated formation of glutathionylated SERCA2b. Such a scenario is consistent

with reports that high glucose and streptozotocin-induced diabetes lead to upregulation of

Grx1 content and activity in the heart and the retina (Li et al., 2005;Shelton et al., 2007).

Since SERCA3 also has implications in diabetes and SERCA3-deficiency leads to

increased islet cell response (Arredouani et al., 2002), deglutathionylation and

subsequent inhibition of SERCA activity via increased Grx1 may be a beneficial tool in

160

diabetes. Thus, comparative analysis of the catalytic efficiency of Grx1 for each of the

SERCA isoforms will be necessary for evaluating the global versus local therapeutic

utility of modulating Grx activity since both decreased SERCA3 activity in islet cells and

increased SERCA2b activity in VSMCs appear to be beneficial in diabetic complications.

(4) S-glutathionylation of PKC- Protein kinase C (PKC) is activated by increased diacylglyceride (DAG), and is a major pathway in the pathogenesis of diabetic complications, primarily in vascular complications as mentioned in Chapter 1 of this thesis (King et al., 1994). PKC also operates upstream in JNK/ERK signaling pathways.

PKC enzymes are classified as calcium-dependent (cPKC), novel calcium-independent

(nPKC), or atypical (aPKC). The cPKCs (α, β1, and β2) and nPKCs (ε, δ) isoforms each

have tissue-specific implications in the diabetic retina, glomerulus, heart, and aorta, but

the β-isoforms in the vasculature have the most profound significance (Koya and King,

1998). Ruboxisaurin is a PKC-β inhibitor being tested in clinical trials for vascular protection in diabetic retinopathy (Clarke and Dodson, 2007). Several isoforms of PKC

(PKC-α, β1, β2, ε, and δ) have been reported to be inactivated by glutathionylation in vitro

(Ward et al., 2000;Chu et al., 2001), and Grx1 was reported to reactivate PKC-α specifically in NIH3T3 cells (Ward et al., 2000;Chu et al., 2001). If Grx1-mediated deglutathionylation of PKC leads to its activation in vivo, Grx1 would likely be an additional therapeutic target for diabetic vascular complications. However, this is a simplified view which does not consider other control points potentially regulated by glutathionylation, such as aldose reductase that can signal to PKC and NFκB (Srivastava et al., 2005).

161

(5) S-glutathionylation of NFκB- IκB kinase (IKK) phosphorylation of IκB promotes subsequent ubiquitination and degradation of IκB, freeing NF-κB for nuclear translocation where it binds DNA and activates transcription. NFκB signaling is activated in many models of diabetes such as the retina (Shelton et al., 2007;Zheng et al.,

2004;Romeo et al., 2002), kidney (Schmid et al., 2006), and liver (Cai et al., 2005).

NFκB activation is a common downstream event in major pathways leading to diabetic complications such as those involving AGEs and PKC; and mitochondrial superoxide has been implicated in NFκB hyperactivity in high glucose (Brownlee, 2001). In vitro glutathionylation of p50 (p50-SSG) was shown to inhibit its specific binding to DNA

(Pineda-Molina et al., 2001). Glutaredoxin restored transcriptional activity of inactivated p65 in pancreatic cancer cells, indicating inactivation via glutathionylation of p50, p65, or a transcriptional co-factor (Qanungo et al., 2007). Furthermore, glutaredoxin was reported to enhance NIK-induced activity of NFκB in HEK293 cells (Hirota et al., 2000).

To date, S-glutathionylation has been reported for 13 proteins within the NFκB pathway in various cells and experimental contexts (Fig 1.3, pg. 44). For example, the ubiquitin/proteasome is increased in diabetic patients (Marfella et al., 2006), and glutathionylation inhibits the ubiquitin-activating (E1) and ubiquitin-carrier (E2) enzymes

(Obin et al., 1998;Jahngen-Hodge et al., 1997), and the 20S proteasome (Demasi et al.,

2003).

(6) Grx1 regulation of insulin exocytosis- Glucose metabolism generates NADPH and induces the release of insulin from pancreatic β-cells. NADPH added intracellularly also

162

has been reported to increase insulin release (Ivarsson et al., 2005) and elicits a similar

extent of membrane depolarization as that associated with calcium mediated exocytosis

in mouse β-cells, suggesting that NAPDH mediates glucose induced exocytosis of insulin

(Ivarsson et al., 2005). The authors rationalized that since Grx1 and Trx are electron acceptors of NAPDH, they would be mediators in NAPDH-induced capacitance.

However, Trx inhibited NAPDH-induced capacitance whereas Grx1 potentiated it.

Whether the effect of Grx1 involves regulation via reversible S-glutathionylation of

specific proteins remains to be discovered, and the basis for the different effect of Trx is

unknown.

(7) S-glutathionylation of PTP, and regulation by Grx1- Protein tyrosine phosphatases

(PTPs) are essential for the deactivation of the insulin receptor once insulin is no longer

present (Goldstein et al., 2005). PTP-1B is identified as the major phosphatase enzyme

that inhibits insulin signaling (Goldstein et al., 2005), and genetic manipulations and

inhibition of PTP-1B increase insulin signaling (Goldstein et al., 2005;Tonks, 2003).

Accordingly, PTP-1B has become a prime candidate for therapeutic intervention in

diabetes and obesity (Tonks, 2003;Zhang and Lee, 2003;Pei et al., 2004;Goldstein,

2001). ROS potentiate insulin signaling (Goldstein et al., 2005), paradoxical to the

deleterious effects of oxidative stress well known in diabetes and associated

complications. Primarily, ROS have been shown to mediate insulin signaling and

inhibition of PTP-1B (Goldstein et al., 2005). PTP-1B can be inactivated by glutathionylation in a Grx1-reversible fashion (Barrett et al., 1999a;Barrett et al., 1999b),

and it is converted to PTP-1B-SSG in situ in A431 cells in response to ROS generated

163

intracellularly. Since PTP-1B inhibition is beneficial in diabetes, inhibition of Grx1

would provide an alternative therapeutic approach, leading to increased glutathionylation

of PTP-1B and its concomitant inactivation.

(8) S-glutathionylation of Ras, and regulation by Grx1- The role of Ras in diabetes is

complex and cell-type specific. For example, insulin is well established for its ability to

activate Ras (Goalstone and Draznin, 1998) and for signal transduction via PI3K in many

cell types including those from the liver, muscle, and adipose tissue (Shepherd et al.,

1998;Farese et al., 2005). In separate studies Ras has been reported to both activate PI3K

and serve as an effector of PI3K (Shepherd et al., 1998), but Ras is not essential in insulin

activation of PI3K in adipocytes (Gnudi et al., 1997;van den Berghe et al., 1994).

Furthermore, Ras is an insufficient trigger for insulin-induced glucose uptake in adipocytes (van den Berghe et al., 1994). In male diabetic mice but not female, Ras is involved in destruction of β-cells (Efrat et al., 1990). Glucose activates H-Ras in retinal capillary endothelial cells, and the data suggest that superoxide plays an important role in this event (Kowluru and Kowluru, 2007). Glutathionylation activates Ras in VSMC, and this activation is decreased in Grx1-over-expressing cells (Adachi et al., 2004a).

Moreover, glutathionylation of Cys118 and corresponding activation of Ras leads to insulin resistance, and insulin signaling was recovered with Grx1 over-expression, implicating a role for Grx1 treatment in diabetes (Clavreul et al., 2006).

(9) S-glutathionylation of MEKK- MEKK mediates Ras-Raf signal transduction to JNK

and c-Jun (Vojtek and Der, 1998), and p21ras is reported to act both downstream (van

164

den Berghe et al., 1994) and upstream (Clavreul et al., 2006) of insulin receptor substrate-1 (IRS-1) in the insulin signaling cascade. S-glutathionylation inhibits MEKK1 in menadione-treated lymph node carcinoma prostrate cells (Cross and Templeton, 2004), and glutaredoxin enhances NFκB activation through MEKK in HEK293 cells (Hirota et

al., 2000). By analogy, reversible glutathionylation may play a role in regulating

MEKK-dependent insulin signaling.

(10) S-glutathionylation of c-Jun- c-Jun is phosphorylated by c-Jun NH2-terminal kinase

(JNK), and JNK has implications in the pathology of β-cells via IL-1β or specifically in

mediating IRS-1-insulin receptor interactions (Aguirre et al., 2000;Major and Wolf,

2001). Perturbation of the redox-regulation of the JNK signaling cascade is important in

diabetes, and inhibition of JNK signaling leads to beneficial effects in type 1 and type 2

diabetic mice (Kaneto et al., 2007). In this regard, c-Jun has been shown to undergo

glutathionylation in vitro (Klatt et al., 1999). Furthermore, c-Jun activation in MCF-

1/ADR cells is reported to be hindered by Grx1 binding to ASK-1 (ASK:Grx1), a

mechanism of regulation distinct from the typical de-glutathionylation activity of Grx1

(Song et al., 2002). Hence, interference with these modes of regulation by alteration in

Grx1 content may be important also in understanding the complications of diabetes.

Summary on diabetes- The majority of studies, including those on aldose reductase,

potassium channels, IKK, and PTP-1B, support the notion that inhibition of Grx1 would

have beneficial effects. However, other reports such as those on insulin exocytosis and

SERCA suggest otherwise. Determining the catalytic efficiency of endogenous Grx1

165 towards individual glutathionylated protein substrates will be critical in evaluating the physiological and pathological outcomes of Grx1 as a therapeutic target. Furthermore, to approach a more complete understanding regarding specificities associated with tissue type, cell type, subcellular compartment, and micro-domain must be taken into consideration.

Many more proteins whose activities can be changed by glutathionylation are implicated in various aspects of diabetes, including metabolism, homeostatic and redox regulation, protein folding, leukocyte activation, transport, and cell death. However, most of these other proteins have not been studied under physiologically relevant conditions, nor have they been tested for reversibility by Grx1. A partial list of proteins that have been reported to be glutathionylated in a variety of conditions include alcohol dehydrogenase

(Klatt and Lamas, 2000), superoxide dismutase (Klatt and Lamas, 2000;Tao and English,

2004), malate dehydrogenase (Eaton and Shattock, 2002), creatine kinase (Klatt and

Lamas, 2000), glycogen phosphorylase b (Klatt and Lamas, 2000), calbindin (Tao and

English, 2004), cathepsin K (Percival et al., 1999), fatty acid binding protein (Fratelli et al., 2002), heat shock protein 60 (Fratelli et al., 2002), pro-caspase-3 (Klatt and Lamas,

2000;Pan and Berk, 2007), and GAPDH (Tao and English, 2004).

Interestingly, diabetes is associated with increased susceptibility to other pathologies such as the other inflammatory diseases discussed below. For example, heart disease and stroke are the leading causes of death of diabetic patients, and atherosclerosis is the prototypic cardiovascular disease involving inflammation. The abundance and

166

significance of research on the combined areas of atherosclerosis and diabetes can be

underscored by the nearly 3000 PubMed entries found for review articles alone. Lung

dysfunction also increases in diabetes (Goldman, 2003). Furthermore, diabetes increases

the risk for Alzheimer’s Disease, and vice versa (Janson et al., 2004;Martins et al., 2006).

A similar relationship seems to exist between diabetes and pancreatic, colorectal,

prostate, endometrial, liver, breast, and renal cell cancers and non-Hodgkin’s lymphoma

(Wang et al., 2003a;Volkers, 2000). How diabetes predisposes patients to other diseases

is unknown, but regulation and dysregulation in cellular redox homeostasis is an area in

common that may be involved.

Atherosclerosis- Atherosclerotic plaques induce signaling pathways in endothelial cells

via toll-like receptors (TLR2 and TLR4) (Seong and Matzinger, 2004). Monocyte

diapedesis and concomitant differentiation into macrophages are hallmarks of the early

stages of atherosclerosis. Subsequent differentiation of macrophages into foam cells is

accompanied by lipid uptake and accumulation, and foam cell aggregation constitutes the

core of atherosclerotic plaques. OxLDL enhances monocyte recruitment and is taken up by macrophages, promoting foam cell development, and oxLDL can be found in established macrophage foam cells within atherosclerotic plaques (Bobryshev, 2006).

Grx1 has been found in endothelial cells, adventitia fibroblasts, and medial smooth muscle cells in both atherosclerotic and non-atherosclerotic coronary arteries of postmortem humans, and localized predominantly to infiltrated macrophages of atherosclerotic lesions (Okuda et al., 2001). OxLDL seems to mediate glutathionylation

of Ras and subsequent Akt signaling in endothelial cells (Clavreul et al., 2006). Ras is

167

activated by glutathionylation, and this is reversible by Grx1 (Adachi et al., 2004a).

Further, Ras has been reported to mediate atherogenesis in part by inducing inflammatory

cytokines (Minamino et al., 2003). Thus, further studies are needed to test whether Grx1-

mediated de-glutathionylation and inactivation of Ras would have therapeutic potential in

atherosclerosis. In addition, Grx1 has been implicated in protection of human

macrophages from oxLDL-induced cell injury (Wang et al., 2006).

Inflammatory lung diseases- Lung tissue is exposed to more oxygen than most other parts of the body, suggesting that redox regulation would play a critical role in physiological and pathological processes of the pulmonary system. Briefly discussed here are the potential roles of Grx1 and reversible protein-SSG formation in two related allergen-induced inflammatory diseases, chronic obstructive pulmonary disease (COPD) and asthma.

COPD- Cigarette smoke, the most common irritant of chronic obstructive pulmonary disease (COPD), leads to accumulation and activation of macrophage and neutrophils, and the subsequent tissue damage in the lung is irreversible (Peltoniemi et al., 2006). A decrease in Grx1 was reported in alveolar macrophages in progression of COPD severity, and Grx1 content was correlated with lung function (Peltoniemi et al., 2006). The decreases in Grx1 observed in whole lung homogenates were similar to those in the

inflammatory cells, indicating that Grx1 is mostly localized in macrophages (Peltoniemi

et al., 2006). The authors speculated that the decrease in Grx1 could be a result of down-

regulation by TGF-β or secretion of Grx1. In contrast to decreases in intracellular Grx1,

168

it was reported to be elevated in sputum supernatants in patients with COPD

exacerbations. However, the source and the function of this extracellular Grx1 must be

further characterized in order to assess its physiological roles and involvement in COPD.

Asthma- Similar to COPD, allergen-induced inflammation in the lung is a common component of asthma, and eosinophil and neutrophil generation of ROS is robust. A number of different changes in Grx1 were observed in different contexts. First, Grx1 was found to be increased in the murine lung with allergic airway disease (Reynaert et al.,

2007). Secondly, treatment with IL-13 and IFNγ led to increases in Grx-like activity

while TGFβ and IL-4 led to the opposite affect in murine tracheal epithelium; and finally,

TNFα was found not to have an effect on Grx1 (Table 1.1, pg. 45, (Reynaert et al.,

2007)). Accordingly, treatment of cells with IFNγ and TGFβ led to a decrease and increase in protein glutathionylation, respectively. In contrast to the report by Peltoniemi et al. (2004) of weak detection of Grx1 in human bronchial epithelial cells, Reynaert et al. (2007) found a substantial amount of Grx1 protein in murine airway epithelium, and these authors speculated that species specificity might account for the difference in observations. Furthermore, Grx1 and glutathionylation of IKK were shown to be involved in NFκB-mediated cytokine production (Reynaert et al., 2006). Therefore, it will be interesting to see how cytokines such as IL-4, IL-13, and IFNγ affect Grx1 and protein glutathionylation in analogous studies of human tissue.

Cancer- The damage and death of cells by cancer treatments leads to the release of

chemokines, cytokines, and factors such as the HMGB1 (Apetoh et al., 2007;Hagemann

169

et al., 2007) (see above for HMGB1-SSG). Inflammatory cells are essential for clearing

damaged cancer cells and cell debris, and macrophages can constitute up to 50% of tumor

masses (Hagemann et al., 2007). HMGB1 may trigger tumor-promoting pro-

inflammatory cytokine production from macrophage and dendritic cells (Hagemann et

al., 2007). However, the release of HMGB1 from dying cancer cells was shown to be an

essential component of the T cell antitumor antigen cross presentation immune response

via toll-like receptor-4 (Apetoh et al., 2007). Characterizing the effects of

glutathionylation and Grx1 on HMGB1 localization, chemotaxis, and receptor binding is

critical in evaluating the utility of redox manipulation in cancer therapeutics. In addition,

the study linking regulation of NFκB activity to Grx1 mediated glutathionylation was

conducted in the context of hypoxic pancreatic cancer cells (Qanungo et al., 2007),

implicating NFκB and modulation of its glutathionylation status as a regulatory

mechanism in cancer. Adding to with this concept, a previous study reported elevated

levels of Grx1 in pancreatic carcinoma tissue (Nakamura et al., 2000). Together, these

studies support the role of NFκB activation (which is inhibited by Grx1-reversible

glutathionylation) in producing pro-survival factors in pancreatic cancer. These limited

examples already demonstrate profound implications for cancer regulation via Grx1 and

glutathionylation.

Alzheimer’s Disease A defining feature of Alzheimer’s disease (AD) is β-amyloid

toxicity. Accumulation of A-beta plaques, neurofibrillary tangles, and reactive microglia are thought to underlie neuroinflammation (Akiyama et al., 2000;Shepherd et al., 2007).

In addition, β-amyloid is a ligand for RAGE (van Beijnum et al., 2008). Elevated Grx1

170 has been reported in Alzheimer’s disease of post-mortem brains (Akterin et al., 2006).

Also, A-beta accumulation in cultured human neuroblastoma SH-SY5Y cells leads to oxidation of Grx1 and apoptosis (Akterin et al., 2006). Oxidized Grx1 is not able to bind to ASK in human prostate adenocarcinoma cells (DU-145), and this was correlated with

ASK-mediated apoptosis (Song and Lee, 2003). These studies suggest that reduced Grx1 prevents apoptosis via protein-protein binding to ASK. In a separate proteomics study, increased glutathionylation was reported for α-enolase, GAPDH, deoxyhemoglobin, and

α-Crystallin B, and concomitant inhibition of GAPDH and enolase activities were reported in AD brain (Newman et al., 2007). The increase in both Grx1 and glutathionylation reported in these two studies is an apparent inconsistency if conditions favor Grx1-mediated deglutathionylation. However, in the event that glutathionyl-radical generating conditions prevail, then Grx1 may promote glutathionylation of proteins

(Qanungo et al., 2007;Starke et al., 2003).

Figure 4.2 (pg. 195) summarizes the changes in Grx1 in these inflammatory diseases.

The findings of most studies support an increase in Grx1 with disease, but whether the increase in Grx1 is reflective of a homeostatic attempt to protect the body from physiological dysfunction, or an attribute of the disease progression is largely speculative. Collectively, the findings on glutaredoxin are not comprehensive enough to determine whether it would best serve as a therapeutic target or a therapeutic agent. It has been relatively recent that inflammation surfaced as a shared theme to such diverse diseases, and it will now be intriguing to discover how much overlap exists in the role of

Grx1 in each of them. Most likely therapeutic manipulation of Grx1 would have to be

171

disease specific and localized, because of its tissue- and cell-specific roles under different

conditions. For example, localized inhibition of Grx1 in the retina may block its pro-

inflammatory role in activation of the NFκB pathway. In contrast, Grx1 appears to

mediate glutathionylation and inactivation of NFκB in hypoxic pancreatic cancer cells

treated with NAC, suggesting that enhancing Grx activity in this case would promote

apoptosis of these cancer cells. Likewise, the paradoxical increase in Grx1 along with an

increase in protein glutathionylation (reported in separate studies) in Alzheimer’s disease

requires further studies to determine whether Grx1 is responsible for the increased

protein-SSG and whether these events contribute to disease progression, thus identifying

Grx1 as a target for inhibition. The ever evolving nature of humans and disease may

keep us forever chasing a satisfactory mechanism for proper inflammatory balance to

ward off infection while avoiding self-mutilation. Nevertheless, Grx1 regulated

glutathionylation likely represents one important mechanism by which inflammatory

pathways and signaling mediators are modulated, and thus further attention to this system

is warranted to discover novel means of therapeutic intervention.

Physiological and pathological relevance of Grx1 and glutathionylation can be

exemplified by the recent explosion in PubMed entries, even though the number of works

addressing the two with respect to one another has been limited (Fig. 1.1, pg. 42). For example, increased Grx activity was reported to modulate potassium channel gating in

diabetic rat hearts (Li et al., 2005), but whether glutathionylation of channel proteins is

the underlying mechanism of action was not addressed. Furthermore, the majority of

studies reporting protein glutathionylation have not tested its reversibility by Grx1, an

172 essential criterion for determining its potential as a regulatory mechanism. Future studies of Grx1 and glutathionylation will inevitably merge and shed light on sulfhydryl redox regulation of protein activity and cellular functions. Furthermore, according to the Center for Disease Control, the number of diagnoses for diabetes has doubled from 8 to nearly

16 million in the USA. This of course may be in part to a decrease in undiagnosed cases, but it is no secret that diabetes is surfacing in epidemic proportions, and therefore the elucidation of mechanisms within diabetes is of paramount importance.

Grx1 effects on neurons and regulation of potassium channels

The current studies have focused on the pro-inflammatory effects of Grx1 in retinal cells.

However, for full elucidation of the role of Grx1 in retinal glial cells and in the retina,

Grx1 regulation of many diverse processes must be taken into consideration. Retinal glial cells are the main support cell within the retina, providing a homeostatic nourishing and supportive environment for other cells within the retina. For example, the majority of the retinal pool of glutathione is found in the Müller cells (Pow and Crook, 1995), and

GSH has been reported to undergo an ischemia-induced transfer to neurons (Schutte and

Werner, 1998). Furthermore, the Müller cells serve as primary storage sites for glycogen, and metabolize glucose to lactate that can be transferred to neurons (Winkler et al.,

2000;Poitry-Yamate et al., 1995;Sarthy et al., 2004). They also release glutamate and

ATP which suppress activity of retinal neurons and help regulate their own cell volume in an autocrine fashion via potassium channels (Wurm et al., 2008). One of the defining features of the Müller cells is their uptake and redistribution of excess extracellular

173

potassium which regulates neuronal excitability (Newman and Reichenbach, 1996).

Furthermore, a decrease in potassium currents via subcellular distribution of potassium

channels and subsequent cell swelling of Müller cells has been linked to diabetic edema

(Pannicke et al., 2006). Therefore, it seems necessary to determine whether changes in

Grx1 affect the regulation of potassium and potassium channels in Müller cells and how

the retinal neurons are affected by these changes in Grx1 in the Müller cell.

Specifically, regulation of neuronal KATP channels by Grx1 would be intriguing since

these channels appear to mediate neuronal activation in response to glucose metabolism

of the glial cell (Burdakov and Ashcroft, 2002), and increased Grx1 in the hearts of

diabetic rats has been correlated with decreased potassium current via voltage gated

potassium channels (Li et al., 2005). Expression of inward rectifier potassium channels

(Kir4.1) is decreased in the perivascular membrane, and voltage-dependent fast transient

(type A) outward potassium currents are only found in Müller cells in the diabetic retina

(Pannicke et al., 2006). Analogous to the correlation between surface expression of

ICAM-1 and glutathionylation of ICAM-1 (Mukherjee et al., 2007), it would be

interesting to see whether glutathionylation and/or Grx1 could regulate membrane

shuttling of potassium channels. We have recently reviewed the evidence for Grx1 in

plasma, sputum, and extracellular milieu (Shelton and Mieyal, 2008), and Grx1-mediated

deglutathionylation of extracellular proteins is conceivable. However, it is not known

whether the other factors involved in catalytic turnover of Grx1 are also in the

extracellular milieu in sufficient quantities, i.e., GSH, GSSG reductase, and NADPH.

174

Significance of the work- The work presented in this thesis contributes to the overall understanding of the role of Grx1 in molecular and cellular processes. The most significant and exciting findings are that Grx1 is upregulated in retinal glial cells cultured in high glucose and in retinae from diabetic rats, Grx1 regulates NFκB signaling and subsequent production of pro-inflammatory proteins (IL-6 and ICAM-1) in retinal glial cells, and IL-6 mediates pro-inflammatory signaling from Müller cells to endothelial cells. Furthermore, these are the first reports that changes in Grx1 leads to increased secretion of the IL-6 cytokine from cells and that endogenous IKKβ from non-transfected and un-stimulated cells is S-glutathionylated. The regulation of NFκB via IKK-SSG by

Grx1, and the subsequent changes in pro-inflammatory mediators (IL-6 and ICAM-1) provide mechanistic insight into the regulatory processes within the cell and suggest that

Grx1 is a potential therapeutic target in diabetic retinopathy.

4.2 Future Directions

The data in this thesis provide insight into understanding the regulation of Grx1 and the regulation by Grx1 within cells and in the context of inflammatory disease, i.e., diabetic retinopathy. However, this study also has brought forth many questions. Future goals are to (1) elucidate whether additional proteins that are products of NFκB-driven transcription are regulated by Grx1; (2) determine the effects of Grx1-regulated products

(i.e. upregulation of ICAM-1 and secretion of IL-6) on cells; (3) characterize the specific response elements and potential co-activators for cytokine-mediated induction of Grx1;

(4) measure quantitatively the S-glutathionylation of IKKβ in cells over-expressing Grx1

175

in normal glucose medium or cultured in normal or high glucose medium; and (5)

analyze the effects of knocked-down Grx1 in the retinae of diabetic animals.

(1) Elucidate whether additional proteins transcriptionally driven by NFκB are regulated by Grx1- This thesis work demonstrates that Grx1 regulates specific transcriptional products of NFκB, namely ICAM-1 and IL-6. Expression of TNFα, IL-

1β, IFNα, and IL-10, were undetected in the medium of Grx1 over-expressing rMC-1 cells, and VEGF was detected but unchanged in response to Grx1 (Table 3.1, pg.144).

Proposals for studies focusing on the role of ICAM-1 and IL-6 in rMC-1 cells are discussed below. However, several additional NFκB-regulated cytokines such as IL-8 and MCP-1 are also potential targets for regulation by Grx1. It will be important to follow up on whether upregulation of Grx1 leads to increased secretion of these cytokines, analogous to that of IL-6.

IL-8 is a pro-inflammatory cytokine that is regulated by NFκB in retinal glial cells

(Yoshida et al., 1998). IL-8 is elevated in the vitreous of diabetic patients (Yoshida et al., 1998;Canataroglu et al., 2005;Murugeswari et al., 2008;Hernandez et al., 2005), and in the glial and endothelial cells of diabetic patients as demonstrated by immunohistochemistry of a section of donated retina (Yoshida et al., 1998). Also, IL-8 is secreted from primary guinea pig Müller cells (Malgorzata Goczalik et al., 2005).

Secretion of IL-8 likely is increased in rMC-1 cells over-expressing Grx1. However, only two standard assays (ELISA) for rat IL-8 protein have been found. One of these

176 assays has been discontinued, and the other is available but expensive (nearly $800, also see below).

Another important pro-inflammatory cytokine, monocyte chemoattractant protein-1

(MCP-1), is also elevated in the vitreous of diabetic patients (Murugeswari et al.,

2008;Hernandez et al., 2005), and MCP-1 from Müller glial cells has been reported to promote neuronal (photoreceptor) cell death via increased infiltration of macrophages

(Nakazawa et al., 2007;Nakazawa et al., 2006). Analysis of changes in MCP-1 secretion in Grx1 over-expressing rMC-1 cells will be helpful in understanding Grx1-regulated processes in the cell (also see below).

Furthermore, Grx1 has been shown to regulate the expression of the chemokines, macrophage inflammatory protein 2 and keratinocyte-derived chemokine, in primary tracheal epithelial cells (Reynaert et al., 2006).

Experimental Approach-In order to determine whether these cytokines are key players in regulation by Grx1 and high glucose, conditioned medium from rMC-1 cells over- expressing Grx1 or cultured in high glucose can readily be assayed by ELISAs for keratinocyte-derived chemokine (R & D Systems, cat RCN100), MCP-1 (GE Healthcare, cat # RPN2740), macrophage inflammatory protein 2 (Alpco Diagnostics, cat # 45-

MIPRT-E01), and IL-8 (USCNLIFE Science and Technology, cat # E0080r, $796).

177

In addition, an IL-6 neutralizing antibody (R & D, cat # AF506) added to the conditioned

medium from rMC-1 cells prior to its application to freshly cultured rMC-1 or endothelial

cells would test whether IL-6 is the primary essential mediator of Grx1-induced ICAM-1

expression. Similar ideas for neutralizing antibodies for IL-1β and MCP-1 have been

used previously in rMC-1 cell experiments (Vincent and Mohr, 2007;Nakazawa et al.,

2007).

Expected outcomes and alternative approaches: Since IL-6 is abundant in the conditioned medium, we expect IL-6 to be one of the predominant cytokines mediating the upregulation of Grx. Investigation of the protein content of additional candidate cytokines will provide data comparative to IL-6. Furthermore, direct treatment of cells with the cytokines (purified recombinant proteins would be used to determine whether the cytokine(s) account for the upregulation of Grx. If a particular cytokine(s) does not upregulate Grx to the same extent as that observed in rMC-1 cells cultured in conditioned medium, we would conclude that additional cytokines are involved. If the data from the

ELISA experiments do not reveal an increase in additional cytokines in the medium, it would seem that the extent of upregulation of these additional cytokines is sufficient to induce upregulation of Grx but below the experimental limit of detectibility. Therefore, cDNA of the cell lysate would be analyzed via microarray to test which cytokine genes

are being upregulated. Also, neutralizing antibodies directed towards the newly

identified cytokines would be incubated in rMC-1 cells cultured in conditioned medium

from Grx over-expressing rMC-1 cells, and the lysate would be analyzed for upregulation

of Grx. The neutralizing antibody for the cytokine(s) mediating Grx gene regulation

178

would be expected to inhibit increases in Grx. Alternatively, redox mediators such as

nitric oxide could be mediating the effects in the conditioned medium and would be

tested.

The current study has shown that conditioned medium from rMC-1 cells over-expressing

Grx1 leads to the upregulation of Grx1 and ICAM-1 in endothelial cells, and upregulation

of IL-6 was identified in the conditioned medium. If IL-6 is confirmed to be the primary mediator of this protein upregulation, subsequent mechanism(s) will need to be elucidated. For example, IL-6 could be leading to upregulation of proteins in the endothelial cells via the soluble IL-6 receptor (sIL-6R) provided by the rMC-1 cells or by membrane-bound IL-6R on the endothelial cells. If sIL-6R is the mediator, it would be interesting to discover whether or not Grx1 is involved in the expression and/or shedding of sIL-6R. A description of IL-6R signaling is described below.

The expression of the IL-6 receptor (IL-6R) once was thought to be restricted primarily to leukocytes, macrophages, and hepatocytes (Bauer et al., 1989a;Bauer et al.,

1989b;Baumann et al., 1990;Yamasaki et al., 1988) and cells of hematopoietic, fibroblastic, epithelial, and neural origin (Snyers et al., 1989). However, more recent studies have reported IL-6R in neurons (Nelson et al., 1999), endothelial progenitor cells

(Fan et al., 2008), and endothelial cells (Saura et al., 2006). IL-6 binds to IL-6R, forming a non-signal transducing complex which must collaborate with the non-ligand binding membrane glycoProtein Gp130 to transduce the IL-6 signal in cells (Taga et al.,

1989;Hibi et al., 1990). Understanding the orchestration of this signal transduction is

179

important since IL6 signaling seems to occur in many more cell types than IL6-R is

expressed, and this can occur when the IL-6 ligand is bridged to the ubiquitously

expressed gp130 in one cell type via a soluble form of IL6-R (sIL-6R) from a second type of cell (Taga, 1992;Rose-John et al., 2006). Whether this type of trans-signaling by the

Müller cells is involved in the upregulation of Grx1 and ICAM-1 in the endothelial cells

remains to be determined.

(2) Determine the effects of Grx1-regulated products on cells- Thus far, we have found

that Grx1 upregulates the protein expression of ICAM-1 and the cellular secretion of IL-

6. The effect of both of these events should be investigated and are discussed below.

Determine the effects of Grx1-upregulated IL-6 on retinal cells- The current study

suggests that Grx1-induced IL-6 mediates the pro-inflammatory effects of ICAM-1 on

endothelial and Müller cells. Other studies have suggested a neuroprotective role for IL-

6 in retinal microglial cells (Sappington and Calkins, 2006;Sanchez et al., 2003), rat

Müller cells (Seki et al., 2006), and neurons (Biber et al., 2008;Wang et al., 2007b). In

addition, IL-6 has been reported to have protective effects on hydrogen peroxide-treated

endothelial cells (Waxman et al., 2003). Elucidating directly the effects of IL-6 on the

individual cell types in the retina will help elucidate whether elevated IL-6 is deleterious

or beneficial to the retina.

Experimental approach: To fully elucidate the effects of Grx1-regulated IL-6 released

from Müller cells, different cell types each cultured in the conditioned medium from

180

Grx1 over-expressing rMC-1 cells would be analyzed for cell death. Namely, retinal

neuronal, Müller, and endothelial cells would be used. To isolate the effects of IL-6 in

the conditioned medium, an IL-6 neutralizing antibody would be used to block effects

attributed to IL-6 as mentioned above. Trypan blue is the quickest and most convenient method of assessing cell death, and would be used for initial analysis. For further

assessment, caspase activity, Hoechst staining, and DNA laddering would be used for

measuring apoptosis.

Expected outcomes and alternative approach: If IL-6 is the predominant protein released

into the medium from rMC-1 cells in response to Grx1 and it has a neuroprotective effect,

treatment with IL-6 would not lead to an increase in cell death. To test for a protective

effect, a stimulus such as light or hydrogen peroxide could be used to induce cell death in

cells cultured in conditioned medium from rMC-1 cells infected with adenoviral-Grx1 or

adenoviral-Empty vector. Treatment of cells with IL-6 or conditioned medium

containing elevated IL-6 would be expected to prevent cell death in response to these

stimuli.

Test whether the Grx1-regulated increases in ICAM-1 result in an increase in

extracellular membrane surface expression of ICAM-1 on rMC-1 cells- Testing whether

the increased ICAM-1 expression in lysates of rMC-1 cells over-expressing Grx1 or

grown in high glucose medium corresponds to an increase in extracellular exposed

membrane-bound ICAM-1 will help determine whether Grx1-induced ICAM-1 could be

involved in cellular signaling or leukostasis.

181

Experimental approach: As described in Chapter 2, pilot studies involving biotin labeling to detect cell surface expression of ICAM-1 were promising (Fig. 2.13, pg. 103 and Fig. 2.14, pg. 104), but must be replicated and extended. The finding that a ligand-

bound ICAM-1 complex can undergo endocytosis (Muro et al., 2005) raises the question of whether the experiments in Chapter 2 are reflecting membrane-bound ICAM-1-biotin or intracellular ICAM-1-biotin that has entered the cell via endocytic vesicles. To address this concern, membrane fractions can be isolated prior to biotin labeling and subsequent processing. Quantitative analysis of membrane-bound ICAM-1 would then need to be determined for rMC-1 cells cultured in normal glucose medium or high glucose medium; or rMC-1 over-expressing glutaredoxin in normal glucose medium.

Biotin is naturally present in all cells, and has the potential to give rise to high experimental background. A modification to the protocols outlined in Chapter 2 would be made to decrease non-specific signals. The source of biotin can be switched from EZ- link Sulfo-NHS-Biotin to EZ-Link NHS-SS-Biotin (Biosuccinimidyl 2-(biotinadmido)- ethyl-1, 3’-dithiopropionate, Pierce cat # 21441). Streptavidin will bind molecules labeled with these biotin compounds as well as endogenous biotin. However, the disulfide in NHS-SS-biotin allows for the molecule to be isolated by cleavage from the streptavidin-biotin complex with a reducing agent (e.g. DTT), leaving behind the endogenous biotin that is bound to streptavidin.

182

Expected outcomes and alternative approach: If there is not an increase in membrane- bound ICAM-1, then the role of ICAM-1 accumulated within the cell would be

addressed. It is possible that ICAM-1 is produced and stored in vesicles that get fused to

the surface membrane upon a second stimulus. However, it is likely that an increase in

membrane-bound ICAM-1 will be found exposed to the extracellular milieu. If this

hypothesis is correct, the role of membrane bound ICAM-1 on the surface of Muller cells would be pursued more aggressively. For example, ICAM-1 could serve as a receptor for intracellular signaling or as an anchor for white blood cell adhesion.

An alternative method for detecting the increased expression of adhesion molecules on the cell surface would be to measure attachment of white blood cells to rMC-1 cells over- expressing Grx1. Using this approach, the specific molecule responsible for the binding of white blood cells would be difficult to discern unless each adhesion protein was selectively knocked down. Alternatively, incubation of the rMC-1 cells with anti-ICAM-

1 antibody prior to incubation with the white blood cells would discriminate whether

ICAM-1 was responsible for binding of the white blood cells. If the changes in ICAM-1 are sufficient in cell culture conditions to bind other cells, this technique would demonstrate functional cell to cell physiology. Furthermore, if Grx1 increased the adhesion of white blood cells to Müller cells, this would help determine the role of Grx1 in inflammatory processes.

(3) Characterize the specific response elements and potential co-activators for cytokine- mediated induction of Grx1- The mechanisms of gene regulation of Grx are largely

183

unknown despite the advancements in this area over the past decade. The human Glrx gene and upstream promoter region have been sequenced from genomic DNA previously

(Park and Levine, 1997). Multiple promoter regions were identified as potential binding sites of transcription factors including SP1-1E and AP-1. However, many of these sites are located in promoter regions that exhibited decreased luciferase-driven promoter activity over that of similar but truncated promoter regions (Park and Levine, 1997).

Several transcription factors such as Nrf2 and the Jun and Fos components in AP-1 bind to an enhancer antioxidant response element (ARE) which mediates transcriptional

products in response to oxidative and inflammatory signals (Rahman, 2005). ARE is involved in the transcriptional regulation of many redox proteins including thioredoxins

and enzymes in GSH synthesis. There has been speculation that ARE might regulate

Grx. However, high glucose selectively upregulates Grx in rMC-1 cells while changes in

thioredoxin were not observed, suggesting that the regulation of Grx is not pleiotropic. In

a drug (MPTP)-induced model of Parkinson’s disease AP-1 appears to be involved in

upregulation of Grx mRNA in the brains of male mice only, but does not mediate the

elevated constitutive expression of Grx in female mice which have more estrogen

(Kenchappa et al., 2004). In H9c2 cardiomyocytes, estrogen-driven gene expression of

Grx is mediated by electrophoretic response element (EpRE)-like 1 element (Urata et al.,

2006). In that study, a related EpRE-like 2 element as well as AP-1 were found not to be

critical factors. Thus, EpRE-like 1 may be important in regulation of Grx in estrogen-

responsive tissues. However, much remains to be learned about the mechanisms of

transcriptional regulation of Grx.

184

Regulation of Glrx gene expression by NFκB has not been investigated previously but likely plays an important role. Increased cellular Grx leads to increased activation of

NFκB and a corresponding increase in its transcriptional product ICAM-1 (Shelton et al.

2007). Furthermore, Grx1 is upregulated in rMC-1 Müller cells cultured in conditioned medium from Grx over-expressing rMC-1 cells. This conditioned medium contains elevated IL-6 whereas TNFα, IL1β, IFNγ, and IL-10 proteins were not detected.

Concurrent with the increase in Grx, treatment with this conditioned medium leads to upregulation of ICAM-1. These data suggest that the conditioned medium containing IL-

6 is leading to activation of NFκB and further suggests that Grx could be transcriptionally regulated by NFκB. In addition, the conserved DNA sequence “GGGRNNYYCC” that binds NFκB is found in the promoter region of the Glrx gene that contained the highest luciferase activity of those tested (Park and Levine, 1997). Taken together, these data suggest that an autoregulatory feedback mechanism may exist for NFκB activation and

Glrx gene expression in retinal Müller cells. This mechanism is similar to that of other transcriptional products of NFκB such as TNFα that in turn leads to activation of NFκB.

Experimental approach: Using rMC-1 Müller cells, the mechanism of transcriptional regulation of the Glrx1 gene by cytokines would be investigated. As described above, analysis for increased expression of additional cytokines would be conducted on the conditioned medium from Grx over-expressing rMC-1 cells, and this conditioned medium leads to upregulation of Grx in freshly cultured rMC-1 or endothelial cells.

These findings would be compared to that of medium from rMC-1 cells cultured in normal or high glucose medium.

185

Grx mRNA will be measured to determine whether changes in translational regulation of

Grx accompany the increase in Grx protein in rMC-1 cells in high glucose medium.

Next, plasmids containing luciferase-driven promoter regions of the Grx gene would be constructed. rMC-1 cells would be transiently transfected with these constructs, cultured in conditioned medium from Grx over-expressing rMC-1 cells or from rMC-1 cells cultured in high glucose medium, and assayed for luciferase activity. Alternatively, transfected rMC-1 cells would be treated with purified cytokines such as IL-6 and assayed for luciferase activity. Thus, the DNA sequences that led to positive regulation of Glrx gene expression would be identified by the promoter constructs of Glrx exhibiting the highest luciferase activity.

To test the regulation of Grx by NFκB specifically, luciferase-driven promoter regions of the Glrx gene would be incubated with recombinant NFκB and analyzed for DNA binding via the EMSA. Next, the same constructs would be used to transfect rMC-1 cells over-expressing NFκB. Conversely, the luciferase activity in rMC-1 cells knocked down in NFκB would be measured. If the upregulation of NFκB is sufficient for regulation of

Glrx gene expression, mutations in the NFκB binding site would reveal the minimal essential sequence. Chromatin immunoprecipitation (CHIP) assays would also be used to determine the DNA sequence bound by NFκB.

Expected outcomes and alternative approaches: If NFκB is a primary regulator of Grx in the rMC-1 cells, enhanced luciferase activity of the Grx promoter regions will be

186 observed in NFκB over-expressing cells. However, it is possible that elevated NFκB is necessary but not sufficient to induce expression of Grx. NF-IL6 (also known as C/EBP proteins) and STAT3 are well known to mediate cellular transcriptional regulation in response to IL-6, and both of these proteins undergo significant protein-protein interactions with NFκB. For example, STAT3 binds to NFκB and they appear to synergistically activate the IL-6 gene (Yang et al., 2007). Therefore, NFκB will be immunoprecipitated from nuclear fractions of rMC-1 cells over-expressing Grx or cultured in high glucose medium, and the eluant will be analyzed by mass spectrometry to identify associated proteins. In addition, western blot analysis for upregulation of additional transcription factors such as STAT3, phospho-STAT3, and NF-IL6 will be conducted on nuclear fractions of cells over-expressing Grx or cultured in high glucose.

(4) Measure Quantitatively the S-glutathionylation of IKKβ in cells over-expressing

Grx1 in normal glucose medium or cultured in normal or high glucose medium- To determine whether changes such as the increase in ICAM-1 protein expression and IL-6 secretion are due to changes in the extent of glutathionylation of IKKβ, quantitative comparisons of the glutathionylation status of IKKβ in cells over-expressing Grx1 in normal glucose medium or cultured in normal or high glucose medium must be determined.

Experimental approach: Relatively pure and abundant IKKβ is needed for quantitative analysis of the glutathionylation status via mass spectrometry. Extensive optimization on the isolation of IKKβ from rMC-1 cell lysates has been done (See Chapter 3 and

187

Appendix). In addition to low yield from the immunoprecipitation, non-specific interactions (e.g. myosin) and interfering IgGs have been problematic. Mass spectral facilities most often prefer receiving a sample from an in-gel digest rather than one in solution because the protein band on the gel is typically more pure even if it leads to a lower yield of sample. Therefore, proteins in the lysate would be separated on SDS-

PAGE gels, and the gel bands at approximately 87 kDa would be excised and digested as done previously. To increase the accuracy of the band excision, standard IKKβ could be processed via SDS PAGE and visualized with coomassie or silver stain. Recombinant

His-IKKβ runs slightly higher than endogenous IKKβ (Fig. A1, pg. 203), but this would provide an estimation of the appropriate protein band/area. Steps such as adding heat and more washes should increase the percent of recovery of the proteins from the gel. Many in gel digests would be pooled. Alternatively, blue native gels could be used to isolate the intact complex of IKK.

Using the above technique to acquire sufficient IKKβ isolated from lysates (containing

IAM) of rMC-1 cells, several methods for quantitative mass spectrometry are then available, including metabolic labeling (SILAC), chemical labeling (ICAT and iTRAQ), enzymatic labeling (O18/O16), and the use of internal standards (AQUA). All of these techniques can give relative quantitation except for AQUA which measures absolute quantitations.

To start, isobaric tag for relative and absolute quantitation (iTRAQ) would be used to label primary amines of the peptides of IKKβ (DeSouza et al., 2005;Ross et al., 2004).

188

Here, IKKβ from cells cultured in normal glucose (NG) or high glucose (HG) medium would be digested separately, and then NG peptides would be labeled with one iTRAQ tag and HG-peptides would be labeled with a second iTRAQ tag (Applied Biosystems,

Forest City, CA). After labeling, the samples would be pooled and run on LC-MS-MS.

The spectrum from iTRAQ labeled peptides simplifies analysis over that of ICAT

(isotope-coded affinity tag) and SILAC (stable isotope labeling by amino acids in cell culture) because a single ion in the MS is observed for the peptides in the two samples.

Alternatively, the peptides of the one sample could be labeled with O16, and the other

with O18 (Miyagi and Rao, 2007). The advantage of this technique is that the labeling

and digestion occur simultaneously, reducing the variability in yield. Furthermore, work

on this technique of quantitative mass spectrometry has already begun via interaction with Dr. Masaru Miyagi (CWRU).

As its name implies, the only mass spectrometry technique that offers absolute

quantification is the newly developed AQUA (Gerber et al., 2003;Kirkpatrick et al.,

2005). First, two heavy isotope-labeled peptides (<15 amino acid) would be synthesized

by Sigma. Two peptides mimicking an amino acid sequence surrounding Cys179 of

IKKβ would be identical except that only one would contain the glutathionylation at

Cys179. The sequence surrounding Cys179 offers either a lysine or leucine to be used

for isotope labeling. The sample of isolated endogenous IKKβ would then be spiked

with a known amount of the labeled peptides that are used as internal standards, digested,

189

and sequenced via LC-MS-MS. The total amount of IKKβ and the extent of

glutathionylation can be determined from the resulting spectrum.

Expected outcomes and alternative approaches- The glutathionylation of IKKβ in cells cultured in high glucose medium is expected to be less than that in cells cultured in normal glucose medium because Grx1 is elevated by high glucose. Direct quantification of the glutathionylation of IKKβ is preferred. However, all of the above techniques

depend largely on the efforts of the mass spectrometer facilities and collaborations.

These experiments can be conducted, but may require a prolonged time frame and a more

extensive budget. Alternatively, changes in activity in IKKβ that result from changes in

the S-glutathionylation status can be measured. Pilot studies to measure the kinase

activity of IKKβ in lysates of rMC-1 cells in the presence or absence of exogenous

purified Grx1 have been initiated (see Chapter 3). Standard kinase assays rely on the

immunoprecipitation of IKK but to circumvent the limited yield of the

immunoprecipitations obtained from rMC-1 cells thus far (see Chapter 3), cell lysate

would be substituted for the IP. Initially, lysate (containing 5 mM NEM) from rMC-1

cells was incubated with [32P] ATP and an IKK substrate (GST-tagged IκB) in the

presence or absence of exogenous Grx1 and/or GSH. However, no Grx1-induced change

in IKK activity was detected because excess NEM in the lysate was inhibiting Grx1. We

have since found that treatment of lysate with 1 M hydroxylamine for 1 hr will prevent

inactivation of exogenous Grx1. Therefore, if Grx1 induced changes in the

glutathionylation of IKK leads to corresponding changes in kinase activity, those changes

should be detectable in lysates containing 5 mM NEM after treatment with 1 M

190

hydroxylamine for 1hr. To confirm kinase activity, a non-phosphorylatable mutant form of IκB will be used as a control.

Prior to conducting experimentation on cell lysates, recombinant IKKβ will be used to

test whether NEM or IAM alter its conformation such that activity is lost, regardless of its

glutathionylation status. Recombinant IKKβ (rIKKβ) will be glutathionylated using

radiolabeled oxidized glutathione ([35S] GSSG) (Fig. 4.3, pg. 196). Since in vitro

glutathionylation by oxidants can often force glutathionylation on cysteines which do not

become glutathionylated under physiological conditions, the rIKKβ protein will be

precipitated with ice cold TCA. Theoretically, glutathionylation on cysteine residues that

are not in the active site should not impact the activity of IKKβ. However, the net

negative charge of glutathione may induce conformational changes in the protein and

affect its activity. The protein pellet will be washed with cold TCA several times to wash

away any residual GSSG. Then, the protein pellet will be resuspended in NaOH, and

analyzed via scintillation counting. Based on the specific radioactivity of [35S] GSSG

and the protein content, a stoichometry of the [35S]-glutathionyl moiety to IKKβ can be

obtained. In addition, a DTNB assay will be used to test whether all the cysteines have become glutathionylated. If there are no free sulfhydryls retained on IKKβ, the time and

concentration of GSSG will be optimized such that not all cysteines are oxidized since

that would not mimic the status of the sulfhydryls on IKKβ in the cell. Then, NEM or

IAM will be added to IKKβ to block remaining sulfhydryl groups (Fig. 4.3 pg. 196).

Hydroxylamine will be used to rid excess NEM, as described above. Next, IKKβ will be

191

de-glutathionylated via treatment with pure Grx1, and assayed for its ability to

phosphorylate IκB, as described above.

(5) Analyze the effects of knocked-down Grx1 in the retinae of diabetic animals- The

induction of ICAM-1 in both the Müller and endothelial cell in response to medium from

Müller cells over-expressing Grx suggests that Grx would have an overall pro-

inflammatory effect in the retina.

Experimental approach: To test whether the suppression of Grx could diminish the lesions in diabetic retinopathy, Grx would be knocked down in the retina of diabetic rats by intravitreal injections of shRNA targeting Grx1 from adenovirus (Ad) vectors, as done previously to knock down VEGF in the eye (Cashman et al., 2006). Müller cells are highly susceptible to infection with Ad vectors (Weise et al., 2000;Di et al., 1998;Dudus et al., 1999). The diabetic rats would be injected with Ad-shRNA-Grx routinely for 1 month. Retinae will be excised, homogenized, and assayed for production of ICAM-1,

IL-6, and Grx1. In separate experiments, retinal vasculature will be isolated from retinae and assayed for TUNEL-positive cells and acellular capillaries. Also, caspase activation will also be measured in Müller cells and photoreceptors isolated from the retinae of diabetic rats compared to those from non-diabetic control rats.

Expected outcomes and alternative approaches: Knocking down Grx1 is expected to result in prevention or prolonged onset of lesions and markers of diabetic retinopathy, namely acellular capillaries, cell death, and increase expression of pro-inflammatory

192 proteins. If Grx1 knock-down results in an exacerbation or accelerated development of these events, additional analysis of the effects of changes in Grx1 in other retinal cells such as astrocytes, photoreceptors, and pericytes would be conducted.

Figure 1.1 (pg. 42) shows that there is an increase in Grx activity collectively in retinal cells of the diabetic rats. However, in order to elucidate which cells in the diabetic retinae are most affected by changes in Grx1 and have the greatest impact on diabetic lesions, Grx1 would be knocked down in a cell selective manner. For example, selective knock-down in the Muller cells may be possible utilizing gene regulatory elements for the glial intermediate filament protein (GFAP) gene or the cellular retinaldehyde-binding protein (CRALBP) gene.

193

Figure 4.1. Proteins involved in insulin secretion and signaling that are implicated for regulation by glutathionylation and Grx1. Proteins involved in the signaling pathway of insulin secretion are illustrated in Cell one. Signaling networks that can be activated by insulin signaling are depicted in Cell two. (Slightly modified from Mieyal et al. 2008).

194

Figure 4.2. Changes in protein glutathionylation and glutaredoxin in inflammatory

diseases. (Modified from Shelton and Mieyal, 2008).

195

Figure 4.3. Process of treating cell lysates in order to determine the amount of IKK activity attributable to de-glutathionylation. Cells will be lysed in 5 mM NEM (or 5 mM IAM) to block free sulfhydryls. Then the lysates will be treated with hydroxylamine to rid excess NEM that is not bound to protein. Finally, lysates will be incubated with

Grx1 and GSH, and subsequently used in an IκΒ kinase assay.

196

Figure 4.4. Modifications to sulfhydryls on IKKβ used to determine the activity of

IKKβ attributable to glutathionylation. Recombinant IKKβ (rIKKβ) will be glutathionylated with glutathione disulfide (GSSG). The remaining free sulfhydryls will be blocked with NEM (or IAM). NEM that does not bind to the protein will be quenched by the addition of hydroxylamine. IKKβ will be deglutathionylated by pure Grx1 and

GSH, and assayed via IκB phosphorylation.

197

Appendix

A.1 Approaches to quantitative analysis of the S-glutathionylation of IKKβ. Once the qualitative glutathionylation of IKKβ was established, quantitative analysis was pursued in order to establish whether the cellular and molecular changes (i.e., increased NFκB activation and ICAM-1 production) correspond to relative differences in the glutathionylation status of IKKβ in cells. We sought the expertise of Masaru Miyagi,

Ph.D. to pursue an O16/O18 labeling technique for quantitative comparisons of the glutathionylation of IKKβ in rMC-1 cells cultured in normal (5 mM) or high glucose (25 mM) medium. Key differences in the standard operating procedures of Dr. Miyagi’s group compared to the technique used by Dr. Distler include single enzyme digests with endoproteinase Lys C (abbreviated as Lys C) and analysis based on peptide matches and scoring to a database of Lys C cleaved peptides. Therefore, the initial goal was to re- establish the qualitative analysis of glutathionylation under these conditions.

Method of isolating IKKβ from cell lysate for mass spectral analysis: In order to maximize the amount of IKKβ immunoprecipitations were performed similar to that described in Chapter 3. Lysate from one 100 mm dish of cells was used for each immunoprecipitation. The eluate from eight immunoprecipitations was pooled for analysis (n=1). This sample was also denatured with 8 M urea prior to digestion as recommended by Dr. Miyagi.

198

Results of mass spectral LC-MS-MS analysis: Albumin was the predominant protein identified in the eluate from the immunoprecipitations of IKKβ. In a subsequent experiment, an immunoprecipitation with lysate from four 150 mm dishes of cells was analyzed without the urea treatment. Myosin was the protein with the highest matching score to the rat database of proteins cleaved with Lys C.

The standard requirement for mass spectral analysis by Dr. Miygai’s group is a purified protein that is observable on a coomassie stained gel, and the limit of detectibility of the coomassie staining is approximately 100 ng protein. Based on the estimated yield of

IKKβ from the immunoprecipitations (32 ng IKKβ per IP, Fig. A1, A, pg. 203), 4 immunoprecipitations with lysate from four 150 mm dishes of cells each would be pooled together to achieve a yield of at least 100 ng of IKKβ, a reasonable amount of working material. Ηοwever, the apparent abundance of non-specific proteins (e.g. Myosin) in the immunoprecipitations creates a dynamic range problem that interferes with the detection of low abundance peptides, including IKKβ and its glutathionylated form. Therefore, several alternative methods were explored to increase the purity and abundance of the

IKKβ isolated from rMC-1 cells for quantitative mass spectral analysis, and will be discussed below (see 1a-e, see below). As an alternative to mass spectrometry, radiolabeled glutathione was tested as a means of detecting glutathionylation of IKKβ immunoprecipitated from cells (see 1f, see below). Furthermore, methods to detect the glutathionylation of IKKβ within lysates of rMC-1 cells were pursued (2a-c, see below).

199

(1) Pursuit of quantitative analysis of S-glutathionylation of IKKβ isolated from rMC-1

cells, and approaches to improved isolation of IKKβ. Studies involving analysis of

glutathionylation require special care. For example, typically, immunoprecipitated

proteins are eluted from the beads by treatment with a reducing agent such as DTT.

However, in addition to the disulfide bonds in the antibodies, DTT also reduces the protein glutathione mixed disulfide bonds. For this reason, DTT must be avoided in the elution of proteins that are processed for subsequent analysis for glutathionylation. In

addition, cells must be lysed in a thiol-blocking reagent (e.g. IAM or NEM) to prevent artificial thiol exchange or oxidation on free sulfhydryls during processing.

1a) Optimization of the immunoprecipitation of IKKβ from lysates of rMC-1 cells: A

series of modifications to immunoprecipitation protocol were tried in order to maximize

the amount of IKKβ isolated from cell lysates. The goal of the immunoprecipitations

(IPs) was to isolate IKK in order to subsequently detect and quantify the amount of S-

glutathionylation.

Methods of the optimization of the immunoprecipitation of IKKβ from lysates of rMC-1 cells: Immunoprecipitations were conducted as described in Chapter 3. Cells were lysed in 1% Triton X-100 lysis buffer containing either 5 mM IAM or 5 mM NEM. The BCA assay was used to determine the total protein content of the lysates.

Immunoprecipitations were incubated with 1 µg IKKγ over-night at 4°C and eluted with

1X sample buffer with or without 5 mM DTT. The standard use of DTT to elute the immunoprecipitated protein from the bead could not be employed for analysis of

200 glutathionylation. Therefore, elution by DTT was used only in the troubleshooting and optimization stages. Furthermore, lysate for immunoprecipitations must be generated in the presence of a thiol blocking agent, and for these purposes, 5 mM NEM or 5 mM IAM in the lysate were compared. Additional steps to optimize the yield of the immunoprecipitations are highlighted below.

• IKKβ was immunoprecipitated from cell lysate of various amounts of total

protein. Cells from different numbers and sizes of culture dishes were pooled,

centrifuged, and lysed in different amounts of lysis buffer. In one experiment,

IKKβ was immunoprecipitated from 2 mg protein lysate in 0.5 ml lysis buffer, 4

mg protein lysates in 0.5 ml lysis buffer or 6 mg protein lysate 0.5 ml lysis buffer

and compared via Western blot analysis. In a separate experiment, IKKβ was

immunoprecipitated from 0.5 ml lysate from 1, 2, 4, or eight 100 mm dishes of

cells and compared to that from eight 100 mm dishes of cells in 2 ml of lysis

buffer and analyzed via Western blot analysis. Additionally, Western blot

analysis was conducted on IKKβ immunoprecipitated from one 150 mm dish

lysed in 0.5 ml or from four 150 mm dishes lysed in 0.5 ml.

• Lysate was pre-cleared with either 60 µl 1:2 slurry of Protein A/G PLUS beads

alone or 60 µl 1:2 slurry of Protein A/G PLUS beads and 1 µg IgG for 1 hr at 4°C.

• The amount of PBS used to wash the beads after immunoprecipitation but prior to

elution was changed from 1.5 ml to 0.4 ml.

• Immunoprecipitations with Protein A/G PLUS beads were compared to those with

Protein G beads.

201

• The amount of immunoprecipitating anti-IKKγ antibody was increased from 1 µg

to 2 µg.

• Protein A/G PLUS beads and anti-IKKγ antibody were pre-bound by an overnight

incubation at 4°C prior to incubation with lysate.

• IgG (Santa Cruz Biotechnology rabbit sc-2027) or anti-IKKγ antibody (Santa

Cruz Biotechnology, sc8830) was covalently cross-linked to Protein G Plus beads

via the Seize X Mammalian Immunoprecipitation Kit according the

manufacturer’s instructions (Pierce, cat # 45225). Briefly, the IgG or IKK

antibody was incubated with beads for 1 hr. Absorbance readings demonstrated a

10-fold decrease in A280 of the IgG or antibody after incubation with the Protein

G beads, suggesting ample binding to the beads. The IgG or antibody was then

cross-linked to the beads using disuccinimidyl suberate, a non-cleavable cross-

linking agent. The immobilized anti-IKKγ was subsequently used to

immunoprecipitate IKKβ from cell lysate.

Results of the optimization of the immunoprecipitation of IKKβ from lysates of rMC-1 cells: The success of a Western blot of immunoprecipitated IKKβ was found to be

highly dependent on the individual sulfhydryl modifying agents in the lysate and whether

DTT was present during the elution from the bead. Immunoprecipitated IKKβ was

detected when the lysate contained IAM regardless of whether DTT was present during

the elution (Fig. A1, B, pg. 203). It also could be detected in the absence of thiol

modifying agents but only if DTT was present during the elution (Fig. A1, B, pg. 203).

202

Figure A1. The estimated yield and the effects of sulfhydryl modifying reagents on

the immunoprecipitation of IKKβ from lysates of rMC-1 cells. IKKβ was co- immunoprecipitated with IKKγ from lysates with or without IAM or NEM. The immunoprecipitated IKKβ was eluted from the beads with 50 µl 1X sample buffer with or without DTT. Western blot analyses using anti-IKKβ (1:1000 dilution) were conducted on the lysate and the immunoprecipitations. Cell pellets from four 150 mm dishes of rMC-1 cells were resuspended in 0.6ml of lysis buffer and used for the immunoprecipitations. Compared to recombinant IKKβ standards, the immunoprecipitation of IKKβ yielded approximately 32 ng (A). IKKβ was detected in the immunoprecipitations from lysate containing IAM, but could not be detected when the lysate contained NEM irrespective of whether DTT was present in the elution (B,

203 lanes 4 and 6). IKKβ was not detected in the immunoprecipitations from lysate in the absence of NEM or IAM if DTT was omitted from the elution (B, lane 2). From whole cell lysates containing either IAM or NEM, IKKβ could be detected regardless of whether DTT was used in the elution (B, lanes 5 and 7). However, in the absence of either sulfhydryl blocking reagent, IKKβ could be detected only when it was eluted with

DTT (B, lane 3).

204

In order to further maximize the amount of IKKβ immunoprecipitated from the lysates, immunoprecipitations were conducted with lysates containing different protein concentrations.

Methods for immunoprecipitation of IKKβ from lysate with different protein content:

The protein concentrations of the lysates were made to be different by pooling and centrifuging different amounts of cells and lysing the cell pellet in different amounts of lysis buffer. rMC-1 cells from one, two, four, or eight confluent 100 mm culture dishes were pooled and centrifuged. The cell pellets were lysed in 0.5 ml lysis buffer. Cells from one 100 mm dish equate roughly to 2 mg protein. The lysate was then used for the immunoprecipitation of IKKβ.

Results of the immunoprecipitation of IKKβ from lysate with different protein content:

Western blot analysis revealed that the yield of IKKβ generally increased with increasing concentrations of cell lysate. For example, the amount of IKKβ immunoprecipitated from lysate from cells in eight 100 mm dishes and lysed in 0.5 ml lysis buffer was significantly increased over that from lysate from the same number of cells lysed in 2 ml lysis buffer (Fig. A2-1, A, pg. 207). Furthermore, the yield of IKKβ of an immunoprecipitation from four pooled 150 mm dishes of cells lysed in 0.5 ml lysis buffer was compared to that from four pooled 150 mm dishes of cells lysed in 0.5 ml lysis buffer and 0.5 ml PBS (Fig. A2-2, compare lanes 3 and 4, pg. 212). These data indicate that the yield of the immunoprecipitation is dependent on the protein concentration of the cell lysate. According to technical literature, detergent can interfere with efficient

205

immunoprecipitation, and it is possible that the efficiency of the immunoprecipitation could be improved by decreasing the amount of detergent from 1% to 0.5% so long as the efficiency in the lysis of the cell was not decreased substantially.

In a similar experiment, rMC-1 cells from one, two, or three 100 mm dishes were pooled, centrifuged, and resuspended in 0.8 ml of lysis buffer. Lysate from cells pooled from one, two, or three 100 mm dishes corresponded to nearly 3 mg, 5 mg, and 7 mg of total protein, respectively. IKKβ was immunoprecipitated from each sample of protein lysate and compared to the lysate before and after immunoprecipitation of IKKβ (Fig. A2-1, B, pg. 207). This experiment led to several interesting findings; (1) IKKβ is still plentiful in the lysate after the immunoprecipitation (2) actin binds non-specifically to the Protein

A/G PLUS beads in a protein concentration-dependent fashion, and (3) GAPDH, another protein in high abundance in the cell, does not bind to Protein A/G PLUS beads.

In subsequent studies, four individual immunoprecipitations from lysate of cells from one

100 mm dish lysed in 0.5 ml of lysis buffer each were pooled. The detection of immunoprecipitated IKKβ was increased over that of an individual immunoprecipitation from lysate from cells from one 100 mm dish lysed in 0.5 ml lysis buffer. The yield of

IKKβ was further increased in immunoprecipitations from lysate from cells from one 150 mm lysed in 0.5 ml lysis buffer. Four pooled immunoprecipitations from lysate of one

150 mm dish lysed in 0.5 ml gave rise to additional IKKβ.

206

Figure A2-1. Optimization of the amount of lysate used in the immunoprecipitation

of IKKβ from lysate of rMC-1 cells. rMC-1 cells from one, two, four, or eight 100 mm

dishes were pooled and resuspended in 0.5 ml lysis buffer. Alternatively, eight 100 mm dishes of cells were pooled and resuspended in 2 ml lysis buffer. These lysates then were used to co-immunoprecipitate IKKβ using an anti-IKKγ antibody. IKKβ was eluted from the beads with 50 µl 1X sample buffer containing 5 mM DTT, and immunoprobed using an anti- IKKβ antibody. Τhe amount of IKKβ immunoprecipitated from the lysates increased with increasing amounts of lysate in 0.5 ml lysis buffer (A). However, more

IKKβ was immunoprecipitated from lysate from eight 100 mm dishes of cells in 0.5 ml lysis buffer than that from eight dishes of cells in 2 ml lysis buffer (A). Then, IKKβ was immunoprecipitated from 2, 4, or 6 mg of total protein in 0.8 ml protein lysate and

207 compared to 40 µl of cell lysate before (Pre-IP) and after the immunoprecipitation (Post-

IP). Forty µl corresponded to 131 µg in the lysate containing 2 mg total protein, 240 µg in the lysate containing 4 mg total protein, and 333 µg in the lysate containing 6 mg total protein. The yield of immunoprecipitated IKKβ increased slightly with increasing protein concentration (B). Actin contamination was found in the immunoprecipitation in a concentration-dependent fashion, but GAPDH was not found in any of the immunoprecipitations (B).

208

The following attempts to optimize the yield of the immunoprecipitations were made using lysate from cells pooled from four 150 mm dishes and lysed in 0.8 ml. Pre-clearing the lysate with Protein A/G PLUS beads plus IgG decreased the non-specific signals, but pre-clearing the lysate with Protein A/G PLUS beads or Protein A/G PLUS beads plus

IgG did not enhance the amount of IKKβ immunoprecipitated (n=1) (Fig. A2-2, compare lane 3 with lanes 8 and 9, pg. 212). Unexpectedly, the eluate from the beads used to pre- clear the lysate (e.g. no anti-IKK antibody was present) gave a robust signal for IKKβ

(n=1) (Fig. A2-2, lanes 6 and 7, pg. 212). These data are quite surprising and would need to be repeated to confirm and attempt to understand this result.

Actin non-specifically bound to the beads in the presence of anti-IKK antibody (Fig. A2-

1, B, pg. 207), and myosin and albumin were the predominant proteins found in the immunoprecipitation via mass spectrometry analysis (see above). IKK is not as abundant in the cell as these proteins, but in samples of such high protein concentrations IKK could be non-specifically binding the beads or aggregating with an abundant protein like actin which is binding to the beads non-specifically.

Further optimizations of the immunoprecipitation were attempted using lysate from two

150 mm dishes of cells lysed in 0.6 ml lysis buffer. The amount of IKKβ from immunoprecipitations using Protein A/G PLUS beads compared to Protein G beads were similar with a slight increase in yield of IKKβ with Protein A/G PLUS beads (n=1) (Fig.

A2-2, compare lane 13 with lane 11, pg. 212). The amount of PBS used to wash the beads after immunoprecipitation but prior to elution was changed from 1.5 ml to 0.4 ml,

209

and this change resulted in a small but worthwhile increase in the yield of IKKβ (n=1)

(Fig. A2-2, compare lane 13 with lane 12, pg. 212). Doubling the amount of anti-

IKKγ antibody to 2 µg also did not increase the yield of immunoprecipitated IKKβ (n=1)

(Fig. A2-2, compare lane 3 with lane 10, pg. 212). It is possible that the anti-IKK antibody bound to cellular IKK but did not get bound to the beads, and therefore was lost in the washes of the immunoprecipitating beads. The range of primary antibody recommended by the manufacturer for these beads is 0.2 µg-2 µg of primary antibody.

Though, lysates that are high in protein concentration such as the ones used in these experiments may decrease the efficiency of the binding of the antibody to the beads. Pre- binding the Protein A/G PLUS beads and anti-IKKγ antibody in a separate overnight incubation did not increase the yield of the immunoprecipitations of IKKβ over immunoprecipitations where each were added simultaneously to the lysate (i.e. standard

IP) (n=1) (Fig. A2-2, compare lanes 13 and 15, pg. 212). Likewise, Protein G beads pre- bound to the antibody did not increase the yield of the immunoprecipitations (Fig. A3, compare lanes 13 and 14, pg. 218). Furthermore, anti-IKKγ antibody covalently bound to

Protein G PLUS beads using a Seize X immunoprecipitation kit was relatively unsuccessful (Fig. A2-2, compare lanes 3 and 5, pg. 212). This was an unexpected result, but is likely due to a conformational change of the antibody, interfering with its affinity for the antigen. There was a 10-fold decrease in A280 nm readings of the IKKγ antibody subsequent to incubation with the beads, suggesting that inefficient antibody binding to the beads was not problematic.

210

In addition, serial incubations of Protein A/G beads, anti-IKK antibody, and fresh lysate from rMC-1 cells from one 150 mm dish lysed in 500 µl lysis buffer were tested to increase the yield of the immunoprecipitations. The immunoprecipitation from three serial incubations of 3 hrs were compared to an immunoprecipitation from three serial incubations followed by an over-night incubation and an immunoprecipitation of an overnight incubation alone. Fresh antibody was added each time the beads were incubated with fresh lysate. No increase in yield of IKKβ was observed.

211

Figure A2-2. Optimization of the immunoprecipitation of IKKβ from lysates of rMC-1 cells. IKKβ was immunoprecipitated from 0.8 ml lysate from four 150 mm dishes of rMC-1 cells (lanes 1-10) or from 0.625 ml (7.3 mg protein) lysate from two 150 mm dishes of the same cells (lanes 11-15). In all samples, the beads were eluted with 50

µl 1X sample buffer, and this elution was loaded into the gel. Various modifications of the immunoprecipitation were made to improve the yield. Lanes 1 and 2 contain lysate in

NP40 and triton lysis buffers, respectively. Lane 3: Typical IP conditions with 1 µg anti-

IKKγ precipitating antibody in 0.5 ml lysate from four 150 mm dishes of cells. Lane 4:

IP of 0.5 ml lysates from four 150 mm dishes of cells diluted in half with PBS. Lane 5:

IP using beads covalently bound to anti-IKKγ. Lane 6: Lysates were pre-cleared with

Protein A/G PLUS beads, and the 50 µl elution of the pre-clearing beads was loaded into

212

the gel. Lane 7: Lysates were pre-cleared with Protein A/G PLUS beads and 1 µg IgG,

and the 50 µl elution of the pre-clearing beads was loaded into the gel. Lane 8: IP from

lysate that was pre-cleared with Protein A/G PLUS beads. Lane 9: IP from lysate that was pre-cleared with Protein A/G PLUS beads and 1 µg IgG. Lane 10: IP with 2 µg anti-IKKγ precipitating antibody. Some steps led to a reduction in background on the

Western blots (compare lanes 3 to lanes 9 and 10). However, none of the changes to the standard IP dramatically improved the efficiency of the immunoprecipitation of IKKβ.

IP, immunoprecipitation.

213

(1b) Approach to separate proteins in lysate of rMC-1 cells via size exclusion column

chromatograph. Proteins in cell lysates were separated by size using column

chromatography in order to isolate IKKβ from the majority of the other proteins.

Methods for column preparation and calibration: Prior to separation of cellular proteins on a size exclusion column, the column was prepared and calibrated. G150 Sephadex in

10 mM Tris-HCl pH=8 was used to fill a 70 cm, 344 ml glass gravity-flow column. The

column was calibrated with a protein mixture containing 4 mg blue dextran (2,000,000

Da), 8 mg hemoglobin (68,000 Da), and 4 mg aprotinin (6,500 Da). The column effluent was collected in fractions with an automated fraction collector. The proteins were identified in the fraction volumes based on color and A280 nm readings.

Methods for separation of protein lysate on the column and subsequent analysis of the

fractions: Lysate was prepared from eight 150 mm dishes of rMC-1 cells lysed in 1 ml

1% NP40 lysis buffer. The final volume of the lysate was approximately 1.3 ml. Lysate

was applied to the G150 Sephadex column, and 1-2 ml fractions were collected and

analyzed for protein via A280 nm readings. 225 µl of each fraction was loaded onto a large 3 mm, 12% SDS-PAGE gel for Western blotting analysis for IKKβ. Alternatively,

fractions were analyzed via slot blotting with an anti-IKKβ antibody on nitrocellulose

membrane according the manufacturer’s protocol (Bio-Rad; BioDot SF Microfiltration

Apparatus). These analyses were conducted to determine which fractions contained

IKKβ. Fractions were pooled and concentrated.

214

Methods for concentrating IKKβ: Different methods of concentrating IKKβ were

compared using approximately 800 ng rIKKβ in 10 ml 10 mM Tris buffer and

concentrated to 200 µl: stirred concentrator under nitrogen pressure (mw cutoff 10,000

Da), spin concentrators (mw cutoff 10,000 Da), polyethylene glycol (PEG, mw 30,000

Da) applied to a dialysis tube containing the sample, and speed vacuum dehydration.

Alternatively, fractions were pooled and concentrated by precipitation with 20% ice cold

tricholoroacetic acid (TCA), and the protein pellet formed after centrifugation was re-

dissolved and neutralized with sodium hydroxide. Recovery of IKKβ from the

concentration procedures was analyzed via slot blotting and compared to the initial un-

concentrated sample.

Results of the column calibration: Proteins in a calibration mixture (dextran, mw 2,000

kDa; hemoglobin, mw 68 kDa; and aprotinin, mw 6.5 kDa) were separated on a G150

Sephadex column. The analysis of the separation of these proteins in fractions collected from the column predicted that the elution of IKKβ from cell lysate would occur at 150

ml of eluant from the time the sample is loaded on the column.

Results of protein lysate separated by the column: Western blotting analysis of column

fractions from cell lysate showed the presence of IKKβ at approximately 20-30 ml into

the flow through. Because IKKβ was eluted from the column much earlier than predicted

for the 87 kDa molecular weight of IKKβ based on the calibration curve, it seemed as if

IKKβ might be complexed to other proteins (e.g. IKKγ, IKKα, etc.) when eluted from the

column. IKKβ was not detected as a larger molecular weight complex in the Western

215

blot analysis likely because the SDS in the sample buffer and gel dissociated the complex

into its respective subunits. Therefore, follow-up experiments were conducted with 8 M

urea in the lysis buffer in order to denature the high molecular weight complexes

containing IKKβ prior to its elution from the column. However, the presence of IKKβ

seemed to become more diffuse in the column fractions with urea treatment.

Results for concentrating IKKβ: In order to discover the optimal method for concentrating IKKβ in fractions from the column, a mock experiment was conducted whereby recombinant IKKβ was concentrated via Amicon stirred concentrators driven by pressure of nitrogen gas, spin concentrators that undergo centrifugation, dialysis tubing containing the sample overlayed with PEG, or speed vacuum. The stirred concentrator and spin concentrators were the preferred method of concentration because samples maintained a constant concentration of salts and low molecular weight species, while concentrating the protein of interest according to differential ultrafiltration (nominal molecular weight exclusion 10 kDa).

The concentrated standards were compared to a standard curve for IKKβ on a slot blot

(Fig. A 3, A and B, pg. 219). The standards increased linearly up to about 15-20 ng at

this exposure of film, and the higher end of the standard curve was linear on film that was

exposed for less time. A 200 µl aliquot of the 10 ml unconcentrated starting material was

estimated to have 8-9 ng IKKβ (Fig. A3, A, pg. 218), and predicted a total of 400 ng in

the 200 µl of concentrated samples. The stirred concentrators yielded no detectable

216

IKKβ, and the three additional concentration methods gave a 3-12% yield of IKKβ (Fig.

A3, A, pg. 218).

Results for concentrating column fractions: When cell lysate was separated by gel filtration chromatography, fractions were pooled and concentrated from approximately

144 ml to 400 µl. This degree of concentration of the protein solution led to protein aggregates that were not seen in the concentration of IKKβ standards. Treatment of the

concentrated sample with cold 20% TCA or formic acid and subsequent neutralization

with sodium hydroxide or ammonia, respectively, disrupted the aggregates and created a

soluble sample. Fractions from two columns (i.e. lysate equivalent to a total of 16-150 mm dishes of cells) led to a yield of roughly 50 ng IKKβ.

In an effort to increase this yield of IKKβ, an alternative means of concentration using

TCA was pursued. The proteins in the pooled fractions from two columns were

precipitated with cold 20% TCA, redissolved with sodium hydroxide, and analyzed on a

dot blot to yield roughly 80 ng IKKβ. This approach gave the best yield among those

tested.

217

Figure A3. Slot blot comparison of recombinant IKKβ that was concentrated by multiple techniques and the quantification of a standard curve for IKKβ. Approximately 800 ng rIKKβ in 10 ml 10 mM Tris buffer was concentrated to 200 µl by multiple means: (a) a nitrogen pressurized stirred concentrator, (b) a centrifugation filter (spin cup), (c) speed vacuum, or (d) dialysis with polyethylene glycol (PEG, mw 30,000 Da) on the outside of the tubing. All membranes had a molecular weight cut off of 10,000 Da. Concentrated samples were applied to a slot blot alongside rIKKβ standards and a portion of the unconcentrated sample; the blot was then immunoprobed for IKKβ. The stirred concentrator was the only concentration device that did not yield any IKKβ (A). Spin concentrators yielded approximately 35 ng or a 9% recovery (A). Only ½ of the concentrated sample was loaded from the PEG dialysis due to the viscosity of the sample, and this technique yielded about 25 ng or a 12% recovery (A). Speed vacuum gave about 10 ng or a 3% recovery (A). Quantification of the rIKKβ standard curve demonstrated a linear pattern up to approximately 15-20 ng at this exposure of film.

218

(1c) Approach to in-gel digest a radiolabeled standard, [35S] BSA-SSG. In order to

determine the optimal method for isolating IKKβ from cell lysates, several techniques

were pursued. However, in order to determine quickly and efficiently the percent yield of

the in gel digest, radiolabeled [35S] BSA-SSG was used instead of cell lysates (IKKβ).

Methods for processing in gel digestions of [35S] BSA-SSG: [35S] BSA-SSG (100 µg per

lane) was processed via SDS PAGE, and the BSA in the 12% gel was stained overnight

with coomassie blue dye. The coomassie stained bands that corresponded to BSA (i.e.

approximately 65 kDa) were excised. The excised bands containing [35S] BSA-SSG

were then put directly into scintillation fluid, but no value (counts per minute, cpm)

above background was detected. Therefore, [35S] BSA-SSG was trypsin-digested in the gel, and the corresponding peptides of [35S] BSA-SSG were eluted and analyzed via

scintillation counting. For digestions, gel pieces were first destained by incubation in

acetonitrile (ACN) for 30 min followed by two washes of 25 mM ammonium bicarbonate

and 50% ACN for 1 hr per wash. [35S] BSA-SSG was then rehydrated in 25 mM

ammonium bicarbonate and digested for at least 24 hr in trypsin (Sigma T6567) at

approximately 0.1 μg/15 μl in 25 mM ammonium bicarbonate. This solution of trypsin was collected, and the gel pieces were incubated in 0.1% trifluoroacetic acid (TFA) and

60% ACN multiple times for at least 30 min per incubation. These solutions of ACN-

TFA were pooled with the solution of trypsin. Gels were treated with an additional wash with 100% ACN, and this solution was pooled with previously pooled solution of ACN,

TFA, and trypsin. The pooled solutions were dried with a speed vacuum, dissolved in

0.1% TFA and 5% ACN, and analyzed via liquid scintillation counting.

219

Results of in gel digestions of [35S] BSA-SSG: In order to project whether in-gel digests of proteins migrating at 87 kDa on an SDS page gel would give an abundant yield of

IKKβ isolated from lysate of rMC-1 cells, coomassie bands of [35S] BSA-SSG were excised and extracted via in-gel trypsic digestions. Scintillation readings gave a range of

0.8- 41% yield with an overall average of 11% yield (n=14). This amount of recovered sample might produce sufficient IKKβ from cell lysates if several digests were pooled.

Yield was variable in these experiments, but is likely to improve with practice and with some slight modifications to the current protocol (see below). However, unlike the BSA used in the pilot studies, IKKβ at best may give a very faint band upon coomassie staining, which is difficult to precisely excise out of the gel without inclusion of additional interfering protein(s). Theoretically, IKKβ standards could be used to accurately identify the band migration. However, recombinant IKKβ standards are His- tagged and migrate slightly above IKKβ in cell lysates (Fig. A1, A, pg. 203). Some modifications to the current protocol may render this technique more useful for isolating

IKK for analysis of glutathionylation. For example, the yield of the in-gel digest might be improved by heating the gel to 37°C during the trypsic digest, increasing ACN washes after the digest, and processing the digest in a sonicating water bath.

(1d) Approach to electro-elute a radiolabeled standard, [35S] BSA-SSG. Similar to the in- gel digestion experiments above, radiolabeled [35S] BSA-SSG was used as a substitute for protein lysate (IKKβ) in order to quickly determine the percent yield of an electro- elution of a protein (BSA) from an SDS PAGE gel.

220

Methods for the electro-elution of [35S] BSA-SSG: [35S] Radiolabeled BSA-SSG (100 µg

per lane) was processed via SDS PAGE. Protein (BSA) in the 8% SDS PAGE gel was stained overnight with coomassie blue dye. Ten coomassie stained gel bands

corresponding to BSA (approximately 65 kDa) were excised, minced, and pooled into a

single glass tube of an electro-eluter (Bio-Rad Model 422). Proteins were electro-eluted in 0.1% SDS ammonium bicarbonate buffer at 30 mAmp through membrane caps with a

12,000-15,000 mw cutoff. Electro-eluted [35S] BSA-SSG was then analyzed via liquid

scintillation counting.

Results of the electro-elution of [35S] BSA-SSG: In order to predict whether the electro-

elution of proteins migrating at approximately 87 kDa on an SDS page gel would give an

abundant yield of IKKβ isolated from lysate of rMC-1 cells, coomassie gel bands of [35S]

BSA-SSG were excised and electro-eluted. Only 1.5% of [35S] BSA-SSG was recovered

when the electro-eluted sample was destained with 1mM HCl and 80% acetone (n=1).

However, a 45% yield of [35S] BSA-SSG was obtained when the coomassie was

destained prior to electro-elution, although complete removal of the stain was not achieved (n=1). Electro-elution of IKKβ from lysates of rMC-1 cells may be useful in

some cases, but it has similar limitations as the in-gel digests (e.g. interfering proteins

from the gel).

(1e) Attempt to partial purify IKKβ from lysates of rMC-1 cells. In order increase the

amount of IKK on the gel when loading 100 µg total protein lysate, lower molecular

weight proteins were removed from the lysate.

221

Methods for the partial purification of cell lysate: In order to partially purify protein

lysate, low molecular weight proteins were filtered from the mixture. rMC-1 cells from

six 100 mm dishes were lysed in 420 µl NP40 lysis buffer, and yielded a total volume of

500 µl and 13.5 mg of protein. Approximately 12 mg of that protein lysate was diluted to

1 ml in lysis buffer. The lysate was then divided in half and centrifuged in two Amicon

ultra YM50 spin concentrators (mw cutoff 50,000 Da) (500 µl lysate per cup).

Additional buffer was added to the lysate once the initial buffer began filtering out of the lysate. The filtered lysates were pooled and appeared to contain white aggregates of precipitated proteins. The filtered lysate was cleared of precipitates by centrifugation for

1 min at 6.5 x g, and 100-400 µg of this lysate was compared to 100-400 µg of unfiltered lysate. Proteins in the lysates were separated on a 4-20% gradient Tris-HCl gel (Bio-Rad, cat # 161-1159) and immunoprobed for anti-IKKβ (1:1000 dilutions).

Results for the partial purification of cell lysate: Detection of 100 µg, 200 µg and 400

µg un-filtered lysate was robust for IKKβ, and showed a concentration-dependence.

However, IKKβ was barely detectable in 100 µg, 200 µg or 400 µg of filtered lysate.

This decrease in IKKβ is likely due to aggregates formed during the filtration and not

retained in the lysate after centrifugation but prior to gel loading. However, Millipore

now offers new high recovery, high speed Amicon Ultra centrifugal filtration devices that

concentrate samples much faster (the new filtration devices were used later to

concentrate IL-6 in cell culture medium). The protein aggregates that were formed

222

during the long centrifugation steps might not aggregate in the new spin concentrators

that decrease the time of samples under gravitational force, yielding an increase in IKKβ.

(1f) Approach to radiolabel detection of S-glutathionylation of IKKβ immunoprecipitated

from lysates of rMC-1 cells. As an alternative to mass spectral analysis, detection of the

glutathionylation of IKKβ was attempted using radiolabeled sulfhydryl modifying

reagents that react with free sulfhydryls after treatment with glutaredoxin.

Method for the radiolabel detection of S-glutathionylation of immunoprecipitated IKKβ -

IKKβ was immunoprecipitated (or IgG control) from IAM (5 mM)-treated lysate from one 100 mm dish of rMC-1 cells, and the Protein A/G PLUS beads bound with IKK were washed several times in PBS, incubated with 2 nM-2000 nM Grx1 with or without 0.5 mM GSH, and subsequently labeled with either [14C] IAM or [14C] NEM. The radiolabel

associated with the immunoprecipitated IKKβ was then determined via scintillation

counting. Values were similar regardless of whether or not IKKβ was eluted from the

bead prior to scintillation counting.

Results the radiolabel detection of S-glutathionylation of immunoprecipitated IKKβ:

IKKβ was immunoprecipitated from IAM-treated lysates of rMC-1 cells, and the endogenously glutathionylated residues were then treated with Grx1. Newly generated free sulfhydryls would be blocked with radiolabeled [14C]-IAM or [14C]-NEM, and

detected via scintillation fluid readings. Values above background (IgG control) were not obtained for any concentration of Grx1. However, these experiments were conducted on

223

lysate from four 150 mm dishes of cells under conditions that were later shown to only

give a yield of around 32 ng of IKKβ (Fig. A1, A, pg. 203). Using the specific activity of

the sulfhydryl labeling compound, the amount of radioactivity expected from this amount

of IKKβ would be only 44 cpm (not above background). Nevertheless, the concept of

these experiments is logical, and if the conditions for the immunoprecipitations could be

worked out such that the yield was sufficiently enhanced and the specific radioactivity

increased, these techniques might prove useful.

(2) Approaches for quantitative detection of S-glutathionylation of IKKβ within lysates

of rMC-1 cells.

(2a) Attempts to analyze glutathionylation proteins by Western blot analysis of GSH in lysate of rMC-1 cells. The analysis of lysates immunoprobed with an anti-GSH antibody

could provide relative differences between two samples, and we sought to detect

glutathionylation of a protein that migrates at 87 kDa (i.e. IKKβ).

Methods for immunoprobing with the anti-GSH antibody: rMC-1 cells were cultured in

normal or high glucose medium for 5 days, lysed in 1% NP40 buffer, and processed via

Western blot analysis for protein glutathionylation using an anti-GSH antibody. The

basic protocol outlined in the above chapters was used for Western blot analysis. The anti-GSH antibody (Virogen, 1:500 dilution) was used to immunoprobe the glutathione

adduct of IKKβ in eluates of immunoprecipitated IKKβ and in cell lysate separated on

12% SDS-PAGE gels. DTT was omitted throughout all processes. In some cases, lower

224 molecular weight proteins (<50 kDa) were run off the gel to increase separation of higher molecular weight proteins such as IKKβ (87 kDa).

Results for detecting glutathionylation of proteins in lysate using the anti-GSH antibody:

A band that corresponded to IKKβ (87 kDa) was not detected in an immunoprecipitation of IKKβ from lysate (1-3 mg) of rMC-1 cells. The percent yield of these immunoprecipitations was found later to be 0.1-0.4% (see above), and therefore this technique was not sensitive enough to detect the low amounts of glutathionylated IKKβ that were present.

Differences in protein glutathionylation in lysates (containing NEM) from normal versus high glucose treated cells were not distinguishable using the anti-GSH antibody (n=1)

(Fig A4-1, pg. 228). The variability in the number of protein bands detected with the anti-GSH antibody can be seen by comparing Western blot analysis of lysates from cells in normal glucose medium that contained either NEM (Fig. A4-1, A-B, pg. 228) or IAM

(Fig. A4-2, A-B, pg. 229). A discrete robust band at 42 kDa was reversible by DTT and was increased in lysates treated with GSSG, (Fig. A4-1, A, arrows, pg. 228). This protein corresponds to actin, an abundant cellular protein that is well known to be S- glutathionylated (Wang et al., 2001;Wang et al., 2003b). Interestingly, lysate containing

IAM does not give rise to a similar signal, although the background in this experiment was notably high (Fig. A4-2, A, lanes 9 and 10, pg. 229). In one experiment, a robust

DTT-reversible band at approximately 50 kDa was immunoreactive with anti-GSH in the presence of NEM but not in its absence (Fig. A4-1, B, pg. 228) or in the presence of IAM

225

(Fig A4-2, A-B, pg. 229) likely due to a change in protein conformation or an artifact

(n=1).

BSA-SSG, a prototypic glutathionylated protein, seemed to be detected at approximately

80 kDa and was reversible with DTT (Fig. A4-2, B, lanes 2 and 6, pg. 229). Also, one band at approximately 100 kDa was found in lysate treated with GSSG and migrated similarly to recombinant IKK treated with GSSG (Fig. A4-2, B, lanes 5 and 7, pg. 229).

Because the migration of both the recombinant IKK (87 kDa) and BSA (66 kDa) on the gel was higher than that corresponding to the molecular weight markers, it seems that either the proteins migrated abnormally, the molecular weight markers are not accurate at the higher molecular weights, or the molecular weight markers are not accurate when the proteins migrated through the gel for an extended electrophoresis time.

More importantly, these data suggest that endogenous IKK can be glutathionylated to a detectable extent by GSSG in cells lysed in IAM, but not NEM, according to measurements with the anti-glutathione antibody. Two possible explanations for this are that the IAM is not efficiently binding IKK, leaving reduced sulfhydryls susceptible to glutathionylation, IAM binding to IKK causes a conformational change that increases the accessibility of anti-GSH to the site of glutathionylation, or NEM binding to IKK causes a conformational change that decreases the accessibility of anti-GSH to the site of glutathionylation. Interestingly, this finding correlates to the data showing that IKKβ could only be co-immunoprecipitated via IKKγ in the presence of IAM, but not NEM

(Fig. A1, B, pg. 203). Furthermore, these results indicate that the anti-glutathione

226

antibody is effective for detection of glutathionylation of abundant proteins (i.e., actin) and possibly lysate oxidized by exogenous GSSG to non-physiological levels but is not a useful tool for measuring endogenous levels of glutathionylated IKK or other proteins in the NFκB pathway.

227

Figure A4-1. Western blot analysis of protein-SSG in lysates of rMC-1 cells. Cell lysates treated with or without GSSG or DTT were analyzed for protein glutathionylation via an anti-GSH antibody (1:500 dilution). (A) Cells cultured in normal or high glucose medium were lysed in 10 mM NEM and treated with or without 20 mM GSSG or 40 mM DTT. Lane 1: 100 µg 1lysate + GSSG, Lane 3: 100 µg 2lysate + GSSG, Lane 5: 100 µg 1lysate + DTT, Lane 6: 100 µg 2lysate + DTT, Lane 8: 100 µg 1lysate, Lane 9: 100 µg 2lysate, Lane 10: mw marker. Arrows indicate a glutathionylated protein (approx. 42 kDa). (B) Cells cultured in normal glucose medium were lysed in 7.5 mM NEM and treated with or without 5 mM DTT. Lane 2: 100 µg lysate, Lane 3: 200 µg, Lane 6: 800 µg lysate + DTT, Lane 8: 200 µg lysate + DTT, Lane 9: 100 µg lysate + DTT, Lane 10: mw marker. A DTT-reversible protein that migrated to approximately 50 kDa on the gel is circled.

228

Figure A4-2. Western blot analysis of protein-SSG in lysates of rMC-1 cells. Protein glutathionylation of lysate of rMC-1 cells cultured in normal glucose (5 mM) medium (containing 5 mM IAM), BSA, and recombinant IKKβ (rIKKβ) was detected in Western blots using the anti-GSH antibody (1:500 dilution). (A) Lysate and our standard glutathionylated protein BSA-SSG were treated with or without 40 mM DTT or 100 mM GSSG. Lane 1: 400 µg lysate + DTT, Lane 2: 200 µg lysate + DTT, Lane 3: 400 µg BSA-SSG + DTT, Lane 4: Molecular weight marker, Lane 5: Empty lane (apparent flow over from other lanes), Lane 6: 400 µg lysate + GSSG, Lane 7: 200 µg lysate + GSSG, Lane 8: 400 µg BSA-SSG, Lane 9: 400 µg lysate , Lane 10: 200 µg lysate. (B) Lysate, BSA-SSG, and rIKKβ were treated with or without 100 mM DTT or 100 mM GSSG. Lane 1: 200 ng rIKKβ + DTT, Lane 2: 200 µg BSA-SSG + DTT, Lane 3: 200 µg lysate + DTT, Lane 5: 200 ng rIKKβ + GSSG, Lane 6: 200 µg BSA-SSG, Lane 7: 200 µg + GSSG, Lane 8: Molecular weight marker, Lane 9: 200 ng rIKKβ, Lane 10: 200 µg lysate

229

(2b) Attempts to measure protein glutathionylation via rMC-1 cellular uptake of

radiolabeled GSH. The net negative charge of GSH prevents GSH from passively

crossing lipid barriers, including membranes of most cells. However, GSH transporters

have been indentified in several types of cells including retinal pigmented epithelial cells

(Lu et al., 1995), cerebrovascular endothelial cells (Kannan et al., 2000), astrocytes

(Kannan et al., 2000), and small intestine epithelial cells (Iantomasi et al., 1997).

Moreover, primary Müller cells and rMC-1 cells specifically have been characterized for

sodium-dependent and sodium-independent GSH transporters (Kannan et al., 1999).

Therefore, theoretically, the endogenous glutathionylation status of proteins could be

measured via uptake of radiolabeled glutathione by rMC-1 cells, and glutathionylation

specific to IKKβ could be detected by scintillation counting of an in gel digestion if the

excised band did not contain interfering proteins.

Methods for [35S] GSH uptake into cells for detection of protein glutathionylation:

rMC-1 cells were cultured in 60 mm dishes and depleted of GSH with 10 mM L-

buthionine sulfoximine (BSO) plus 30 mM dimethyl fumarate (DMF) in NaCl buffer

(100 mM NaCl, 1.2 mM MgCl2, 0.81 mM MgSO4, and 25 mM HEPES-Tris pH= 7.4) for

30-40 min at 37°C. The BSO inhibits -glutamylcysteine synthetase ( -GCS, glutamate-

cysteine ligase), an enzyme that catalyzes the first and rate-limiting step in glutathione

biosynthesis. DMF rapidly depletes intracellular GSH via covalent modification by

GSH-S-transferase and subsequent export of the GS-DMF adduct. Cells used in these experiments were 50-100% confluent. Control experiments were pre-equilibrated for 5-

10 min at 4°C and conducted at 4°C. Cells were rinsed once in NaCl buffer, and

230

incubated with 1 mM (6 µCi) [35S] GSH for 30 min at 37°C. Cells were washed four

times with 5 mM non-radiolabeled GSH, lysed in 1% NP40 buffer, and incubated with

20% ice old TCA for 10 min (1:1 volume). Precipitated proteins were redissolved and

neutralized with NaOH. The radiolabel associated with these proteins was measured by

liquid scintillation counting. The removal of the glutathione was tested with 100 mM

DTT. Grx would have been used instead of DTT in subsequent experiments to

demonstrate disulfide reduction attributed selectively to protein glutathionylation.

Results for [35S] GSH uptake into cells and subsequent detection of protein

glutathionylation: A small amount of radiolabel was detected to be associated with

cellular proteins after [35S] GSH was applied to the outside of intact GSH-depleted rMC-

1 cells, but it was not releasable with DTT, indicating that the radiolabeled glutathione

was not predominantly in the form of S-glutathionylated proteins. A longer incubation

time of the radiolabeled glutathione might enhance the glutathione equilibration with

endogenously glutathionylated proteins. Nevertheless, the sensitivity of this technique likely is not sufficient to detect endogenous glutathionylation of IKKβ. Additionally, the

limitations of purity and yield of in-gel digestions have been discussed above.

(2c) Approach to measuring the activity of endogenous IKKβ in lysates of rMC-1 cells in

the absence or presence of purified Grx1. As an alternative to detection of glutathionylation on IKK, changes in the IKK kinase activity in response to treatment

with Grx1 (specific de-glutathionylation) would indirectly reflect differences in the

glutathionylation status of IKK. Standard IKK activity assays depend on

231 immunoprecipitation of IKK and [32P] ATP to phosphorylate an IKK substrate, IκB.

However, due to the poor yield of the immunoprecipitations in this system (see above) and the anticipation that immunoprecipitations might work differently with glutathionylated and non-glutathionylated forms of IKK, modifications were made to the standard kinase assay. Cell lysate was substituted for the immunoprecipitation, a change that is not expected to cause a large increase in non-specific phosphorylation because

IKK is the only enzyme known to catalyze phosphorylation of the serines in the active site of IκB. C-Src and other Src family kinases phosphorylate Tyr 42 on IκBα (Fan et al., 2004), but a tyrosine phosphatase inhibitor is presence in the reaction mixture.

Method for detecting IKKβ activation via IκB phosphorylation: IκB is an immediate down-stream substrate of IKKβ, and its phosphorylation was used to assay activity of

IKKβ in lysates of rMC-1 cells. Cells were lysed in NEM to prevent thiol exchange and artifactual glutathionylation of cysteines during processing. All reduced and accessible cysteines become labeled with NEM (Fig. 4.4, pg. 197). Addition of Grx1 then reduces the glutathionylated proteins in the lysate. Since glutathionylation of IKKβ has been correlated with a decrease in its activity (Reynaert et al., 2006), an increase in IκB phosphorylation is expected to be observed after incubation with Grx1.

First, a GST-tagged IκB plasmid and a mutant GST-tagged IκB plasmid for a negative control were obtained and used to express the corresponding proteins. Then, 20 µl of a

1:2 slurry of GSH-agarose beads was incubated with 5-10 µg IκB-GST protein overnight.

The IκB-GST-GSH-agarose bead mixture was then incubated with protein lysate (15-120

232

µg in 1% NP40 buffer and 5 mM NEM), 34 µl kinase buffer (20 mM HEPES pH =7.5,

50 mM KCl, 5 mM MgCl2, 1 mM MnCl2, 2 mM NaF, 0.1 mM Na3VO4, 2 mM PMSF,

0.02 mM leupeptin, and 5 µCi [32P] ATP), and 10 nM to 10 µM of Grx1 (13 µl) for 10

min, 30 min, or 1 hr at RT. The IκB-GST-GSH-agarose beads were washed with 200 µl

PBS for up to six times, and the radioactivity associated with the immobilized IκB was

detected by scintillation fluid counting.

Results of IKKβ activation by Grx1 according to IκB phosphorylation: In pilot studies,

lysate from rMC-1 cells and containing 5 mM NEM was incubated with [32P] ATP and

an IKK substrate (IκB-GST) in the presence or absence of exogenous Grx1. No

difference in IKK kinase activity was observed in response to a range of concentrations

of Grx1 (10 nM to 10 µM). We speculated that excess NEM from the lysate was likely

binding to and inactivating the exogenous Grx1. Indeed Grx1 pre-incubated in 5 mM

NEM -lysates did not maintain activity in the standard plate reader assay.

Methods of quenching NEM with hydroxylamine: NEM primarily reacts with sulfhydryl

groups at pH 7 via a Michael’s addition but also can also undergo hydrolysis by reaction

with amino groups such as that of hydroxylamine (NH2OH). Hydroxylamine (NH2OH)

has been used previously to selectively reduce thioesters without cleaving disulfide bonds

(Wan et al., 2007). Therefore, we hypothesized that hydroxylamine might bind the excess NEM without removing glutathione from Cys-SSG residues in this system. To test the ability of hydroxylamine to quench the excess NEM in the lysate, lysis buffer (i.e. no cells) with 5 mM NEM was pre-incubated with 0.0001 M to 10 M NH2OH for 30 min-

233

1 hr, and then purified Grx1 (1 U/ml) was added for an additional 5-30 min. Grx activity

was measured via a standard plate reader assay with CSSG as a substrate and coupled to

NADPH oxidation.

Results of quenching NEM with hydroxylamine: No Grx1 activity was observed in

buffers containing 5 mM NEM that were incubated with 0.1-100 mM NH2OH. However,

1-10 M NH2OH pre-incubated with an NEM solution for 30 min prevented inhibition of

Grx1 activity, but the activity of Grx1 demonstrated a trend of inverse proportion to the

time that exogenous Grx1 was in the presence of the NEM- NH2OH mixture. Increasing

the pre-incubation time of NEM and NH2OH to 1 hr circumvented the time-dependent

inhibition on Grx1 activity when Grx1 was incubated in the NEM- NH2OH mix for up to

30 min, and Grx1 maintained full activity under these conditions. Also, 1 M NH2OH (in

the absence of the Grx1 system) does not remove radiolabeled glutathione from [35S]

BSA-SSG after a 30 min-1 hr incubation and subsequent precipitation with 20% cold

TCA. Thus, a standard protocol was developed (a 1 hr incubation of lysate containing 5 mM NEM with 1M NH2OH), and can be subsequently used to quench NEM in lysates

prior to addition of exogenous Grx1 in IKKβ activity assays. The alternation of the

conformation of IKK by modification with NEM such that it cannot display activity is a possible complication.

234

A.2 Attempts to measure the effects of Grx1 on cell death

Increases in Grx1 in rMC-1 cells promote pro-inflammatory effects in rMC-1 and

TRiBRB endothelial cells. To test whether changes in Grx1 affect cell death, the following experiments were conducted. rMC-1 cells over-expressing Grx1 were tested for cell death (A.2-1a) and apoptosis specifically (A.2-1b). In addition, TRiBRB cells were cultured in the conditioned medium from rMC-1 cells over-expressing Grx1 and analyzed for apoptosis (A.2-2).

A.2-1a Methods for measuring the uptake of trypan blue in Grx1 over-expressing rMC-1 cells- rMC-1 cells in (about 0.1 million cells per well) 6-well dishes were transfected with Ad-Grx1 or Ad-Empty at MOI 10 for 1 hr as discussed above, and cultured for an additional 24-48 hr. The 2 ml medium from the cells was collected. Cells were washed with 300 µl PBS, and removed from the plate with 300 µl trypsin for 5 min at 37°C. The

PBS, trypsin, and trypsinized cells were combined with the 2 ml medium, and 10 µl of this mixture was incubated with 10 µl 0.4% trypan blue solution (Gibco cat # 15250-061) for 5 min at RT. Cells in 10 µl of the mixture were counted in a hemacytometer in duplicate. The number of blue cells was divided by the total number of cells to determine the fraction of non-viable cells.

Results of the effects of Grx1 over-expression on trypan blue uptake in rMC-1 cells.

Over-expression of Grx1 in rMC-1 cells leads to increased nuclear expression and activity of NFκB. NFκB is well known to be involved in both apoptotic events and

235

inflammatory responses. To test whether the activation of NFκB by Grx1 affects cell

death, trypan blue exclusion was used to report viability (defined here by intact

membranes) of Grx1 over-expressing rMC-1 cells. However, no substantial difference in

the percent of cell death was observed (n=6). Increasing the number of cells in these

experiments might make small differences in cell death more apparent, but overall Grx1

and Grx1-regulated NFκB does not seem to affect overall cell death in rMC-1 cells.

2A-1b Methods for Hoechst staining of Grx1 over-expressing rMC-1 cells. rMC-1 cells

in (about 0.1 million cells per well) 6-well plates were infected with Ade-Grx1 or Ad-

Empty at MOI of 10, MOI 20 or MOI 40 for 1 hr as described above. 24-48 hr after the

infection, 2 µl 10 mM Hoechst dye was added to 2 ml medium on the cells and incubated

for 10 min at 37°C. Cells were visualized with a Leica fluorescent microscope, and some

fields of cells were analyzed for the number of cells with condensed chromatin in a

blinded fashion by Elizabeth Sabens in our laboratory in order to obtain unbiased results.

Results of the effects of Grx1 on condensed chromatin in rMC-1 cells. The lack of

changes observed with trypan blue exclusion could result if a change in apoptotic death

was counter-acted by other forms of death such as necrosis. Caspase activity and

Annexin V staining is increased in rMC-1 cells cultured in high glucose (Mohr et al.,

2002), suggesting that these cells can die via apoptotic pathways. To selectively address the effects of Grx1 on apoptosis, Hoechst staining was used to visualize condensed nuclear chromatin. Apoptotic versus non-apoptotic cells were clearly distinguishable according to condensed chromatin. No changes in the number of cells with condensed

236

chromatin were observed at MOI 10 (n=7), MOI 20 (n=1), or MOI 40 (n=1). These data

suggest that Grx1 and Grx1-regulated activation of NFκB do not lead to changes in

apoptotic events detected by chromatin condensation in rMC-1 cells. Confirming that

similar results are observed according to Annexin V staining and Caspase activation as

was done by Mohr et al. (2002) would suggest that NFκB activation is not leading to

apoptosis in rMC-1 cells.

2A-2 Methods for Hoechst staining of TRiBRB endothelial cells that were cultured in

conditioned medium from rMC-1 cells over-expressing Grx1- rMC-1 cells (0.2 million

cells per well) in 6-well plates were infected with Ad-Grx1 or Ad-Empty at MOI 10 for 1

hr as described above. 24 hr after the infection, the 2 ml medium was collected and

assayed for glucose concentrations (see Ch. 2 Methods). The conditioned medium was

supplemented with glucose to restore a 5 mM concentration when necessary and with 15

µg/ml endothelial cell growth supplement (ECGS, Sigma, cat # E0760).

The conditioned medium from the rMC-1 cells was sterile filtered onto wells of a 6-well

gelatin-coated (0.1%) dishes of TRiBRB endothelial cells (0.2 million cells per well), and the endothelial cells were cultured for an additional 24 hr. In some cases, cells were treated with conditioned medium from Grx1 over-expressing rMC-1 cells for two

consecutive days. For a positive control for cell death, the conditioned medium was substituted for chemicals known to induce cell death in other contexts; 0.5 mM or 1 mM hydrogen peroxide, 20 µM etoposide, or 12-24 µg/ml cycloheximide. Etoposide and hydrogen peroxide have been used to induce apoptosis via NFκB signaling (Shishodia

237

and Aggarwal, 2002). Two µl 10 mM Hoechst dye was added to 2ml medium per well of

TRiBRB cells and incubated for 10 min at 37°C. Cells were visualized with a Leica fluorescent microscope.

Results of the effects of conditioned medium from rMC-1 cells over-expressing Grx1 on

Hoechst staining of condensed chromatin in TRiBRB endothelial cell. Upregulation of

Grx1 leads to increased activation of NFκB in rMC-1 cells. We set out to test whether

the increase in NFκB activity in rMC-1 cells leads to apoptosis of neighboring

endothelial cells. Thus, rat TRiBRB endothelial cells were analyzed for chromatin

condensation after being cultured in conditioned medium from Grx1 over-expressing

rMC-1 cells. No change in the number of endothelial cells containing condensed

chromatin was observed in response to changes in Grx1 expression in rMC-1 cells. As

expected, increased chromatin condensation was found in endothelial cells in response to

positive control stimuli i.e., hydrogen peroxide, etoposide, and cycloheximide.

238

Bibliography

Abu el Asrar,A.M., Maimone,D., Morse,P.H., Gregory,S., Reder,A.T. (1992) Cytokines in the vitreous of patients with proliferative diabetic retinopathy. Am.J.Ophthalmol. 114(6): 731-736.

Adachi,T., Pimentel,D.R., Heibeck,T., Hou,X., Lee,Y.J., Jiang,B., Ido,Y., Cohen,R.A. (2004a) S-glutathiolation of Ras mediates redox-sensitive signaling by angiotensin II in vascular smooth muscle cells. J.Biol.Chem. 279(28): 29857-29862.

Adachi,T., Weisbrod,R.M., Pimentel,D.R., Ying,J., Sharov,V.S., Schoneich,C., Cohen,R.A. (2004b) S-Glutathiolation by peroxynitrite activates SERCA during arterial relaxation by nitric oxide. Nat.Med. 10(11): 1200-1207.

Aguirre,V., Uchida,T., Yenush,L., Davis,R., White,M.F. (2000) The c-Jun NH(2)- terminal kinase promotes insulin resistance during association with insulin receptor substrate-1 and phosphorylation of Ser(307). J.Biol.Chem. 275(12): 9047-9054.

Akiyama,H., Barger,S., Barnum,S., Bradt,B., Bauer,J., Cole,G.M., Cooper,N.R., Eikelenboom,P., Emmerling,M., Fiebich,B.L., Finch,C.E., Frautschy,S., Griffin,W.S., Hampel,H., Hull,M., Landreth,G., Lue,L., Mrak,R., Mackenzie,I.R., McGeer,P.L., O'Banion,M.K., Pachter,J., Pasinetti,G., Plata- Salaman,C., Rogers,J., Rydel,R., Shen,Y., Streit,W., Strohmeyer,R., Tooyoma,I., Van Muiswinkel,F.L., Veerhuis,R., Walker,D., Webster,S., Wegrzyniak,B., Wenk,G., Wyss-Coray,T. (2000) Inflammation and Alzheimer's disease. Neurobiol.Aging 21(3): 383-421.

Akterin,S., Cowburn,R.F., Miranda-Vizuete,A., Jimenez,A., Bogdanovic,N., Winblad,B., Cedazo-Minguez,A. (2006) Involvement of glutaredoxin-1 and thioredoxin-1 in beta-amyloid toxicity and Alzheimer's disease. Cell Death.Differ. 13(9): 1454-1465.

Apetoh,L., Ghiringhelli,F., Tesniere,A., Obeid,M., Ortiz,C., Criollo,A., Mignot,G., Maiuri,M.C., Ullrich,E., Saulnier,P., Yang,H., Amigorena,S., Ryffel,B., Barrat,F.J., Saftig,P., Levi,F., Lidereau,R., Nogues,C., Mira,J.P., Chompret,A., Joulin,V., Clavel-Chapelon,F., Bourhis,J., Andre,F., Delaloge,S., Tursz,T., Kroemer,G., Zitvogel,L. (2007) Toll-like receptor 4- dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat.Med. 13(9): 1050-1059.

Aracena,P., Sanchez,G., Donoso,P., Hamilton,S.L., Hidalgo,C. (2003) S- glutathionylation decreases Mg2+ inhibition and S-nitrosylation enhances Ca2+ activation of RyR1 channels. J.Biol.Chem. 278(44): 42927-42935.

239

Arkan,M.C., Hevener,A.L., Greten,F.R., Maeda,S., Li,Z.W., Long,J.M., Wynshaw- Boris,A., Poli,G., Olefsky,J., Karin,M. (2005) IKK-beta links inflammation to obesity-induced insulin resistance. Nat.Med. 11(2): 191-198.

Arredouani,A., Guiot,Y., Jonas,J.C., Liu,L.H., Nenquin,M., Pertusa,J.A., Rahier,J., Rolland,J.F., Shull,G.E., Stevens,M., Wuytack,F., Henquin,J.C., Gilon,P. (2002) SERCA3 ablation does not impair insulin secretion but suggests distinct roles of different sarcoendoplasmic reticulum Ca(2+) pumps for Ca(2+) homeostasis in pancreatic beta-cells. Diabetes 51(11): 3245-3253.

Ashcroft,F.M. (2005) ATP-sensitive potassium channelopathies: focus on insulin secretion. J.Clin.Invest 115(8): 2047-2058.

Aviram,M., Rosenblat,M., Billecke,S., Erogul,J., Sorenson,R., Bisgaier,C.L., Newton,R.S., La,D.B. (1999) Human serum paraoxonase (PON 1) is inactivated by oxidized low density lipoprotein and preserved by antioxidants. Free Radic.Biol.Med. 26(7-8): 892-904.

Baldwin,A.S., Jr. (1996) The NF-kappa B and I kappa B proteins: new discoveries and insights. Annu.Rev.Immunol. 14: 649-683.

Bandyopadhyay,S., Starke,D.W., Mieyal,J.J., Gronostajski,R.M. (1998) Thioltransferase (glutaredoxin) reactivates the DNA-binding activity of oxidation-inactivated nuclear factor I. J.Biol.Chem. 273(1): 392-397.

Barrett,W.C., DeGnore,J.P., Keng,Y.F., Zhang,Z.Y., Yim,M.B., Chock,P.B. (1999a) Roles of superoxide radical anion in signal transduction mediated by reversible regulation of protein-tyrosine phosphatase 1B. J.Biol.Chem. 274(49): 34543-34546.

Barrett,W.C., DeGnore,J.P., Konig,S., Fales,H.M., Keng,Y.F., Zhang,Z.Y., Yim,M.B., Chock,P.B. (1999b) Regulation of PTP1B via glutathionylation of the active site cysteine 215. Biochemistry 38(20): 6699-6705.

Bauer,J., Bauer,T.M., Kalb,T., Taga,T., Lengyel,G., Hirano,T., Kishimoto,T., Acs,G., Mayer,L., Gerok,W. (1989a) Regulation of interleukin 6 receptor expression in human monocytes and monocyte-derived macrophages. Comparison with the expression in human hepatocytes. J.Exp.Med. 170(5): 1537-1549.

Bauer,J., Lengyel,G., Bauer,T.M., Acs,G., Gerok,W. (1989b) Regulation of interleukin-6 receptor expression in human monocytes and hepatocytes. FEBS Lett. 249(1): 27-30.

Baumann,M., Baumann,H., Fey,G.H. (1990) Molecular cloning, characterization and functional expression of the rat liver interleukin 6 receptor. J.Biol.Chem. 265(32): 19853-19862.

240

Biaglow,J.E., Donahue,J., Tuttle,S., Held,K., Chrestensen,C., Mieyal,J. (2000) A method for measuring disulfide reduction by cultured mammalian cells: relative contributions of glutathione-dependent and glutathione-independent mechanisms. Anal.Biochem. 281(1): 77-86.

Biber,K., Pinto-Duarte,A., Wittendorp,M.C., Dolga,A.M., Fernandes,C.C., Von Frijtag Drabbe,K.J., Keijser,J.N., de,V.R., Ijzerman,A.P., Ribeiro,J.A., Eisel,U., Sebastiao,A.M., Boddeke,H.W. (2008) Interleukin-6 upregulates neuronal adenosine A1 receptors: implications for neuromodulation and neuroprotection. Neuropsychopharmacology 33(9): 2237-2250.

Bidasee,K.R., Dincer,U.D., Besch,H.R., Jr. (2001) Ryanodine receptor dysfunction in hearts of streptozotocin-induced diabetic rats. Mol.Pharmacol. 60(6): 1356- 1364.

Blaber,R., Stylianou,E., Clayton,A., Steadman,R. (2003) Selective regulation of ICAM-1 and RANTES gene expression after ICAM-1 ligation on human renal fibroblasts. J.Am.Soc.Nephrol. 14(1): 116-127.

Bobryshev,Y.V. (2006) Monocyte recruitment and foam cell formation in atherosclerosis. Micron. 37(3): 208-222.

Boothby,M. (2001) Specificity of sn50 for NF-kappa B? Nat.Immunol. 2(6): 471-472.

Brambilla,R., Bracchi-Ricard,V., Hu,W.H., Frydel,B., Bramwell,A., Karmally,S., Green,E.J., Bethea,J.R. (2005) Inhibition of astroglial nuclear factor kappaB reduces inflammation and improves functional recovery after spinal cord injury. J.Exp.Med. 202(1): 145-156.

Brownlee,M. (2001) Biochemistry and molecular cell biology of diabetic complications. Nature 414(6865): 813-820.

Burdakov,D., Ashcroft,F.M. (2002) Shedding new light on brain metabolism and glial function. J.Physiol 544(Pt 2): 334.

Cai,D., Yuan,M., Frantz,D.F., Melendez,P.A., Hansen,L., Lee,J., Shoelson,S.E. (2005) Local and systemic insulin resistance resulting from hepatic activation of IKK-beta and NF-kappaB. Nat.Med. 11(2): 183-190.

Cameron,N.E., Cotter,M.A. (2008) Pro-inflammatory mechanisms in diabetic neuropathy: focus on the nuclear factor kappa B pathway. Curr.Drug Targets. 9(1): 60-67.

Canataroglu,H., Varinli,I., Ozcan,A.A., Canataroglu,A., Doran,F., Varinli,S. (2005) Interleukin (IL)-6, interleukin (IL)-8 levels and cellular composition of the vitreous humor in proliferative diabetic retinopathy, proliferative vitreoretinopathy, and traumatic proliferative vitreoretinopathy. Ocul.Immunol.Inflamm. 13(5): 375-381.

241

Caplan,J.F., Filipenko,N.R., Fitzpatrick,S.L., Waisman,D.M. (2004) Regulation of annexin A2 by reversible glutathionylation. J.Biol.Chem. 279(9): 7740-7750.

Cashman,S.M., Bowman,L., Christofferson,J., Kumar-Singh,R. (2006) Inhibition of choroidal neovascularization by adenovirus-mediated delivery of short hairpin RNAs targeting VEGF as a potential therapy for AMD. Invest Ophthalmol.Vis.Sci. 47(8): 3496-3504.

Catley,M.C., Sukkar,M.B., Chung,K.F., Jaffee,B., Liao,S.M., Coyle,A.J., Haddad,e., Barnes,P.J., Newton,R. (2006) Validation of the anti-inflammatory properties of small-molecule IkappaB Kinase (IKK)-2 inhibitors by comparison with adenoviral-mediated delivery of dominant-negative IKK1 and IKK2 in human airways smooth muscle. Mol.Pharmacol. 70(2): 697-705.

Chai,Y.C., Ashraf,S.S., Rokutan,K., Johnston,R.B., Jr., Thomas,J.A. (1994a) S- thiolation of individual human neutrophil proteins including actin by stimulation of the respiratory burst: evidence against a role for glutathione disulfide. Arch.Biochem.Biophys. 310(1): 273-281.

Chai,Y.C., Hendrich,S., Thomas,J.A. (1994b) Protein S-thiolation in hepatocytes stimulated by t-butyl hydroperoxide, menadione, and neutrophils. Arch.Biochem.Biophys. 310(1): 264-272.

Chrestensen,C.A., Starke,D.W., Mieyal,J.J. (2000) Acute cadmium exposure inactivates thioltransferase (Glutaredoxin), inhibits intracellular reduction of protein-glutathionyl-mixed disulfides, and initiates apoptosis. J.Biol.Chem. 275(34): 26556-26565.

Chu,F., Ward,N.E., O'Brian,C.A. (2001) Potent inactivation of representative members of each PKC isozyme subfamily and PKD via S-thiolation by the tumor-promotion/progression antagonist glutathione but not by its precursor cysteine. Carcinogenesis 22(8): 1221-1229.

Church,L.D., Cook,G.P., McDermott,M.F. (2008) Primer: inflammasomes and interleukin 1beta in inflammatory disorders. Nat.Clin.Pract.Rheumatol. 4(1): 34-42.

Claiborne,A., Yeh,J.I., Mallett,T.C., Luba,J., Crane,E.J., III, Charrier,V., Parsonage,D. (1999) Protein-sulfenic acids: diverse roles for an unlikely player in enzyme catalysis and redox regulation. Biochemistry 38(47): 15407- 15416.

Clarke,M., Dodson,P.M. (2007) PKC inhibition and diabetic microvascular complications. Best.Pract.Res.Clin.Endocrinol.Metab 21(4): 573-586.

Clavreul,N., Bachschmid,M.M., Hou,X., Shi,C., Idrizovic,A., Ido,Y., Pimentel,D., Cohen,R.A. (2006) S-glutathiolation of p21ras by peroxynitrite mediates

242

endothelial insulin resistance caused by oxidized low-density lipoprotein. Arterioscler.Thromb.Vasc.Biol. 26(11): 2454-2461.

Clayton,A., Evans,R.A., Pettit,E., Hallett,M., Williams,J.D., Steadman,R. (1998) Cellular activation through the ligation of intercellular adhesion molecule-1. J.Cell Sci. 111 ( Pt 4): 443-453.

Coetzee,W.A., Nakamura,T.Y., Faivre,J.F. (1995) Effects of thiol-modifying agents on KATP channels in guinea pig ventricular cells. Am.J.Physiol 269(5 Pt 2): H1625-H1633.

Cotgreave,I.A., Gerdes,R.G. (1998) Recent trends in glutathione biochemistry-- glutathione-protein interactions: a molecular link between oxidative stress and cell proliferation? Biochem.Biophys.Res.Commun. 242(1): 1-9.

Coughlan,M.T., Cooper,M.E., Forbes,J.M. (2007) Renal microvascular injury in diabetes: RAGE and redox signaling. Antioxid.Redox.Signal. 9(3): 331-342.

Couper,K.N., Blount,D.G., Riley,E.M. (2008) IL-10: the master regulator of immunity to infection. J.Immunol. 180(9): 5771-5777.

Cross,J.V., Templeton,D.J. (2004) Oxidative stress inhibits MEKK1 by site-specific glutathionylation in the ATP-binding domain. Biochem.J. 381(Pt 3): 675-683.

Cruz,C.M., Rinna,A., Forman,H.J., Ventura,A.L., Persechini,P.M., Ojcius,D.M. (2007) ATP activates a reactive oxygen species-dependent oxidative stress response and secretion of proinflammatory cytokines in macrophages. J.Biol.Chem. 282(5): 2871-2879.

Dafre,A.L., Sies,H., Akerboom,T. (1996) Protein S-thiolation and regulation of microsomal glutathione transferase activity by the glutathione redox couple. Arch.Biochem.Biophys. 332(2): 288-294.

Dai,J., Wang,P., Bai,F., Town,T., Fikrig,E. (2008) Icam-1 participates in the entry of west nile virus into the central nervous system. J.Virol. 82(8): 4164-4168.

Davis,D.A., Dorsey,K., Wingfield,P.T., Stahl,S.J., Kaufman,J., Fales,H.M., Levine,R.L. (1996) Regulation of HIV-1 protease activity through cysteine modification. Biochemistry 35(7): 2482-2488.

Davis,D.A., Newcomb,F.M., Starke,D.W., Ott,D.E., Mieyal,J.J., Yarchoan,R. (1997) Thioltransferase (glutaredoxin) is detected within HIV-1 and can regulate the activity of glutathionylated HIV-1 protease in vitro. J.Biol.Chem. 272(41): 25935-25940.

Demasi,M., Silva,G.M., Netto,L.E. (2003) 20 S proteasome from Saccharomyces cerevisiae is responsive to redox modifications and is S-glutathionylated. J.Biol.Chem. 278(1): 679-685.

243

DeSouza,L., Diehl,G., Rodrigues,M.J., Guo,J., Romaschin,A.D., Colgan,T.J., Siu,K.W. (2005) Search for cancer markers from endometrial tissues using differentially labeled tags iTRAQ and cICAT with multidimensional liquid chromatography and tandem mass spectrometry. J.Proteome.Res. 4(2): 377- 386.

Di Simplicio,P., de Giorgio,L.A., Cardaioli,E., Lecis,R., Miceli,M., Rossi,R., Anichini,R., Mian,M., Seghieri,G., Franconi,F. (1995) Glutathione, glutathione utilizing enzymes and thioltransferase in platelets of insulin- dependent diabetic patients: relation with platelet aggregation and with microangiopatic complications. Eur.J.Clin.Invest 25(9): 665-669.

Di,P.A., Aigner,L.J., Dunn,R.J., Bray,G.M., Aguayo,A.J. (1998) Prolonged delivery of brain-derived neurotrophic factor by adenovirus-infected Muller cells temporarily rescues injured retinal ganglion cells. Proc.Natl.Acad.Sci.U.S.A 95(7): 3978-3983.

Du,X., Matsumura,T., Edelstein,D., Rossetti,L., Zsengeller,Z., Szabo,C., Brownlee,M. (2003a) Inhibition of GAPDH activity by poly(ADP-ribose) polymerase activates three major pathways of hyperglycemic damage in endothelial cells. J.Clin.Invest 112(7): 1049-1057.

Du,Y., Miller,C.M., Kern,T.S. (2003b) Hyperglycemia increases mitochondrial superoxide in retina and retinal cells. Free Radic.Biol.Med. 35(11): 1491- 1499.

Du,Y., Sarthy,V.P., Kern,T.S. (2004) Interaction between NO and COX pathways in retinal cells exposed to elevated glucose and retina of diabetic rats. Am.J.Physiol Regul.Integr.Comp Physiol 287(4): R735-R741.

Du,Y., Smith,M.A., Miller,C.M., Kern,T.S. (2002) Diabetes-induced nitrative stress in the retina, and correction by aminoguanidine. J.Neurochem. 80(5): 771- 779.

Dudus,L., Anand,V., Acland,G.M., Chen,S.J., Wilson,J.M., Fisher,K.J., Maguire,A.M., Bennett,J. (1999) Persistent transgene product in retina, optic nerve and brain after intraocular injection of rAAV. Vision Res. 39(15): 2545-2553.

Eaton,P., Shattock,M.J. (2002) Purification of proteins susceptible to oxidation at cysteine residues: identification of malate dehydrogenase as a target for S- glutathiolation. Ann.N.Y.Acad.Sci. 973: 529-532.

Efrat,S., Fleischer,N., Hanahan,D. (1990) Diabetes induced in male transgenic mice by expression of human H-ras oncoprotein in pancreatic beta cells. Mol.Cell Biol. 10(4): 1779-1783.

244

Eichler,W., Yafai,Y., Wiedemann,P., Reichenbach,A. (2004) Angiogenesis-related factors derived from retinal glial (Muller) cells in hypoxia. Neuroreport 15(10): 1633-1637.

Erkasap,N. (2007) SERCA in genesis of arrhythmias: what we already know and what is new? Anadolu.Kardiyol.Derg. 7 Suppl 1: 43-46.

Fagerlund,R., Kinnunen,L., Kohler,M., Julkunen,I., Melen,K. (2005) NF-{kappa}B is transported into the nucleus by importin {alpha}3 and importin {alpha}4. J.Biol.Chem. 280(16): 15942-15951.

Famiglietti,E.V., Stopa,E.G., McGookin,E.D., Song,P., LeBlanc,V., Streeten,B.W. (2003) Immunocytochemical localization of vascular endothelial growth factor in neurons and glial cells of human retina. Brain Res. 969(1-2): 195- 204.

Fan,C., Li,Q., Zhang,Y., Liu,X., Luo,M., Abbott,D., Zhou,W., Engelhardt,J.F. (2004) IkappaBalpha and IkappaBbeta possess injury context-specific functions that uniquely influence hepatic NF-kappaB induction and inflammation. J.Clin.Invest 113(5): 746-755.

Fan,Y., Ye,J., Shen,F., Zhu,Y., Yeghiazarians,Y., Zhu,W., Chen,Y., Lawton,M.T., Young,W.L., Yang,G.Y. (2008) Interleukin-6 stimulates circulating blood- derived endothelial progenitor cell angiogenesis in vitro. J.Cereb.Blood Flow Metab 28(1): 90-98.

Farese,R.V., Sajan,M.P., Standaert,M.L. (2005) Insulin-sensitive protein kinases (atypical protein kinase C and protein kinase B/Akt): actions and defects in obesity and type II diabetes. Exp.Biol.Med.(Maywood.) 230(9): 593-605.

Feit-Leichman,R.A., Kinouchi,R., Takeda,M., Fan,Z., Mohr,S., Kern,T.S., Chen,D.F. (2005) Vascular damage in a mouse model of diabetic retinopathy: relation to neuronal and glial changes. Invest Ophthalmol.Vis.Sci. 46(11): 4281-4287.

Finkel,T. (2000) Redox-dependent signal transduction. FEBS Lett. 476(1-2): 52-54.

Flekac,M., Skrha,J., Zidkova,K., Lacinova,Z., Hilgertova,J. (2007) Paraoxonase 1 gene polymorphisms and enzyme activities in diabetes mellitus. Physiol Res.

Foell,D., Wittkowski,H., Roth,J. (2007) Mechanisms of disease: a 'DAMP' view of inflammatory arthritis. Nat.Clin.Pract.Rheumatol. 3(7): 382-390.

Fratelli,M., Demol,H., Puype,M., Casagrande,S., Eberini,I., Salmona,M., Bonetto,V., Mengozzi,M., Duffieux,F., Miclet,E., Bachi,A., Vandekerckhove,J., Gianazza,E., Ghezzi,P. (2002) Identification by redox proteomics of glutathionylated proteins in oxidatively stressed human T lymphocytes. Proc.Natl.Acad.Sci.U.S.A 99(6): 3505-3510.

245

Gallogly,M.M., Mieyal,J.J. (2007) Mechanisms of reversible protein glutathionylation in redox signaling and oxidative stress. Curr.Opin.Pharmacol. 7(4): 381-391.

Gao,Z., Hwang,D., Bataille,F., Lefevre,M., York,D., Quon,M.J., Ye,J. (2002) Serine phosphorylation of insulin receptor substrate 1 by inhibitor kappa B kinase complex. J.Biol.Chem. 277(50): 48115-48121.

Gerber,S.A., Rush,J., Stemman,O., Kirschner,M.W., Gygi,S.P. (2003) Absolute quantification of proteins and phosphoproteins from cell lysates by tandem MS. Proc.Natl.Acad.Sci.U.S.A 100(12): 6940-6945.

Gerhardinger,C., Costa,M.B., Coulombe,M.C., Toth,I., Hoehn,T., Grosu,P. (2005) Expression of acute-phase response proteins in retinal Muller cells in diabetes. Invest Ophthalmol.Vis.Sci. 46(1): 349-357.

Ghosh,S., Karin,M. (2002) Missing pieces in the NF-kappaB puzzle. Cell 109 Suppl: S81-S96.

Ghosh,S., May,M.J., Kopp,E.B. (1998) NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu.Rev.Immunol. 16: 225-260.

Gilbert,H.F. (1995) Thiol/disulfide exchange equilibria and disulfide bond stability. Methods Enzymol. 251: 8-28.

Giustarini,D., Rossi,R., Milzani,A., Colombo,R., Dalle-Donne,I. (2004) S- glutathionylation: from redox regulation of protein functions to human diseases. J.Cell Mol.Med. 8(2): 201-212.

Gnudi,L., Frevert,E.U., Houseknecht,K.L., Erhardt,P., Kahn,B.B. (1997) Adenovirus-mediated gene transfer of dominant negative ras(asn17) in 3T3L1 adipocytes does not alter insulin-stimulated P13-kinase activity or glucose transport. Mol.Endocrinol. 11(1): 67-76.

Goalstone,M.L., Draznin,B. (1998) What does insulin do to Ras? Cell Signal. 10(5): 297-301.

Goldman,M.D. (2003) Lung dysfunction in diabetes. Diabetes Care 26(6): 1915-1918.

Goldstein,B.J. (2001) Protein-tyrosine phosphatase 1B (PTP1B): a novel therapeutic target for type 2 diabetes mellitus, obesity and related states of insulin resistance. Curr.Drug Targets.Immune.Endocr.Metabol.Disord. 1(3): 265-275.

Goldstein,B.J., Mahadev,K., Wu,X. (2005) Redox paradox: insulin action is facilitated by insulin-stimulated reactive oxygen species with multiple potential signaling targets. Diabetes 54(2): 311-321.

246

Gravina,S. (1993) Characterization and Kinetic mechanism of Thioltransferase. In. Case Western Reserve University.

Gravina,S.A., Mieyal,J.J. (1993) Thioltransferase is a specific glutathionyl mixed disulfide oxidoreductase. Biochemistry 32(13): 3368-3376.

Gustavsson,C., Agardh,C.D., Hagert,P., Agardh,E. (2008) Inflammatory markers in nondiabetic and diabetic rat retinas exposed to ischemia followed by reperfusion. Retina 28(4): 645-652.

Hagemann,T., Balkwill,F., Lawrence,T. (2007) Inflammation and cancer: a double- edged sword. Cancer Cell 12(4): 300-301.

Hamnell-Pamment,Y., Lind,C., Palmberg,C., Bergman,T., Cotgreave,I.A. (2005) Determination of site-specificity of S-glutathionylated cellular proteins. Biochem.Biophys.Res.Commun. 332(2): 362-369.

Harja,E., Bu,D.X., Hudson,B.I., Chang,J.S., Shen,X., Hallam,K., Kalea,A.Z., Lu,Y., Rosario,R.H., Oruganti,S., Nikolla,Z., Belov,D., Lalla,E., Ramasamy,R., Yan,S.F., Schmidt,A.M. (2008) Vascular and inflammatory stresses mediate atherosclerosis via RAGE and its ligands in apoE-/- mice. J.Clin.Invest 118(1): 183-194.

Harris,H.E., Andersson,U. (2004) Mini-review: The nuclear protein HMGB1 as a proinflammatory mediator. Eur.J.Immunol. 34(6): 1503-1512.

Harris,H.E., Raucci,A. (2006) Alarmin(g) news about danger: workshop on innate danger signals and HMGB1. EMBO Rep. 7(8): 774-778.

Henquin,J.C. (2000) Triggering and amplifying pathways of regulation of insulin secretion by glucose. Diabetes 49(11): 1751-1760.

Hernandez,C., Segura,R.M., Fonollosa,A., Carrasco,E., Francisco,G., Simo,R. (2005) Interleukin-8, monocyte chemoattractant protein-1 and IL-10 in the vitreous fluid of patients with proliferative diabetic retinopathy. Diabet.Med. 22(6): 719-722.

Hibi,M., Murakami,M., Saito,M., Hirano,T., Taga,T., Kishimoto,T. (1990) Molecular cloning and expression of an IL-6 signal transducer, gp130. Cell 63(6): 1149-1157.

Hirota,K., Matsui,M., Murata,M., Takashima,Y., Cheng,F.S., Itoh,T., Fukuda,K., Yodoi,J. (2000) Nucleoredoxin, glutaredoxin, and thioredoxin differentially regulate NF-kappaB, AP-1, and CREB activation in HEK293 cells. Biochem.Biophys.Res.Commun. 274(1): 177-182.

247

Hoppe,G., Talcott,K.E., Bhattacharya,S.K., Crabb,J.W., Sears,J.E. (2006) Molecular basis for the redox control of nuclear transport of the structural chromatin protein Hmgb1. Exp.Cell Res. 312(18): 3526-3538.

Hosoya,K., Tomi,M., Ohtsuki,S., Takanaga,H., Ueda,M., Yanai,N., Obinata,M., Terasaki,T. (2001) Conditionally immortalized retinal capillary endothelial cell lines (TR-iBRB) expressing differentiated endothelial cell functions derived from a transgenic rat. Exp.Eye Res. 72(2): 163-172.

Humphries,K.M., Juliano,C., Taylor,S.S. (2002) Regulation of cAMP-dependent protein kinase activity by glutathionylation. J Biol.Chem. 277(45): 43505- 43511.

Hurd,T.R., Requejo,R., Filipovska,A., Brown,S., Prime,T.A., Robinson,A.J., Fearnley,I.M., Murphy,M.P. (2008) Complex I within oxidatively stressed bovine heart mitochondria is glutathionylated on Cys-531 and Cys-704 of the 75-kDa subunit: potential role of CYS residues in decreasing oxidative damage. J.Biol.Chem. 283(36): 24801-24815.

Iantomasi,T., Favilli,F., Marraccini,P., Magaldi,T., Bruni,P., Vincenzini,M.T. (1997) Glutathione transport system in human small intestine epithelial cells. Biochim.Biophys.Acta 1330(2): 274-283.

Ivarsson,R., Quintens,R., Dejonghe,S., Tsukamoto,K., in,'., V, Renstrom,E., Schuit,F.C. (2005) Redox control of exocytosis: regulatory role of NADPH, thioredoxin, and glutaredoxin. Diabetes 54(7): 2132-2142.

Jahngen-Hodge,J., Obin,M.S., Gong,X., Shang,F., Nowell,T.R., Jr., Gong,J., Abasi,H., Blumberg,J., Taylor,A. (1997) Regulation of ubiquitin-conjugating enzymes by glutathione following oxidative stress. J.Biol.Chem. 272(45): 28218-28226.

Janson,J., Laedtke,T., Parisi,J.E., O'Brien,P., Petersen,R.C., Butler,P.C. (2004) Increased risk of type 2 diabetes in Alzheimer disease. Diabetes 53(2): 474- 481.

Jao,S.C., English Ospina,S.M., Berdis,A.J., Starke,D.W., Post,C.B., Mieyal,J.J. (2006) Computational and mutational analysis of human glutaredoxin (thioltransferase): probing the molecular basis of the low pKa of cysteine 22 and its role in catalysis. Biochemistry 45(15): 4785-4796.

Jeon,K.I., Byun,M.S., Jue,D.M. (2003) Gold compound auranofin inhibits IkappaB kinase (IKK) by modifying Cys-179 of IKKbeta subunit. Exp.Mol.Med. 35(2): 61-66.

Jockusch,B.M., Schoenenberger,C.A., Stetefeld,J., Aebi,U. (2006) Tracking down the different forms of nuclear actin. Trends Cell Biol. 16(8): 391-396.

248

Joussen,A.M., Poulaki,V., Le,M.L., Koizumi,K., Esser,C., Janicki,H., Schraermeyer,U., Kociok,N., Fauser,S., Kirchhof,B., Kern,T.S., Adamis,A.P. (2004) A central role for inflammation in the pathogenesis of diabetic retinopathy. FASEB J. 18(12): 1450-1452.

Joussen,A.M., Poulaki,V., Mitsiades,N., Kirchhof,B., Koizumi,K., Dohmen,S., Adamis,A.P. (2002) Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression. FASEB J. 16(3): 438-440.

Jung,C.H., Thomas,J.A. (1996) S-glutathiolated hepatocyte proteins and insulin disulfides as substrates for reduction by glutaredoxin, thioredoxin, protein disulfide isomerase, and glutathione. Arch.Biochem.Biophys. 335(1): 61-72.

Kaneto,H., Katakami,N., Kawamori,D., Miyatsuka,T., Sakamoto,K., Matsuoka,T.A., Matsuhisa,M., Yamasaki,Y. (2007) Involvement of oxidative stress in the pathogenesis of diabetes. Antioxid.Redox.Signal. 9(3): 355-366.

Kannan,R., Bao,Y., Wang,Y., Sarthy,V.P., Kaplowitz,N. (1999) Protection from oxidant injury by sodium-dependent GSH uptake in retinal Muller cells. Exp.Eye Res. 68(5): 609-616.

Kannan,R., Chakrabarti,R., Tang,D., Kim,K.J., Kaplowitz,N. (2000) GSH transport in human cerebrovascular endothelial cells and human astrocytes: evidence for luminal localization of Na+-dependent GSH transport in HCEC. Brain Res. 852(2): 374-382.

Kapahi,P., Takahashi,T., Natoli,G., Adams,S.R., Chen,Y., Tsien,R.Y., Karin,M. (2000) Inhibition of NF-kappa B activation by arsenite through reaction with a critical cysteine in the activation loop of Ikappa B kinase. J.Biol.Chem. 275(46): 36062-36066.

Kauffmann,D.J., van Meurs,J.C., Mertens,D.A., Peperkamp,E., Master,C., Gerritsen,M.E. (1994) Cytokines in vitreous humor: interleukin-6 is elevated in proliferative vitreoretinopathy. Invest Ophthalmol.Vis.Sci. 35(3): 900-906.

Kenchappa,R.S., Diwakar,L., Annepu,J., Ravindranath,V. (2004) Estrogen and neuroprotection: higher constitutive expression of glutaredoxin in female mice offers protection against MPTP-mediated neurodegeneration. FASEB J. 18(10): 1102-1104.

Kenchappa,R.S., Ravindranath,V. (2003) Glutaredoxin is essential for maintenance of brain mitochondrial complex I: studies with MPTP. FASEB J. 17(6): 717- 719.

Kern,T.S., Engerman,R.L. (2001) Pharmacological inhibition of diabetic retinopathy: aminoguanidine and aspirin. Diabetes 50(7): 1636-1642.

249

King,G.L., Shiba,T., Oliver,J., Inoguchi,T., Bursell,S.E. (1994) Cellular and molecular abnormalities in the vascular endothelium of diabetes mellitus. Annu.Rev.Med. 45: 179-188.

Kirkpatrick,D.S., Gerber,S.A., Gygi,S.P. (2005) The absolute quantification strategy: a general procedure for the quantification of proteins and post- translational modifications. Methods 35(3): 265-273.

Klatt,P., Lamas,S. (2000) Regulation of protein function by S-glutathiolation in response to oxidative and nitrosative stress. Eur.J.Biochem. 267(16): 4928- 4944.

Klatt,P., Molina,E.P., De Lacoba,M.G., Padilla,C.A., Martinez-Galesteo,E., Barcena,J.A., Lamas,S. (1999) Redox regulation of c-Jun DNA binding by reversible S-glutathiolation. FASEB J. 13(12): 1481-1490.

Kolenko,V., Bloom,T., Rayman,P., Bukowski,R., Hsi,E., Finke,J. (1999) Inhibition of NF-kappa B activity in human T lymphocytes induces caspase-dependent apoptosis without detectable activation of caspase-1 and -3. J.Immunol. 163(2): 590-598.

Kowluru,R.A., Tang,J., Kern,T.S. (2001) Abnormalities of retinal metabolism in diabetes and experimental galactosemia. VII. Effect of long-term administration of antioxidants on the development of retinopathy. Diabetes 50(8): 1938-1942.

Kowluru,V., Kowluru,R.A. (2007) Increased oxidative stress in diabetes regulates activation of a small molecular weight G-protein, H-Ras, in the retina. Mol.Vis. 13: 602-610.

Koya,D., King,G.L. (1998) Protein kinase C activation and the development of diabetic complications. Diabetes 47(6): 859-866.

Kusner,L.L., Sarthy,V.P., Mohr,S. (2004) Nuclear translocation of glyceraldehyde- 3-phosphate dehydrogenase: a role in high glucose-induced apoptosis in retinal Muller cells. Invest Ophthalmol.Vis.Sci. 45(5): 1553-1561.

Lawrence,T., Bebien,M., Liu,G.Y., Nizet,V., Karin,M. (2005) IKKalpha limits macrophage NF-kappaB activation and contributes to the resolution of inflammation. Nature 434(7037): 1138-1143.

Lee,K., Ozanne,S.E., Hales,C.N., Ashford,M.L. (1994) Effects of chemical modification of amino and sulfhydryl groups on KATP channel function and sulfonylurea binding in CRI-G1 insulin-secreting cells. J Membr.Biol. 139(3): 167-181.

Li,X., Massa,P.E., Hanidu,A., Peet,G.W., Aro,P., Savitt,A., Mische,S., Li,J., Marcu,K.B. (2002) IKKalpha, IKKbeta, and NEMO/IKKgamma are each

250

required for the NF-kappa B-mediated inflammatory response program. J.Biol.Chem. 277(47): 45129-45140.

Li,X., Xu,Z., Li,S., Rozanski,G.J. (2005) Redox regulation of Ito remodeling in diabetic rat heart. Am.J.Physiol Heart Circ.Physiol 288(3): H1417-H1424.

Li,Z.W., Chu,W., Hu,Y., Delhase,M., Deerinck,T., Ellisman,M., Johnson,R., Karin,M. (1999) The IKKbeta subunit of IkappaB kinase (IKK) is essential for nuclear factor kappaB activation and prevention of apoptosis. J.Exp.Med. 189(11): 1839-1845.

Liu,B., Hannun,Y.A. (1997) Inhibition of the neutral magnesium-dependent sphingomyelinase by glutathione. J.Biol.Chem. 272(26): 16281-16287.

Liu,S.Y., Tsai,M.Y., Chuang,K.P., Huang,Y.F., Shieh,C.C. (2008) Ligand binding of leukocyte integrin very late antigen-4 involves exposure of sulfhydryl groups and is subject to redox modulation. Eur.J.Immunol. 38(2): 410-423.

Lu,S.C., Bao,Y., Huang,Z.Z., Sarthy,V.P., Kannan,R. (1999) Regulation of gamma- glutamylcysteine synthetase subunit gene expression in retinal Muller cells by oxidative stress. Invest Ophthalmol.Vis.Sci. 40(8): 1776-1782.

Lu,S.C., Sun,W.M., Nagineni,C.N., Hooks,J.J., Kannan,R. (1995) Bidirectional glutathione transport by cultured human retinal pigment epithelial cells. Invest Ophthalmol.Vis.Sci. 36(12): 2523-2530.

Mackness,B., Mackness,M.I., Arrol,S., Turkie,W., Julier,K., Abuasha,B., Miller,J.E., Boulton,A.J., Durrington,P.N. (1998) Serum paraoxonase (PON1) 55 and 192 polymorphism and paraoxonase activity and concentration in non-insulin dependent diabetes mellitus. Atherosclerosis 139(2): 341-349.

Major,C.D., Wolf,B.A. (2001) Interleukin-1beta stimulation of c-Jun NH(2)- terminal kinase activity in insulin-secreting cells: evidence for cytoplasmic restriction. Diabetes 50(12): 2721-2728.

Malgorzata Goczalik,I., Raap,M., Weick,M., Milenkovic,I., Heidmann,J., Enzmann,V., Wiedemann,P., Reichenbach,A., Francke,M. (2005) The activation of IL-8 receptors in cultured guinea pig Muller glial cells is modified by signals from retinal pigment epithelium. J.Neuroimmunol. 161(1- 2): 49-60.

Marfella,R., Cacciapuoti,F., Grassia,A., Manfredi,E., De Maio,G., Caruso,G., Pepe,M., Nittolo,G., Cacciapuoti,F. (2006) Role of the ubiquitin-proteasome system in carotid plaque instability in diabetic patients. Acta Cardiol. 61(6): 630-636.

251

Mariathasan,S., Monack,D.M. (2007) Inflammasome adaptors and sensors: intracellular regulators of infection and inflammation. Nat.Rev.Immunol. 7(1): 31-40.

Martins,I.J., Hone,E., Foster,J.K., Sunram-Lea,S.I., Gnjec,A., Fuller,S.J., Nolan,D., Gandy,S.E., Martins,R.N. (2006) Apolipoprotein E, cholesterol metabolism, diabetes, and the convergence of risk factors for Alzheimer's disease and cardiovascular disease. Mol.Psychiatry 11(8): 721-736.

Matsuo,M., Tucker,S.J., Ashcroft,F.M., Amachi,T., Ueda,K. (1999) NEM modification prevents high-affinity ATP binding to the first nucleotide binding fold of the sulphonylurea receptor, SUR1. FEBS Lett. 458(3): 292- 294.

Meissner,F., Molawi,K., Zychlinsky,A. (2008) Superoxide dismutase 1 regulates caspase-1 and endotoxic shock. Nat.Immunol. 9(8): 866-872.

Meister,A., Anderson,M.E. (1983) Glutathione. Annu.Rev.Biochem. 52: 711-760.

Mieyal,J.J., Gallogly,M.M., Qanungo,S., Sabens,E.A., Shelton,M.D. (2008) Molecular mechanisms and clinical implications of reversible protein S- glutathionylation. Antioxid.Redox.Signal. 10(11): 1941-1988.

Mieyal,J.J., Srinivasan,U., Starke,D.W., Gravina,S.A., Mieyal,P.A. (1995) Glutathionyl specificity of the Thioltransferases: Mechanistic and Physiological Implications. In: Packer,L., Cadenas,E. (eds). pp. 305-372. Marcel Dekker, Inc.

Mieyal,J.J., Starke,D.W., Gravina,S.A., Dothey,C., Chung,J.S. (1991) Thioltransferase in human red blood cells: purification and properties. Biochemistry 30(25): 6088-6097.

Miklossy,J., Doudet,D.D., Schwab,C., Yu,S., McGeer,E.G., McGeer,P.L. (2006) Role of ICAM-1 in persisting inflammation in Parkinson disease and MPTP monkeys. Exp.Neurol. 197(2): 275-283.

Minamino,T., Yoshida,T., Tateno,K., Miyauchi,H., Zou,Y., Toko,H., Komuro,I. (2003) Ras induces vascular smooth muscle cell senescence and inflammation in human atherosclerosis. Circulation 108(18): 2264-2269.

Miyagi,M., Rao,K.C. (2007) Proteolytic 18O-labeling strategies for quantitative proteomics. Mass Spectrom.Rev. 26(1): 121-136.

Mohr,S., Hallak,H., de Boitte,A., Lapetina,E.G., Brune,B. (1999) Nitric oxide- induced S-glutathionylation and inactivation of glyceraldehyde-3-phosphate dehydrogenase. J.Biol.Chem. 274(14): 9427-9430.

252

Mohr,S., Xi,X., Tang,J., Kern,T.S. (2002) Caspase activation in retinas of diabetic and galactosemic mice and diabetic patients. Diabetes 51(4): 1172-1179.

Moorthy,A.K., Savinova,O.V., Ho,J.Q., Wang,V.Y., Vu,D., Ghosh,G. (2006) The 20S proteasome processes NF-kappaB1 p105 into p50 in a translation- independent manner. EMBO J. 25(9): 1945-1956.

Morigi,M., Angioletti,S., Imberti,B., Donadelli,R., Micheletti,G., Figliuzzi,M., Remuzzi,A., Zoja,C., Remuzzi,G. (1998) Leukocyte-endothelial interaction is augmented by high glucose concentrations and hyperglycemia in a NF-kB- dependent fashion. J.Clin.Invest 101(9): 1905-1915.

Mukherjee,T.K., Mishra,A.K., Mukhopadhyay,S., Hoidal,J.R. (2007) High concentration of antioxidants N-acetylcysteine and mitoquinone-Q induces intercellular adhesion molecule 1 and oxidative stress by increasing intracellular glutathione. J.Immunol. 178(3): 1835-1844.

Murata,H., Ihara,Y., Nakamura,H., Yodoi,J., Sumikawa,K., Kondo,T. (2003) Glutaredoxin exerts an antiapoptotic effect by regulating the redox state of Akt. J.Biol.Chem. 278(50): 50226-50233.

Muro,S., Gajewski,C., Koval,M., Muzykantov,V.R. (2005) ICAM-1 recycling in endothelial cells: a novel pathway for sustained intracellular delivery and prolonged effects of drugs. Blood 105(2): 650-658.

Murugeswari,P., Shukla,D., Rajendran,A., Kim,R., Namperumalsamy,P., Muthukkaruppan,V. (2008) PROINFLAMMATORY CYTOKINES AND ANGIOGENIC AND ANTI-ANGIOGENIC FACTORS IN VITREOUS OF PATIENTS WITH PROLIFERATIVE DIABETIC RETINOPATHY AND EALES' DISEASE. Retina 28(6): 817-824.

Mysliwiec,M., Balcerska,A., Zorena,K., Mysliwska,J., Lipowski,P., Raczynska,K. (2008) The role of vascular endothelial growth factor, tumor necrosis factor alpha and interleukin-6 in pathogenesis of diabetic retinopathy. Diabetes Res.Clin.Pract. 79(1): 141-146.

Nakamura,H., Bai,J., Nishinaka,Y., Ueda,S., Sasada,T., Ohshio,G., Imamura,M., Takabayashi,A., Yamaoka,Y., Yodoi,J. (2000) Expression of thioredoxin and glutaredoxin, redox-regulating proteins, in pancreatic cancer. Cancer Detect.Prev. 24(1): 53-60.

Nakazawa,T., Hisatomi,T., Nakazawa,C., Noda,K., Maruyama,K., She,H., Matsubara,A., Miyahara,S., Nakao,S., Yin,Y., Benowitz,L., Hafezi- Moghadam,A., Miller,J.W. (2007) Monocyte chemoattractant protein 1 mediates retinal detachment-induced photoreceptor apoptosis. Proc.Natl.Acad.Sci.U.S.A 104(7): 2425-2430.

253

Nakazawa,T., Matsubara,A., Noda,K., Hisatomi,T., She,H., Skondra,D., Miyahara,S., Sobrin,L., Thomas,K.L., Chen,D.F., Grosskreutz,C.L., Hafezi- Moghadam,A., Miller,J.W. (2006) Characterization of cytokine responses to retinal detachment in rats. Mol.Vis. 12: 867-878.

Nelson,T.E., Campbell,I.L., Gruol,D.L. (1999) Altered physiology of Purkinje neurons in cerebellar slices from transgenic mice with chronic central nervous system expression of interleukin-6. Neuroscience 89(1): 127-136.

Newman,E., Reichenbach,A. (1996) The Muller cell: a functional element of the retina. Trends Neurosci 19(8): 307-312.

Newman,S.F., Sultana,R., Perluigi,M., Coccia,R., Cai,J., Pierce,W.M., Klein,J.B., Turner,D.M., Butterfield,D.A. (2007) An increase in S-glutathionylated proteins in the Alzheimer's disease inferior parietal lobule, a proteomics approach. J Neurosci Res. 85(7): 1506-1514.

Nulton-Persson,A.C., Starke,D.W., Mieyal,J.J., Szweda,L.I. (2003) Reversible inactivation of alpha-ketoglutarate dehydrogenase in response to alterations in the mitochondrial glutathione status. Biochemistry 42(14): 4235-4242.

Obin,M., Shang,F., Gong,X., Handelman,G., Blumberg,J., Taylor,A. (1998) Redox regulation of ubiquitin-conjugating enzymes: mechanistic insights using the thiol-specific oxidant diamide. FASEB J. 12(7): 561-569.

Okuda,M., Inoue,N., Azumi,H., Seno,T., Sumi,Y., Hirata,K., Kawashima,S., Hayashi,Y., Itoh,H., Yodoi,J., Yokoyama,M. (2001) Expression of glutaredoxin in human coronary arteries: its potential role in antioxidant protection against atherosclerosis. Arterioscler.Thromb.Vasc.Biol. 21(9): 1483-1487.

Ondrej,V., Lukasova,E., Krejci,J., Kozubek,S. (2008) Intranuclear trafficking of plasmid DNA is mediated by nuclear polymeric proteins lamins and actin. Acta Biochim.Pol. 55(2): 307-315.

Packer,L., Kraemer,K., Rimbach,G. (2001) Molecular aspects of lipoic acid in the prevention of diabetes complications. Nutrition 17(10): 888-895.

Pan,S., Berk,B.C. (2007) Glutathiolation regulates tumor necrosis factor-alpha- induced caspase-3 cleavage and apoptosis: key role for glutaredoxin in the death pathway. Circ.Res. 100(2): 213-219.

Pannicke,T., Iandiev,I., Wurm,A., Uckermann,O., vom,H.F., Reichenbach,A., Wiedemann,P., Hammes,H.P., Bringmann,A. (2006) Diabetes alters osmotic swelling characteristics and membrane conductance of glial cells in rat retina. Diabetes 55(3): 633-639.

254

Park,J.B., Levine,M. (1997) The human glutaredoxin gene: determination of its organization, transcription start point, and promoter analysis. Gene 197(1- 2): 189-193.

Pederson,T., Aebi,U. (2005) Nuclear actin extends, with no contraction in sight. Mol.Biol.Cell 16(11): 5055-5060.

Pei,Z., Liu,G., Lubben,T.H., Szczepankiewicz,B.G. (2004) Inhibition of protein tyrosine phosphatase 1B as a potential treatment of diabetes and obesity. Curr.Pharm.Des 10(28): 3481-3504.

Peltoniemi,M., Kaarteenaho-Wiik,R., Saily,M., Sormunen,R., Paakko,P., Holmgren,A., Soini,Y., Kinnula,V.L. (2004) Expression of glutaredoxin is highly cell specific in human lung and is decreased by transforming growth factor-beta in vitro and in interstitial lung diseases in vivo. Hum.Pathol. 35(8): 1000-1007.

Peltoniemi,M.J., Rytila,P.H., Harju,T.H., Soini,Y.M., Salmenkivi,K.M., Ruddock,L.W., Kinnula,V.L. (2006) Modulation of glutaredoxin in the lung and sputum of cigarette smokers and chronic obstructive pulmonary disease. Respir.Res. 7: 133-144.

Percival,M.D., Ouellet,M., Campagnolo,C., Claveau,D., Li,C. (1999) Inhibition of cathepsin K by nitric oxide donors: evidence for the formation of mixed disulfides and a sulfenic acid. Biochemistry 38(41): 13574-13583.

Petrilli,V., Dostert,C., Muruve,D.A., Tschopp,J. (2007) The inflammasome: a danger sensing complex triggering innate immunity. Curr.Opin.Immunol. 19(6): 615-622.

Pineda-Molina,E., Klatt,P., Vazquez,J., Marina,A., Garcia,d.L., Perez-Sala,D., Lamas,S. (2001) Glutathionylation of the p50 subunit of NF-kappaB: a mechanism for redox-induced inhibition of DNA binding. Biochemistry 40(47): 14134-14142.

Poitry-Yamate,C.L., Poitry,S., Tsacopoulos,M. (1995) Lactate released by Muller glial cells is metabolized by photoreceptors from mammalian retina. J.Neurosci 15(7 Pt 2): 5179-5191.

Pow,D.V., Crook,D.K. (1995) Immunocytochemical evidence for the presence of high levels of reduced glutathione in radial glial cells and horizontal cells in the rabbit retina. Neurosci.Lett. 193(1): 25-28.

Pujades,C., Teixido,J., Bazzoni,G., Hemler,M.E. (1996) Integrin alpha 4 cysteines 278 and 717 modulate VLA-4 ligand binding and also contribute to alpha 4/180 formation. Biochem.J. 313 ( Pt 3): 899-908.

255

Qanungo,S., Starke,D.W., Pai,H.V., Mieyal,J.J., Nieminen,A.L. (2007) Glutathione supplementation potentiates hypoxic apoptosis by S-glutathionylation of p65- NFkappaB. J.Biol.Chem. 282(25): 18427-18436.

Rahman,I. (2005) Regulation of glutathione in inflammation and chronic lung diseases. Mutat.Res. 579(1-2): 58-80.

Rao,R.K., Clayton,L.W. (2002) Regulation of protein phosphatase 2A by hydrogen peroxide and glutathionylation. Biochem.Biophys.Res.Commun. 293(1): 610- 616.

Ray,A. (2001) Response to 'Specificity of SN50 for NF-kB?'. In. pp. 471-472.

Reynaert,N.L., van,d., V, Guala,A.S., McGovern,T., Hristova,M., Pantano,C., Heintz,N.H., Heim,J., Ho,Y.S., Matthews,D.E., Wouters,E.F., Janssen- Heininger,Y.M. (2006) Dynamic redox control of NF-kappaB through glutaredoxin-regulated S-glutathionylation of inhibitory kappaB kinase beta. Proc.Natl.Acad.Sci.U.S.A 103(35): 13086-13091.

Reynaert,N.L., Wouters,E.F., Janssen-Heininger,Y.M. (2007) Modulation of glutaredoxin-1 expression in a mouse model of allergic airway disease. Am.J.Respir.Cell Mol.Biol. 36(2): 147-151.

Rohl,M., Pasparakis,M., Baudler,S., Baumgartl,J., Gautam,D., Huth,M., De,L.R., Krone,W., Rajewsky,K., Bruning,J.C. (2004) Conditional disruption of IkappaB kinase 2 fails to prevent obesity-induced insulin resistance. J.Clin.Invest 113(3): 474-481.

Rokutan,K., Johnston,R.B., Jr., Kawai,K. (1994) Oxidative stress induces S- thiolation of specific proteins in cultured gastric mucosal cells. Am.J.Physiol 266(2 Pt 1): G247-G254.

Romeo,G., Liu,W.H., Asnaghi,V., Kern,T.S., Lorenzi,M. (2002) Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes. Diabetes 51(7): 2241-2248.

Rose-John,S., Scheller,J., Elson,G., Jones,S.A. (2006) Interleukin-6 biology is coordinated by membrane-bound and soluble receptors: role in inflammation and cancer. J.Leukoc.Biol. 80(2): 227-236.

Ross,P.L., Huang,Y.N., Marchese,J.N., Williamson,B., Parker,K., Hattan,S., Khainovski,N., Pillai,S., Dey,S., Daniels,S., Purkayastha,S., Juhasz,P., Martin,S., Bartlet-Jones,M., He,F., Jacobson,A., Pappin,D.J. (2004) Multiplexed protein quantitation in Saccharomyces cerevisiae using amine- reactive isobaric tagging reagents. Mol.Cell Proteomics. 3(12): 1154-1169.

256

Rossi,A., Kapahi,P., Natoli,G., Takahashi,T., Chen,Y., Karin,M., Santoro,M.G. (2000) Anti-inflammatory cyclopentenone prostaglandins are direct inhibitors of IkappaB kinase. Nature 403(6765): 103-108.

Rozenberg,O., Aviram,M. (2006) S-Glutathionylation regulates HDL-associated paraoxonase 1 (PON1) activity. Biochem.Biophys.Res.Commun. 351(2): 492- 498.

Rungger-Brandle,E., Dosso,A.A., Leuenberger,P.M. (2000) Glial reactivity, an early feature of diabetic retinopathy. Invest Ophthalmol.Vis.Sci. 41(7): 1971-1980.

Sakai,T., Matsuura,B., Onji,M. (1998) Serum paraoxonase activity and genotype distribution in Japanese patients with diabetes mellitus. Intern.Med. 37(7): 581-584.

Sanchez,R.N., Chan,C.K., Garg,S., Kwong,J.M., Wong,M.J., Sadun,A.A., Lam,T.T. (2003) Interleukin-6 in retinal ischemia reperfusion injury in rats. Invest Ophthalmol.Vis.Sci. 44(9): 4006-4011.

Sano,H., Nakagawa,N., Chiba,R., Kurasawa,K., Saito,Y., Iwamoto,I. (1998) Cross- linking of intercellular adhesion molecule-1 induces interleukin-8 and RANTES production through the activation of MAP kinases in human vascular endothelial cells. Biochem.Biophys.Res.Commun. 250(3): 694-698.

Sappington,R.M., Calkins,D.J. (2006) Pressure-induced regulation of IL-6 in retinal glial cells: involvement of the ubiquitin/proteasome pathway and NFkappaB. Invest Ophthalmol.Vis.Sci. 47(9): 3860-3869.

Sarthy,V.P., Dudley,V.J., Tanaka,K. (2004) Retinal glucose metabolism in mice lacking the L-glutamate/aspartate transporter. Vis.Neurosci. 21(4): 637-643.

Saura,M., Zaragoza,C., Bao,C., Herranz,B., Rodriguez-Puyol,M., Lowenstein,C.J. (2006) Stat3 mediates interleukin-6 [correction of interelukin-6] inhibition of human endothelial nitric-oxide synthase expression. J.Biol.Chem. 281(40): 30057-30062.

Sawada,M., Imamura,K., Nagatsu,T. (2006) Role of cytokines in inflammatory process in Parkinson's disease. J.Neural Transm.Suppl(70): 373-381.

Scheidereit,C. (2006) IkappaB kinase complexes: gateways to NF-kappaB activation and transcription. Oncogene 25(51): 6685-6705.

Schmid,H., Boucherot,A., Yasuda,Y., Henger,A., Brunner,B., Eichinger,F., Nitsche,A., Kiss,E., Bleich,M., Grone,H.J., Nelson,P.J., Schlondorff,D., Cohen,C.D., Kretzler,M. (2006) Modular activation of nuclear factor-kappaB transcriptional programs in human diabetic nephropathy. Diabetes 55(11): 2993-3003.

257

Schuppe,I., Moldeus,P., Cotgreave,I.A. (1992) Protein-specific S-thiolation in human endothelial cells during oxidative stress. Biochem.Pharmacol. 44(9): 1757- 1764.

Schutte,M., Werner,P. (1998) Redistribution of glutathione in the ischemic rat retina. Neurosci.Lett. 246(1): 53-56.

Seki,T., Hinohara,Y., Taki,C., Nakatani,M., Ozawa,M., Nishimura,S., Takaki,A., Itho,H., Takenoya,F., Shioda,S. (2006) PACAP stimulates the release of interleukin-6 in cultured rat Muller cells. Ann.N.Y.Acad.Sci. 1070: 535-539.

Seong,S.Y., Matzinger,P. (2004) Hydrophobicity: an ancient damage-associated molecular pattern that initiates innate immune responses. Nat.Rev.Immunol. 4(6): 469-478.

Shelton,M.D., Chock,P.B., Mieyal,J.J. (2005) Glutaredoxin: role in reversible protein s-glutathionylation and regulation of redox signal transduction and protein translocation. Antioxid.Redox.Signal. 7(3-4): 348-366.

Shelton,M.D., Kern,T.S., Mieyal,J.J. (2007) Glutaredoxin regulates nuclear factor kappa-B and intercellular adhesion molecule in Muller cells: model of diabetic retinopathy. J.Biol.Chem. 282(17): 12467-12474.

Shelton,M.D., Mieyal,J.J. (2008) Regulation by reversible S-glutathionylation: molecular targets implicated in inflammatory diseases. Mol.Cells 25(3): 332- 346.

Shepherd,C.E., Grace,E.M., Mann,D.M., Halliday,G.M. (2007) Relationship between neuronal loss and 'inflammatory plaques' in early onset Alzheimer's disease. Neuropathol.Appl.Neurobiol. 33(3): 328-333.

Shepherd,P.R., Withers,D.J., Siddle,K. (1998) Phosphoinositide 3-kinase: the key switch mechanism in insulin signalling. Biochem.J. 333 ( Pt 3): 471-490.

Shin,H.M., Minter,L.M., Cho,O.H., Gottipati,S., Fauq,A.H., Golde,T.E., Sonenshein,G.E., Osborne,B.A. (2006) Notch1 augments NF-kappaB activity by facilitating its nuclear retention. EMBO J. 25(1): 129-138.

Shishodia,S., Aggarwal,B.B. (2002) Nuclear factor-kappaB activation: a question of life or death. J.Biochem.Mol.Biol. 35(1): 28-40.

Shoelson,S.E., Lee,J., Yuan,M. (2003) Inflammation and the IKK beta/I kappa B/NF-kappa B axis in obesity- and diet-induced insulin resistance. Int.J.Obes.Relat Metab Disord. 27 Suppl 3: S49-S52.

Shrikant,P., Benos,D.J., Tang,L.P., Benveniste,E.N. (1996) HIV glycoprotein 120 enhances intercellular adhesion molecule-1 gene expression in glial cells.

258

Involvement of Janus kinase/signal transducer and activator of transcription and protein kinase C signaling pathways. J.Immunol. 156(3): 1307-1314.

Shrikant,P., Weber,E., Jilling,T., Benveniste,E.N. (1995) Intercellular adhesion molecule-1 gene expression by glial cells. Differential mechanisms of inhibition by IL-10 and IL-6. J.Immunol. 155(3): 1489-1501.

Singh,S., Venketesh,S., Verma,J.S., Verma,M., Lellamma,C.O., Goel,R.C. (2007) Paraoxonase (PON1) activity in north west Indian Punjabis with coronary artery disease & type 2 diabetes mellitus. Indian J.Med.Res. 125(6): 783-787.

Snyers,L., Fontaine,V., Content,J. (1989) Modulation of interleukin-6 receptors in human cells. Ann.N.Y.Acad.Sci. 557: 388-393.

Song,J.J., Lee,Y.J. (2003) Differential role of glutaredoxin and thioredoxin in metabolic oxidative stress-induced activation of apoptosis signal-regulating kinase 1. Biochem.J. 373(Pt 3): 845-853.

Song,J.J., Rhee,J.G., Suntharalingam,M., Walsh,S.A., Spitz,D.R., Lee,Y.J. (2002) Role of glutaredoxin in metabolic oxidative stress. Glutaredoxin as a sensor of oxidative stress mediated by H2O2. J.Biol.Chem. 277(48): 46566-46575.

Srinivasan,S., Hatley,M.E., Reilly,K.B., Danziger,E.C., Hedrick,C.C. (2004) Modulation of PPARalpha expression and inflammatory interleukin-6 production by chronic glucose increases monocyte/endothelial adhesion. Arterioscler.Thromb.Vasc.Biol. 24(5): 851-857.

Srinivasan,U., Mieyal,P.A., Mieyal,J.J. (1997) pH profiles indicative of rate-limiting nucleophilic displacement in thioltransferase catalysis. Biochemistry 36(11): 3199-3206.

Srivastava,S.K., Ramana,K.V., Bhatnagar,A. (2005) Role of aldose reductase and oxidative damage in diabetes and the consequent potential for therapeutic options. Endocr.Rev. 26(3): 380-392.

Starke,D.W., Chock,P.B., Mieyal,J.J. (2003) Glutathione-thiyl radical scavenging and transferase properties of human glutaredoxin (thioltransferase). Potential role in redox signal transduction. J.Biol.Chem. 278(17): 14607- 14613.

Taga,T. (1992) IL6 signalling through IL6 receptor and receptor-associated signal transducer, gp130. Res.Immunol. 143(7): 737-739.

Taga,T., Hibi,M., Hirata,Y., Yamasaki,K., Yasukawa,K., Matsuda,T., Hirano,T., Kishimoto,T. (1989) Interleukin-6 triggers the association of its receptor with a possible signal transducer, gp130. Cell 58(3): 573-581.

259

Takashima,Y., Hirota,K., Nakamura,H., Nakamura,T., Akiyama,K., Cheng,F.S., Maeda,M., Yodoi,J. (1999) Differential expression of glutaredoxin and thioredoxin during monocytic differentiation. Immunol.Lett. 68(2-3): 397-401.

Tao,L., English,A.M. (2004) Protein S-glutathiolation triggered by decomposed S- nitrosoglutathione. Biochemistry 43(13): 4028-4038.

Taylor,E.R., Hurrell,F., Shannon,R.J., Lin,T.K., Hirst,J., Murphy,M.P. (2003) Reversible glutathionylation of complex I increases mitochondrial superoxide formation. J.Biol.Chem. 278(22): 19603-19610.

Teppo,A.M., von,W.E., Honkanen,E., Ahonen,J., Gronhagen-Riska,C. (2001) Soluble intercellular adhesion molecule-1 (sICAM-1) after kidney transplantation: the origin and role of urinary sICAM-1? Transplantation 71(8): 1113-1119.

Tesch,G.H. (2007) Role of macrophages in complications of type 2 diabetes. Clin.Exp.Pharmacol.Physiol 34(10): 1016-1019.

Thannickal,V.J., Fanburg,B.L. (2000) Reactive oxygen species in cell signaling. Am.J.Physiol Lung Cell Mol.Physiol 279(6): L1005-L1028.

Thomas,J.A., Poland,B., Honzatko,R. (1995) Protein sulfhydryls and their role in the antioxidant function of protein S-thiolation. Arch.Biochem.Biophys. 319(1): 1-9.

Thomas,J.S., Mallis,R. (2003) S-Nitrosylation, and Irreversible Sulfhydryl Oxidation: Roles in Redox Regulation. In: Gitler,C., Danon,A. (eds) Cellular Implications of Redox Signaling. pp. 258-264. Imperial College Press, London.

Tomlinson,D.R., Gardiner,N.J. (2008) Glucose neurotoxicity. Nat Rev Neurosci 9(1): 36-45.

Tong,X., Ying,J., Pimentel,D.R., Trucillo,M., Adachi,T., Cohen,R.A. (2008) High glucose oxidizes SERCA cysteine-674 and prevents inhibition by nitric oxide of smooth muscle cell migration. J.Mol.Cell Cardiol. 44(2): 361-369.

Tonks,N.K. (2003) PTP1B: from the sidelines to the front lines! FEBS Lett. 546(1): 140-148.

Torgerson,T.R., Colosia,A.D., Donahue,J.P., Lin,Y.Z., Hawiger,J. (1998) Regulation of NF-kappa B, AP-1, NFAT, and STAT1 nuclear import in T lymphocytes by noninvasive delivery of peptide carrying the nuclear localization sequence of NF-kappa B p50. J.Immunol. 161(11): 6084-6092.

Trapp,S., Tucker,S.J., Ashcroft,F.M. (1998) Mechanism of ATP-sensitive K channel inhibition by sulfhydryl modification. J.Gen.Physiol 112(3): 325-332.

260

Urata,Y., Ihara,Y., Murata,H., Goto,S., Koji,T., Yodoi,J., Inoue,S., Kondo,T. (2006) 17Beta-estradiol protects against oxidative stress-induced cell death through the glutathione/glutaredoxin-dependent redox regulation of Akt in myocardiac H9c2 cells. J.Biol.Chem. 281(19): 13092-13102.

van Beijnum,J.R., Buurman,W.A., Griffioen,A.W. (2008) Convergence and amplification of toll-like receptor (TLR) and receptor for advanced glycation end products (RAGE) signaling pathways via high mobility group B1 (HMGB1). Angiogenesis. 11(1): 91-99.

van den Berghe,N., Ouwens,D.M., Maassen,J.A., van Mackelenbergh,M.G., Sips,H.C., Krans,H.M. (1994) Activation of the Ras/mitogen-activated protein kinase signaling pathway alone is not sufficient to induce glucose uptake in 3T3-L1 adipocytes. Mol.Cell Biol. 14(4): 2372-2377.

Vincent,J.A., Mohr,S. (2007) Inhibition of caspase-1/interleukin-1beta signaling prevents degeneration of retinal capillaries in diabetes and galactosemia. Diabetes 56(1): 224-230.

Viner,R.I., Williams,T.D., Schoneich,C. (1999) Peroxynitrite modification of protein thiols: oxidation, nitrosylation, and S-glutathiolation of functionally important cysteine residue(s) in the sarcoplasmic reticulum Ca-ATPase. Biochemistry 38(38): 12408-12415.

Vojtek,A.B., Der,C.J. (1998) Increasing complexity of the Ras signaling pathway. J.Biol.Chem. 273(32): 19925-19928.

Volkers,N. (2000) Diabetes and cancer: scientists search for a possible link. J.Natl.Cancer Inst. 92(3): 192-194.

Walker,R.J., Steinle,J.J. (2007) Role of beta-adrenergic receptors in inflammatory marker expression in Muller cells. Invest Ophthalmol.Vis.Sci. 48(11): 5276- 5281.

Wan,J., Roth,A.F., Bailey,A.O., Davis,N.G. (2007) Palmitoylated proteins: purification and identification. Nat.Protoc. 2(7): 1573-1584.

Wang,F., Herrington,M., Larsson,J., Permert,J. (2003a) The relationship between diabetes and pancreatic cancer. Mol.Cancer 2: 4-9.

Wang,J., Boja,E.S., Tan,W., Tekle,E., Fales,H.M., English,S., Mieyal,J.J., Chock,P.B. (2001) Reversible glutathionylation regulates actin polymerization in A431 cells. J.Biol.Chem. 276(51): 47763-47766.

Wang,J., Pan,S., Berk,B.C. (2007a) Glutaredoxin mediates Akt and eNOS activation by flow in a glutathione reductase-dependent manner. Arterioscler.Thromb.Vasc.Biol. 27(6): 1283-1288.

261

Wang,J., Tekle,E., Oubrahim,H., Mieyal,J.J., Stadtman,E.R., Chock,P.B. (2003b) Stable and controllable RNA interference: Investigating the physiological function of glutathionylated actin. Proc.Natl.Acad.Sci.U.S.A 100(9): 5103- 5106.

Wang,X.C., Qiu,Y.H., Peng,Y.P. (2007b) Interleukin-6 protects cerebellar granule neurons from NMDA-induced neurotoxicity. Sheng Li Xue.Bao. 59(2): 150- 156.

Wang,Y., Qiao,M., Mieyal,J.J., Asmis,L.M., Asmis,R. (2006) Molecular mechanism of glutathione-mediated protection from oxidized low-density lipoprotein- induced cell injury in human macrophages: role of glutathione reductase and glutaredoxin. Free Radic.Biol.Med. 41(5): 775-785.

Ward,N.E., Stewart,J.R., Ioannides,C.G., O'Brian,C.A. (2000) Oxidant-induced S- glutathiolation inactivates protein kinase C-alpha (PKC-alpha): a potential mechanism of PKC isozyme regulation. Biochemistry 39(33): 10319-10329.

Waxman,A.B., Mahboubi,K., Knickelbein,R.G., Mantell,L.L., Manzo,N., Pober,J.S., Elias,J.A. (2003) Interleukin-11 and interleukin-6 protect cultured human endothelial cells from H2O2-induced cell death. Am.J.Respir.Cell Mol.Biol. 29(4): 513-522.

Weik,R., Neumcke,B. (1989) ATP-sensitive potassium channels in adult mouse skeletal muscle: characterization of the ATP-binding site. J.Membr.Biol. 110(3): 217-226.

Weise,J., Isenmann,S., Klocker,N., Kugler,S., Hirsch,S., Gravel,C., Bahr,M. (2000) Adenovirus-mediated expression of ciliary neurotrophic factor (CNTF) rescues axotomized rat retinal ganglion cells but does not support axonal regeneration in vivo. Neurobiol.Dis. 7(3): 212-223.

Winkler,B.S., Arnold,M.J., Brassell,M.A., Puro,D.G. (2000) Energy metabolism in human retinal Muller cells. Invest Ophthalmol.Vis.Sci. 41(10): 3183-3190.

Wurm,A., Pannicke,T., Wiedemann,P., Reichenbach,A., Bringmann,A. (2008) Glial cell-derived glutamate mediates autocrine cell volume regulation in the retina: activation by VEGF. J.Neurochem. 104(2): 386-399.

Xiao,C., Bator,C.M., Bowman,V.D., Rieder,E., He,Y., Hebert,B., Bella,J., Baker,T.S., Wimmer,E., Kuhn,R.J., Rossmann,M.G. (2001) Interaction of coxsackievirus A21 with its cellular receptor, ICAM-1. J.Virol. 75(5): 2444- 2451.

Yamasaki,K., Taga,T., Hirata,Y., Yawata,H., Kawanishi,Y., Seed,B., Taniguchi,T., Hirano,T., Kishimoto,T. (1988) Cloning and expression of the human interleukin-6 (BSF-2/IFN beta 2) receptor. Science 241(4867): 825-828.

262

Yang,J., Liao,X., Agarwal,M.K., Barnes,L., Auron,P.E., Stark,G.R. (2007) Unphosphorylated STAT3 accumulates in response to IL-6 and activates transcription by binding to NFkappaB. Genes Dev. 21(11): 1396-1408.

Yang,Y., Jao,S., Nanduri,S., Starke,D.W., Mieyal,J.J., Qin,J. (1998) Reactivity of the human thioltransferase (glutaredoxin) C7S, C25S, C78S, C82S mutant and NMR solution structure of its glutathionyl mixed disulfide intermediate reflect catalytic specificity. Biochemistry 37(49): 17145-17156.

Yaras,N., Ugur,M., Ozdemir,S., Gurdal,H., Purali,N., Lacampagne,A., Vassort,G., Turan,B. (2005) Effects of diabetes on ryanodine receptor Ca release channel (RyR2) and Ca2+ homeostasis in rat heart. Diabetes 54(11): 3082-3088.

Yoshida,A., Yoshida,S., Khalil,A.K., Ishibashi,T., Inomata,H. (1998) Role of NF- kappaB-mediated interleukin-8 expression in intraocular neovascularization. Invest Ophthalmol.Vis.Sci. 39(7): 1097-1106.

Yoshitake,S., Nanri,H., Fernando,M.R., Minakami,S. (1994) Possible differences in the regenerative roles played by thioltransferase and thioredoxin for oxidatively damaged proteins. J.Biochem.(Tokyo) 116(1): 42-46.

Yuan,M., Konstantopoulos,N., Lee,J., Hansen,L., Li,Z.W., Karin,M., Shoelson,S.E. (2001) Reversal of obesity- and diet-induced insulin resistance with salicylates or targeted disruption of Ikkbeta. Science 293(5535): 1673-1677.

Zhang,Z.Y., Lee,S.Y. (2003) PTP1B inhibitors as potential therapeutics in the treatment of type 2 diabetes and obesity. Expert.Opin.Investig.Drugs 12(2): 223-233.

Zheng,L., Howell,S.J., Hatala,D.A., Huang,K., Kern,T.S. (2007) Salicylate-based anti-inflammatory drugs inhibit the early lesion of diabetic retinopathy. Diabetes 56(2): 337-345.

Zheng,L., Szabo,C., Kern,T.S. (2004) Poly(ADP-ribose) polymerase is involved in the development of diabetic retinopathy via regulation of nuclear factor- kappaB. Diabetes 53(11): 2960-2967.

Zheng,M., Aslund,F., Storz,G. (1998) Activation of the OxyR transcription factor by reversible disulfide bond formation. Science 279(5357): 1718-1721.

263