Open Thesis.Pdf

The Pennsylvania State University

The Graduate School

College of Medicine

A ROLE FOR STANNIN IN CELLULAR SIGNALING

A Thesis in

Integrative Biosciences

Brian Reese

 2005 Brian Reese

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

August 2005

The thesis of Brian Eric Reese was reviewed and approved* by the following:

Melvin L. Billingsley Professor of Pharmacology Thesis Advisor Co-Chair of Committee

Jong K. Yun Assistant Professor of Pharmacology Co-Chair of Committee

Robert J. Milner Professor of Neural and Behavioral Sciences

James R. Connor Professor and Vice-Chair of Neurosurgery

J. Kyle Krady Assistant Professor of Neural and Behavioral Sciences

Anita K. Hopper Professor of Biochemistry and Molecular Biology Co-Director of Integrative Biosciences Graduate Program

*Signatures are on file in the Graduate School

ABSTRACT

Trimethyltin (TMT) is a selective and potent neurotoxicant capable of inducing

apoptotic cell death in the hippocampal formation, neocortex, amygdala, and olfactory

tubercle. Subtractive hybridization was used in previous studies to uncover any common

gene products present in TMT-sensitive tissues that might account for TMT’s selective

toxicity. These studies showed that the gene product Stannin (Snn) was preferentially

expressed in TMT sensitive tissues, with higher levels of Snn expression correlating with

higher regional levels of TMT sensitivity. In addition, using antisense oligonucleotides,

Snn was found to be necessary, but not sufficient, for TMT toxicity.

Snn is an 88 amino acid, membrane-bound protein with a molecular weight of

9.497 kDa. Snn is found in vertebrate species and is highly conserved across vertebrates,

with human and rat Snn showing 98% identity at the amino acid level. Snn is widely

expressed in the developing embryo; Snn expression becomes more restricted during

maturation to adulthood. In the adult, Snn is expressed in the spleen, immune cells,

brain, kidney and lung. Snn has no significant homology to any other known protein.

Tumor necrosis factor alpha (TNFα) was shown to induce Snn mRNA expression in human umbilical vein endothelial cells (HUVECs). We used quantitative, real-time

PCR (QRT-PCR) to quantify the induction of Snn by TNFα in multiple cell lines. We observed significant increases in Snn mRNA in a time-dependent manner in HUVECs and Jurkat T-cells. To better define a potential mechanism underlying this induction, chemical inhibitors of protein kinase C (PKC) were used to determine if PKC played a role in TNFα-mediated Snn gene expression. The results of these experiments indicated

iii

that one or more of the βII, δ, γ, or ε isoforms was responsible for TNFα-mediated Snn gene expression. To determine specifically which isoform(s) of PKC were involved, we utilized short interfering RNA technology (siRNA) and found that PKCε was responsible for mediating TNFα-induced Snn gene expression.

TMT can induce the expression of TNFα in mixed neuronal/glial cultures and so we hypothesized that TMT exposure would cause an increase in Snn gene expression, potentially as a necessary component of TMT toxicity. Again, QRT-PCR was used and we found that TMT significantly upregulated Snn within 9 hours of exposure. However, blocking TNFα with neutralizing antibodies resulted in only partial protection against

TMT.

By placing Snn in a defined TNFα-PKCε signaling pathway, several hypotheses arise concerning the function of Snn, such as Snn being part of a cell death or cell survival-signaling pathway. Given the high level of conservation of Snn, generation of high-affinity, Snn-specific antisera has proven difficult. In order to assess a potential role for Snn in the HUVEC response to TNFα, microarray technology was used. We found that knocking Snn down via siRNA significantly altered HUVEC gene expression in response to TNFα. After normalization and statistical analysis, we found that several genes altered by Snn knockdown are involved in the modulation of the cell cycle and cell growth. Specifically, several genes are involved with p53 and Cyclin D1, known G1/S checkpoint proteins. Functional assays showed that Snn knockdown resulted in significantly less HUVEC growth relative to other treatment. Further, analysis via flow

cytometry indicated that a significant portion of TNFα-, Snn siRNA co-treated HUVECs

were halted in the G1 phase of the cell cycle.

Together, the data presented outline a signaling pathway leading to enhanced Snn gene expression as well as a potential role for Snn in normal cellular function. A role in modulating the cell cycle would explain the high degree of conservation of Snn across vertebrate evolution as well as the tissue-specific pattern of expression of Snn during different life stages. This work details the first known interaction of Snn in a cellular signaling pathway as well as the first evidence of a potential functional role of Snn in normal cells.

TABLE OF CONTENTS

List of Figures...... …ix

List of Tables ...... …xi

List of Abbreviations …………………………………………………………………..xii

Acknowledgements...... xiv

Chapter 1: Introduction……………………………………………………………………1

Chapter 2: Literature Review……………………………………………………………..4

A. Trimethyltin 1. Organotins as toxicants…………………………………………………4 2. Trimethyltin - a potent, selective neurotoxicant………………………..5 3. Trimethyltin toxicity - cellular mechanisms……………………………6 4. Trimethyltin toxicity - behavioral effects………………………………8 5. Trimethyltin - summary…………………………………………….…..8 B. Stannin 1. The discovery of stannin and characterization of expression…………..9 2. Characterization of the stannin gene…………………………………..10 3. Stannin's role in trimethyltin toxicity……………………………….…13 4. Stannin in cellular signaling…………………………………………...14 C. Tumor necrosis factor-α 1. Tumor necrosis factor-α - origin and plieotropism…………...………17 2. Tumor necrosis factor-α receptor 1: signal transduction and major functions………………………………………………………..17 3. Tumor necrosis factor-α receptor 2: signal transduction and known functions…………………………………………………….…21 D. Protein Kinase C 1. Classes of protein kinase c and activation………………………….…22 2. Functions of protein kinase c…………………………………….……25 3. Protein kinase C - summary………………………………………..….30

Chapter 3: Protein Kinase C Epsilon Regulates TNFα-Induced Stannin Gene Expression……………………………………………………………………31 A. Introduction……………………………………………………………...……31 B. Methods 1. Cell culture………………………………………………………...…..35 2. Reagents…………………………………………………………...…..35 3. Cell viability…………………………………………………………...36

4. RNA isolation/cDNA synthesis………………………………………36 5. Quantitative real-time PCR……………………………………...……37 6. siRNA construction………………………………………………...…37 7. Transfection of siRNA…………………………………………..……39 8. Statistical analysis………………………………………………..…...39

C. Results……………………………………………………………………..…40 1. Tumor necrosis factor-α contributes to trimethyltin toxicity…….…..40 2. Trimethyltin induces stannin gene expression in human umbilical vein endothelial cells……………………………………....40 3. Stannin knockdown rescues HUVECs from trimethyltin toxicity..….41 4. Induction of stannin mRNA by tumor necrosis factor-α………….…47 5. Protein kinase c is required for TNFα-induced stannin mRNA expression…………………………………………………....52 6. Knockdown of protein kinase c epsilon via siRNA prevents TNFα-mediated stannin upregulation………………………………..55

D. Discussion………………………………………………………………..…..59

Chapter 4: Microarray Analysis of Stannin Knockdown in Human Umbilical Vein Endothelial Cells in TNFα Response; Implications for Cell Cycle Control…………………………………………………………………….....65 A. Introduction………………………………………………………………...... 65

B. Methods 1. Cell culture……………………………………………………………67 2. Microarray fabrication………………………………………………..67 3. Microarray cDNA probe synthesis and indirect labeling with AlexaFluor555 and 647…………………………………………67 4. Gene expression analysis……………………………………………..68 5. Cell Growth…………………………………………………………...69 6. RNA isolation/cDNA synthesis………………………………………69 7. Quantitative real-time PCR…………………………………………...70 8. siRNA construction…………………………………………………...70 9. Transfection of siRNA………………………………………………..71 10. Statistical analysis……………………………………………..….…72

C. Results 1. Stannin knockdown results in significantly altered HUVEC gene expression in response to TNFα……………………………..…73 2. Stannin knockdown significantly alters several genes involved in cell growth……………………………………………………..…..76

vii

3. Loss of stannin gene expression functionally affects HUVEC response to TNFα………………………………………………..…84 4. Knockdown of stannin inhibits the ability of TNFα-treated HUVECs to progress through the cell cycle…………………….….87 D. Discussion……………………………………………………………….…92

Chapter 5: Discussion…………………………………………………………………95 A. Stannin expression is highly regulated in a spatial and temporal manner…95 B. Stannin as a mediator of trimethyltin toxicity……………………………...96 C. Stannin as a component of cell cycle progression…………………………98 D. Overall Conclusions……………………………………………………….106

References……………………………………………………………………………..102

Appendix - Complete list of significantly altered genes from the microarray………..118

viii

LIST OF FIGURES

1. Developmental alteration in stannin expression pattern……………………………..12

2. Northern blotting analysis for the expression of 5 novel cytokine-responsive genes in resting and TNFα-activated HUVEC………………………………………16

3. Steps in TNF-R1-induced signaling…………………………………………………20

4. Signaling pathways modulated by TNF-R2…………………………………………24

5. Schematic representation of the primary structures of protein kinase c isoforms……………………………………………………………………………...27

6. TNFα is important for trimethyltin toxicity………………………………………....43

7. Trimethyltin exposure results in increased stannin gene expression…………….….45

8. Knockdown of stannin mRNA expression in HUVECs………………………….…49

9. Temporal pattern of stannin mRNA expression following treatment with TNFα….51

10. Pharmacological PKC inhibitors block TNFα-induced stannin mRNA expression……………………………………………………………………….…54

11. Protein kinase c epsilon is required for TNFα-mediated upregulation of stannin mRNA……………………………………………………………………..57

12. Jurkat T-cells respond to TNFα with a PKCε-mediated increase in stannin mRNA……………………………………………………………………………...59

13. A proposed model of trimethyltin toxicity………………………………………...64

14. Microarray experimental design…………………………………………………..75

15. A cluster-based representation of the altered genes across control, TNFα only, and combo treatment groups………………………………………………………78

16. Gene ontologies of significantly different genes (TNFα vs. combo groups)……...80

17. Differential gene expression of known cell growth effector genes………………..83

18. Differential expression of genes involved in cell growth, transcription, and cell adhesion………………………………………………………………….86

19. Alteration of HUVEC cell growth by treatment with TNFα and/or stannin siRNA…………………………………………………………………………….89

20. Co-treatment with TNFa and stannin siRNA alters HUVEC progression through the cell cycle……………………………………………………………..91

21. A model of normal stannin function…………………………………………….102

22. A model of stannin perturbation during trimethyltin exposure…………………105

LIST OF TABLES

1. Alignment and conservation of stannin protein sequences…………………………..33

List of Abbreviations

ARNT (aryl hydrocarbon receptor nuclear translocator) ATP (adenosine triphosphate) Cdc (cyclin-dependent kinase) cDNA (cloned deoxyribonucleic acid) CNS (central nervous system) DAG (diacylglycerol) DAVID (database for annotation, visualization, and integrated discovery) DD (death domain) DED (death effector domain) DISC (death inducing signaling complex) DMSO (dimethyl sulfoxide) DMT (dimethyltin) ERK (extracellular signal-related kinase) FADD (fas-associated death domain) FLIP (FLICE-like inhibitory protein) GABA (gamma amino butyric acid) GAPDH (glyseraldehyde-3-phosphate dehydrogenase) HUVEC (human umbilical vein endothelial cells) IAP (inhibitor of apoptosis) IκB (inhibitor of κB) IL (interleukin) KD (knockdown) KO (knockout) LDL (low density lipoprotein)

xii

MADD (MAP-kinase activating death domain) MAP (mitogen activated protein) MAPK (mitogen-activated protein kinase) MEM (minimum essential medium) mRNA (messenger ribonucleic acid) MZF-1 (myeloid zinc finger-1) NF-κB (nuclear factor kappa-B) NIK (NF-κB-inducing kinase) NO (nitric oxide) QRT (quantitiatie real-time) PBS (phosphate buffered saline) PC-12 (pheochromocytoma cells) PCR (polymerase chain reaction) PKC (protein kinase C) PMA (phorbol 12-myristate, 13-acetate) PND (post natal day) PVC (polyvinylchloride) RT (reverse transcriptase) RIP (receptor interacting protein) RNA (ribonucleic acid) siRNA (short interfering RNA) Snn (stannin) TMT (trimethyltin) TNFα (tumor necrosis factor-alpha) TNFR (tumor necrosis factor receptor) TRADD (TNFR-associated death domain) TRAF (TNFR-associated factor) UTR (untranslated region)

xiii

ACKNOWLEDGEMENTS

Firstly, I would like to thank my committee for guiding my project and helping me to become a better scientist. The discussions we have had over the years have helped me to view scientific problems at both the day-to-day experimental level as well as the “big picture” macro level. The critical thinking and problem solving skills will aid me throughout my career and I truly thank each of you for this.

To my co-mentors Mel Billingsley and Jong Yun. Dr. B., your research talk during orientation was al I needed to hear; I knew then that I had to join your lab for my doctoral studies. The dynamic zest for science and your scope of knowledge gave me all the confidence that you would train me as a quality scientist. During my time in your lab, you taught me independence, problem-solving, and perseverance. Your open attitude toward alternative experiences/careers has allowed me to explore opportunities that many other professors may have decided against. Without these experiences, I would not be where I am today. I cannot thank you enough for being so supportive and I appreciate all you have done for me. Jong, your willingness to become my co-mentor transformed my graduate school experience. While I looked forward to the challenges of being in an independent lab, you came in exactly when I realized I needed help. Your patience and comfortable teaching style made me feel more then your student; you made me feel like family. Thank you for taking my pre-conceptions and turning them on their ear; I will value our talks always.

To my wife, Monique you are my everything and I owe all I do to you. These first few years of marriage have been tough and I will always appreciate the sacrifices you made to get me through this. This new chapter in our lives begins now, and I am thankful I get to spend the rest of my life with you.

To my parents, your love has always made me feel comfortable and safe, even when I moved away and found myself in some fairly uncomfortable situations! The countless talks over dinner or walks through the woods are some of my most treasured memories. The smell of Mom’s cooking makes any day like Christmas and I love you both dearly.

To my other parents, thank you for taking me into your family and making me feel as if I always was. Your support of my endeavors is appreciated and I love you all.

To my friends…where to begin? Since I do not have a whole chapter to dedicate, I will say that I have been blessed to have such wonderful people in my life. No situation is too tough for me with you all by my side and my life is rich and wonderful in no small part due to you.

xiv

CHAPTER I

INTRODUCTION

A cell maintains homeostasis and reproduces only through a complex regulation of its genetic material, and associated proteins. While scientific observation has yielded a staggering array of data inferring the function and interplay of many known proteins, our knowledge of the functional proteome is incomplete. The protein Stannin (Snn) is one protein that has eluded a functional classification to date. Part of the reason for this is the high level of conservation that Snn exhibits across all known vertebrate organisms that express Snn, (www.ensembl.org; Toggas et al., 1992). The importance of a protein can be inferred from the degree of conservation over the course of evolution, thus Snn likely plays a vital role in cellular function.

Snn was originally discovered as a common factor present in cells and tissues that were sensitive to the toxicant, trimethyltin (TMT; Toggas et al., 1992). Through a combination of subtractive hybridization, in situ hybridization, and antisense knockdown experiments, Snn was shown to be necessary, but not sufficient, for TMT toxicity

(Dejeneka et al., 1997; Thompson et al., 1996; Toggas et al., 1992). Recently, the mechanism behind Snn’s mediation of TMT toxicity became more apparent as it was shown that Snn peptides may able to directly bind TMT and demethylate it to dimethyltin

(DMT; Buck et al., 2004; Buck et al., 2003). Further, a recent study showed Snn localization to the mitochondria and indicated that transient transfection of Snn into NIH-

3T3 cells, a cell line resistant to organotins, sensitized the cells to organotin toxicity

(Davidson et al., 2004). Thus, it is possible that Snn converts TMT to DMT in vivo and that this reaction results in mitochondrial dysfunction and, subsequently, cell death.

The Snn protein is an 88 amino acid, 10kDa peptide and is encoded by a 2.9kb mRNA (Dejneka et al., 1998; Toggas et al., 1992). The pattern of expression of Snn is expansive in the developing embryo, with expression becoming more focused as development continues, until adulthood, when Snn is detected in specific tissues including the spleen, hippocampus, neocortex, cerebellum, striatum, midbrain, kidney, and lung Dejneka et al., 1997). Snn is expressed in several cell types including B-cells,

T-cells, and endothelial cells (Horrevoets et al., 1999; Thompson et al., 1996). The observed pattern of Snn expression implies that a complex and tightly regulated control mechanism may exist. This hypothesis is strengthened by the identification of more than

30 potential transcription factor binding sites in the Snn gene, which may help to explain how Snn is expressed in a tissue-specific manner later in development (Dejneka et al.,

1998).

Although our understanding of the nature of Snn’s involvement in TMT toxicity has advanced, very little is known regarding the normal cellular function or the regulation of Snn gene expression. The high degree of conservation has made the generation of a high affinity antibody specific to Snn very difficult. One primary antiserum was developed, but suffered from some non-specific binding and recent attempts to make either mono- or polyclonal antibodies have not proved successful. As such, observing native Snn at the protein level has been difficult.

Recently, one study examining the endothelial cell reaction to TNFα, at the mRNA level, identified Snn as one of five novel genes induced by TNFα (Horrevoets et

al., 1999). This observation was the first link made between Snn gene expression and any signaling molecule. The work described in this thesis begins where Horrevoets et al. left off and examines the specificity and mechanism of TNFα−induced Snn gene expression, as well as how Snn may be involved in the cellular response to TNFα.

CHAPTER II

LITERATURE REVIEW

A. TRIMETHYLTIN

1. Organotins as Toxicants

Organotins are a class of compounds characterized as having at least one tin atom in the +4 state (Arakawa, 1981; Arakawa, 1993). Several classifications of organotins exist including: mono-, di-, tri-, and tetra- alkyltins and each has seen use in industrial and agricultural settings (Arakawa, 1981; Arakawa, 1993). The organotins most commonly used in industry and subsequently released to the environment are: triethyltin, tributyltin, triphenyltin, and trimethyltin (Arakawa, 1993). These compounds have been used primarily as biocides/antifouling agents and PVC stabilizers and catalysts (Fent,

1996). In the past, annual consumption of organotins was very high; in 1985, global consumption of organotins was estimated to be 35,000 tons (Fent, 1996). However, most organotins are highly toxic and the use of these compounds has recently been banned internationally as an antifouling agent on seafaring vessels (Champ, 2000). Data regarding the use of these compounds has not been reported consistently, so it has been difficult to track any decline in usage that may have occurred. Enforcement of these laws has been difficult, since ships can be treated in non-abiding nations and then traverse international and regulated national waters, continuing the problem. In addition, the added expense and difficulty of finding/using alternatives to organotin anti-fouling agents has caused a shift of use to poorer, less regulated nations, whom are least equipped to deal with organotin buildup. Organotin pollution is now a recognized problem and

learning how these compounds elicit their toxic effects is important if we are to better deal with cases of human exposure and to develop policy to effectively dispose of these chemicals.

2. Trimethyltin – A Potent, Selective Neurotoxicant

Trimethyltin (TMT) is a particularly potent organotin formerly used for a variety of industrial and agricultural purposes. The first reported case of TMT poisoning, involving 2 chemists, was in 1978 by Fortemps and colleagues (Fortemps et al., 1978).

The observations by Fortemps, along with those of others whom have observed victims of TMT poisoning, resulted in the defining of a collection of symptoms and pathology commonly observed after TMT poisoning as “TMT Syndrome” (Nishimura et al., 2001).

This “TMT Syndrome” consists of most or all of the following: hearing loss, disorientation, amnesia, aggressive behavior, hyperphagia, disturbed sexual behavior, complex partial and tonic-clonic seizures, nystagmus, ataxia, and mild sensory neuropathy (Besser et al., 1987). These symptoms are caused primarily by a dysregulation/degradation of limbic/cerebellar regions of the brain and lead to an interesting characteristic of TMT poisoning, that of TMT’s specificity.

Areas of the brain known to be affected by TMT include: the hippocampus, neocortex, basal ganglia, cerebellum, brain stem, spinal cord, dorsal root ganglia, olfactory cortex, retina, and the inner ear (Chang, 1990). This pattern of toxicity is dissimilar to other organotins. Other organotins, such as tributyltin and triphenyltin, can enter the brain to a greater degree (Mushak et al., 1982) suggesting that TMT toxicity is not simply a matter of CNS distribution. It was also shown that TMT’s specific pattern

of damage is not due to preferential distribution or uptake of TMT to sensitive areas

(Cook et al., 1984). Those observations led to the theory of an underlying factor, inherent in sensitive cell types, mediating TMT sensitivity (Toggas et al., 1992).

3. Trimethyltin Toxicity – Cellular Mechanisms

It is accepted that TMT causes cell death, though the mechanisms remain unclear.

Some observers have postulated that the dose of TMT affects the method of cell death

(with low doses leading to apoptotic death, and high doses leading to necrotic death).

Among the literature pertaining to TMT, two major theories have arisen: that TMT kills cells due to excitotoxicity (Patel et al., 1990) or that TMT causes a buildup of reactive oxygen species, which, in turn, result in cell death (Ali et al., 1992). The evidence in favor of these hypotheses is briefly discussed below.

Excitotoxicity

The excitotoxic model of TMT toxicity was based on the finding that TMT administration results in seizure and increased glutamate release in murine models of neurotoxicity (Brodie et al., 1990; Feldmen et al., 1993). A second observation in support of an excitotoxic model of TMT damage was the observation that calpain- induced cleavage of caspase-3, a known signaling component of glutamate mediated cytotoxic cascades, was increased in TMT-exposed hippocampal slices (Munirathinam et al., 2002). A downregulation of inhibitory GABA-ergic pathways can also contribute to excitotoxicity through a decreased ability to cope with excitatory glutamatergic signals.

Observers noted that TMT administration resulted in a decrease in several subunits of the

GABAA and GABAB receptors, supporting the notion that TMT may cause excitotoxic

cell death by both increasing glutamate activity while decreasing compensatory GABA

activity (Nishimura et al., 2001). Though several lines of evidence point to an excitotxic

mechanism of TMT cytotoxicity, additional studies show that other factors, such as

oxidative stress may also be involved.

Oxidative Stress

It has been shown that TMT causes an upregulation of nitric oxide activity,

formation of oxidative radicals, and hydrogen peroxide generation (Gunasekar et al.,

2001). Through the use of chelerythrine, a potent PKC inhibitor, PKC was found to be

essential in the formation of reactive oxygen species in TMT-exposed cerebellar granule

cells (Gunasekar et al., 2001). Further, lipid peroxidation, a hallmark of damage caused

by oxidative radicals, was found on TMT-exposed cells within 24 hours of exposure. In

order to assess the importance of oxidative damage in causing TMT toxicity, antioxidants

such as L-NAME, catalase, and MCPG were used to protect cells. Blocking NO

formation and oxidative radical formation (via L-NAME and catalase) partially attenuated TMT-induced cell death (Gunasekar et al., 2001). Thus, formation of reactive oxygen species and NO plays an important role in TMT toxicity.

Experimental evidence suggests a multi-factor mechanism underlying TMT toxicity. Both excitotoxicity and oxidative stress play a role in mediating cell death after

TMT exposure. Regardless of the underlying cellular mechanisms, TMT is capable of causing profound behavioral change in exposed mammals.

4. Trimethyltin Toxicity – Behavioral Effects

Trimethyltin causes severe disruption of several cognitive processes. Among these are: response to visual stimuli (Dyer et al., 1982), response to painful stimuli

(Walsh et al., 1984), and impairment of the development of rat pups when TMT was titrated into a female rat’s drinking water (Noland et al., 1982). In addition to these impairments, TMT altered learning and behavior; studies of reinforced learning showed that a single-dose of TMT was sufficient to decrease the ability of a rat to respond to cues in order to receive a food reward (Wenger et al., 1985). Memory deficits became obvious when rats were placed into a radial arm maze after single dose TMT intoxication (Walsh et al., 1982). Rats exposed to TMT needed between 50-100% more arm entries to complete the task vs. their unexposed counterparts. In addition, this impairment lasted for over 2 months with only minor improvement (Walsh et al., 1982). Studies employing a Hebb-Williams maze showed that TMT treated rats not only showed an increase in errors over control rats, but also were not nearly as adept at reducing those errors when exposed to the same maze configuration repeatedly (Swartzwelder et al., 1982).

5. Trimethyltin - Summary

It is obvious that TMT is a dangerous and selective toxicant and that human TMT exposure, though rare, is severely debilitating. As a result, several groups have studied the effects of TMT on mammals and attempted to discover the mechanism underlying the selective toxicity of TMT. However, correlations between the amount of tin accumulating in brain as a result of TMT exposure, the distribution of that tin, and the chemical structure of TMT provide no insight as to why TMT is so selective in the tissues

it destroys. Given these observations, the hypothesis of a common factor being expressed in TMT-sensitive cells and tissues arose. That hypothesis proved correct and resulted in the discovery and characterization of the protein stannin.

B. STANNIN

1. The Discovery of Stannin and Characterization of Expression

Stannin (Snn) was discovered initially as a gene product present in TMT-sensitive tissues (Toggas et al, 1992). Using avidin-biotin subtractive hybridization on control and

TMT-treated rat brain a 2912 bp clone was isolated and characterized. The Snn cDNA was found to code for an 88 amino acid product and was found to contain a putative origin of replication from bases 148 to 411. In addition, analysis of the cDNA via a

Hopp-Woods algorithm indicated the presence of a 25 amino acid hydrophobic domain in the amino terminus. The expression of Snn mRNA was examined in rat tissues using northern blot and in situ hybridization and was localized to (in descending order of detected expression level): spleen, hippocampus, neocortex, cerebellum, striatum, midbrain, kidney, and lung (Dejneka et al., 1997). Tissues found to express low levels of

Snn in rat included: liver, heart, skeletal muscle, and testis.

The expression of Snn during development in rat was examined; Snn is selectively expressed in adult rat brain. The pattern of Snn expression changes from widespread expression in the embryonic brain, to a more focused pattern of expression in the adult brain (Dejneka et al., 1997). The tissue specificity and developmental distribution of Snn suggest a complex level of regulatory controls modulating Snn expression in vivo.

2. Characterization of the Stannin Gene

Since the Snn protein is highly conserved and expressed in a tissue and developmentally-regulated manner, a complex system of controls must exist to modulate

Snn expression at the levels of transcription and translation. To better understand possible transcriptional regulation of Snn, a comprehensive analysis of the Snn gene and genomic clone was performed (Dejneka et al., 1998). This analysis indicated the presence of two exons in the Snn gene.

The first exon spans bases 1-61 of the 2.8kb mRNA and the second exon is separated from exon 1 by a 5.8kb intron. A search of the region 5’ to exon 1 revealed a putative promoter start site which contained the CCAAT and GGGCGG consensus sequences at –92 and –85 (relative to the first base of exon 1; Dejneka et al., 1998). In addition, this region has six potential transcription factor binding sites, one each, for:

MZF-1, CP2, AP-4, ARNT, N-Myc, and USF. Analysis of the intron indicated the presence of three additional promoter start sites at bases 5501, 6377, and 6485. Unlike the 5’ region, the intron contained a TATA box at position –29 (relative to the first base of exon 2). Additional features of the intron include: a CCAAT consensus sequence at

Figure 1: - Developmental Alteration in Stannin Expression Pattern (Figure reprinted from Dejneka et al., 1997). In situ hybridization of stannin ni developing rat brain. Coronal hemisections of brain were hybridized with an antisense stannin 35S

cDNA probe. Autoradiographs of sections through the medial hippocampus and piriform

cortex are shown for animals postnatal day (PND) 1, 5, 12, and 20. A diffuse

hybridization signal was seen throughout the brain on PND1 and 5. This signal

converges to specific brain regions by PND12 and 20.

–74, a metallothionein metal-responsive element at –337, and 26 transcription factor binding sites (Dejneka et al., 1998). The transcription factor binding sites include those specific to: Nkx-2.5, δEF1, RORα2, GATA-X, four MZF-1 sites, six AP-4 sites, IK-2, and MyoD. Chromosomal mapping located the Snn gene on the A2 band of chromosome 16 in rat (Chr 16p13 in humans).

The diverse array of potential transcription factor binding sites on the Snn gene may help to explain Snn’s tissue-specific expression patterns. The presence of MZF-1,

IK-2, GATA, δEF1, and CP2 sites provide a theoretical mechanism underlying expression of Snn in spleen and brain. It is important to note that the functionality of these transcription factor-binding sites have not been shown experimentally.

3. Stannin’s Role in Trimethyltin Toxicity

Trimethyltin (TMT) is an organotin with potent neuro-, nephro-, and lympho- toxicity (Liu et al., 2004; Robertson et al., 1987; Hioe et al., 1984). Snn was originally characterized during studies defining common factors that are present in TMT-sensitive cells (Toggas et al., 1992) and the expression pattern of Snn correlates very closely with

TMT-sensitive tissues in vivo (Dejneka et al., 1997). Further, Snn expression decreased dramatically in rat hippocampal pyramidal and dentate gyrus neurons following a single

8mg/kg dose of TMT (Patanow et al., 1997). To determine if Snn was a necessary component of TMT toxicity, antisense oligonucleotides directed against Snn were used.

It was found that 0.5 µM of Snn-specific antisense oligonucleotides was sufficient to provide statistically significant protection against a single TC50 dose of TMT in primary neurons (Viviani et al., 1998; Thompson et al., 1996). This result suggests that Snn is a

necessary component of TMT toxicity in primary neurons. However, expression of Snn

alone in TMT-insensitive cells was not sufficient to cause sensitivity to TMT.

Several mechanistic studies have looked at how Snn may be mediating TMT

toxicity in vitro. Previous work had shown that vicinal thiols constitute a general binding target for organotins (Stridh et al., 2001) and the Snn protein contains such a moiety at the proposed membrane interface of the protein (Davidson et al., 2004; Toggas et al.,

1992). With this in mind, Buck et al. 2004 created a synthetic peptide containing the vicinal thiols corresponding to amino acids 29-37 of the Snn protein. This peptide was then used in binding studies and it was found that the Snn peptide bound to, and dealkylated, TMT to dimethyltin (DMT, Buck et al., 2001; Buck et al., 2004). Snn is localized to the mitochondria; Snn transfection conferred sensitivity to both TMT and

DMT in PC12 cells (Davidson et al., 2004). Thus, Snn may mediate TMT toxicity through direct binding and dealkylation of TMT to DMT which, in turn, causes mitochondrial dysfunction and cell death.

4. Stannin in Cellular Signaling

The normal function of Snn in cells is not currently known. One study presented a clue, showing an increase of Snn expression by tumor necrosis factor α (TNFα,

Horrevoets et al., 1999). This work characterized changes in the gene expression patterns of human umbilical vein endothelial cells (HUVECs) after stimulation with

TNFα. After a single dose of 200ng/ml TNFα, HUVECs responded with an increase of

Snn mRNA expression of approximately 10-fold within 3 hours of exposure. Such a

FIGURE 2 - Northern blotting analysis for the expression of 5 novel cytokine- responsive genes in resting and TNF-α-activated HUVEC (reprinted from Horrevoets et al, 1999). Northern blotting analysis of 10 ug HUVEC total RNA with probes from

DD-fragments or corresponding EST clones. Time periods of continuous stimulation by

TNF-a are indicated, and GAPDH analysis is given as control for equal loading.

Key:

CA2_1 = hIAP1 (inhibitor of apoptosis 1)

GG10_2 = Rabkinesin-6

GG2_1 = Novel gene

CG12_1 = APOL-like gene

AG8_1 = Stannin

GAPDH = GAPDH

potent induction implies that a direct signaling relationship may exist between TNFα and

Snn.

C. TUMOR NECROSIS FACTOR α

1. Tumor Necrosis Factor – Origin and Plieotropism

Tumor necrosis factor alpha (TNFα) was first discovered in 1984 as a factor

capable of regressing bacteria-infected tumor masses (Old, 1985). The primary source of

TNFα is activated macrophages in vivo, and TNFα is typically expressed as a 26-kDa

transmembrane protein that can be cleaved to release a 17-kDa soluble form (Idriss and

Naismith, 2000). Though macrophages are the primary source of TNFα, several other cell types are capable of producing it as well including: lymphocytes, keratinocytes, firoblasts and endothelial cells (Tracey and Cerami, 1993). As a signaling molecule,

TNFα is unparalleled in the breadth and power of biological effect it can have, depending upon the presence of cofactors, nature of stimulus, and state of the cell or tissue at the time of stimulus. These effects include, but are not limited to, lymphocyte and leukocyte activation and migration, fever, inflammation, cell proliferation, differentiation, and apoptosis (Tracey and Cerami, 1993). Interestingly, the degree of pleiotropism of TNFα is mediated through only two known TNFα receptors, TNFR1 (p55) and TNFR2 (p75;

Grell et al., 1994).

2. Tumor Necrosis Factor Receptor 1: Signal Transduction and Major Functions

The TNFR1 receptor is best known for its ability to induce cell death via the death inducing signaling complex (DISC). The assembly of the DISC and initiation of

apoptosis is a multi-step process. First, binding of TNFα to TNFR1 results in trimerization of the receptor (Chan et al., 2000). TNF-R-associated death domain, or

TRADD, is then recruited to the death domain (DD) on TNFR1 and TRADD then recruits fas-associated death domain (FADD, Stanger et al., 1995). FADD then recruits procaspase-8 via homologous death effector domain (DED) to form the completed DISC

(Locksley et al., 2001). During the assembly of the complete DISC, procaspase 8 is autolytically cleaved to yield active caspase 8, which is then released from the DISC to the cytoplasm and serves as an activator for the effector caspases 3, 6, and 7 (Gupta S.,

2001). These effector caspases then cleave a number of protein substrates and drive the cell to apoptosis (Locksley et al., 2001). Whether a classic caspase-mediated apoptosis or a mitochondria-mediated form of apoptosis results is cell-type specific (Gupta S., 2001).

Although TNFR1 signaling is classically thought to result in cell death, in certain cell types and in response to specific stimuli, TNFR1 signaling can also result in survival or even activation of cells (MacEwan DJ, 2001). The major pathways through which

TNFα can promote cell survival and cellular activation are through TNF-R-associated factor-2 (TRAF-2) and receptor interacting protein (RIP; MacEwan DJ, 2001). TRAF-2 is known to stimulate the activation of MAP kinase (Natoli et al., 1997). MAP kinase has been shown to be both proapoptotic (Ichijo et al., 1997) and antiapoptotic (Natoli et al.,

1997) with the specific outcome determined by additional cellular and environmental factors. The primary survival pathway mediated by TNFR1 involves binding of MAP- kinase activating death domain protein (MADD) directly to TNFR1 (Schievella et al.,

Figure 3: Steps in TNF-R1-Induced Signaling (reprinted from Gupta 2002). An outline of pro-and anti-apoptotic pathways modulated through TNF-R1.

1997). MADD stimulation then results in extracellular signal-regulated kinase (ERK) activation, a known survival signal (Schievella et al., 1997).

RIP is known to result in nuclear factor kappa B (NF-κB) activation (Kreuz et al.,

2004). NF-κB is a known inhibitor of apoptosis and is capable of inducing transcription of many targets (Liu et al., 1996; Beg and Baltimore, 1996). TNFα regulates NF-κB through TRAF2-mediated NF-κB-inducing kinase (NIK) activation (MacEwan DJ,

2002). NIK activates NF-κB by phosphorylating inhibitor of κB kinase (IKK), resulting in the dissociation of IκB from NF-κB due to ubiquitination and subsequent degradation, thereby releasing active NF-κB (Ling et al., 1998). NF-kB, in turn, activates proteins such as inhibitor of apoptosis 2 (IAP2) and FLICE-like inhibitory protein (FLIP), which act to promote cell survival (Karin and Lin, 2002).

3. Tumor Necrosis Factor Receptor 2: Signal Transduction and Known Functions

There is much less known about the TNFR2 receptor compared to TNFR1.

TNFR2 differs structurally from TNFR1 in that TNFR2 has no death domain. It was discovered that the membrane-bound form of TNFα is superior to the soluble form in terms of ability to activate TNFR2 (Grell et al., 1995). This fact may be part of the reason that TNFR2 is more mysterious, since most laboratory experiments employ the soluble form of TNFα. Signaling through TNFR2 typically involves the binding of

TRAF1 and TRAF2 to the intracellular domain of TNFR2 (Rothe et al., 1994). This complex then recruits the proteins inhibitor of apoptosis –1 and –2 (IAP1, IAP2), which act to promote cell survival (Deveraux et al., 1999; Rothe et al., 1995). Similar to

TNFR1, the TNFR2 receptor is drastically affected by RIP (Pimentel-Muinos and Seed,

1999). In fact, it has been shown that the presence or absence of RIP can result in

apoptosis or the activation of NF-κB, respectively (Pimentel-Muinos and Seed, 1999).

Another interesting function attributed to TNFR2 is that TNFR2 may act as a high

affinity trap for TNFα and deliver it to the TNFR1 receptor, serving to enhance TNFR1-

mediated apoptosis (Tartaglia et al., 1993).

Clearly, TNFR2 is involved in many cellular responses; the most clearly defined

functions, that of NF-κB activation and promotion of cell survival, are not exclusive to

TNFR2 and neither are the signaling pathways involved in these processes. In fact, the only currently known signaling pathway that is exclusive to TNFR2 and not TNFR1, is that of phosphorylation of the TNFR2 receptor by casein kinase and p80TRAK (Darnay et al., 1997; Beyaert et al., 1995). Both of these stimuli bind to a 44 amino acid site in the cytoplasmic domain of TNFR2 and the direct results of these couplings are currently unknown. The results gained through the use of TNFR2 knockout mice clearly implicate

TNFR2 as being important in multi organ inflammation (Douni and Kollias, 1998), thymocyte proliferation (Grell et al., 1998), and microvascular endothelial cell damage

(Lucas et al., 1998); however, much more work needs to be done to separate the specific

effects of TNFR2 from those of TNFR1.

D. PROTEIN KINASE C

1. Classes of Protein Kinase C & Activation

Three main classes of protein kinase C (PKC) exist: the classical, the novel, and

the atypical. These classes are defined according to the structure and cofactor

Figure 4 – Signaling Pathways Modulated by TNF-R2 (reprinted from MacEwan

2002).

requirement for activation of each isotype (Soh and Weinstein, 2003). The classical PKC

isotypes can be activated by diacylglycerol (DAG), calcium, or phospholipids and

include the α, βI, βII, and γ isoforms (Spitaler and Cantrell, 2004). The novel PKCs are activated by DAG or phospholipids but are insensitive to calcium and include the δ, ε, η,

θ, and µ isoforms. The atypical isoforms of PKC are not responsive to either DAG or calcium and are known to include the ζ and ι isoforms. Each of the above isoforms has

an N-terminal regulatory domain and a C-terminal catalytic domain. These domains are

in turn linked through the “third variable region” (V3). The regulatory domain is thought

to act in autoinhibition of the enzyme, and the presence of DAG, calcium, or another

activator is thought to relieve this inhibition and activate the enzyme (Soh and Weinstein,

2003). Additionally, cleavage of the V3 region, thereby releasing the catalytic domain

from the regulatory domain, can result in activation of some PKC isoforms (Emoto et al.,

1996).

2. Functions of Protein Kinase C

Protein kinase C has been implicated in functions as diverse as cell proliferation

and regulation of cell-cell contacts, to regulating insulin secretion in response to glucose

(Poole et al., 2004). These diverse functions are of great importance to the cell and are

served by one or more of the 12 known isoforms of PKC. PKC δ, ε, ζ were examined in

this body of work and a brief description of known/suspected functional roles of each is

presented below.

Figure 5: Schematic representation of the primary structures of protein kinase C isoforms (reprinted from Aksoy et al., 2004). The PKC structure can be divided into an

N-terminal (N) regulatory domain and C-terminal (C) catalytic domain. Indicated are the substrate domain, the C1 domain which binds diacylglycerol and phorbol esters, the C2 domain which binds to acidic lipids and for classic PKCs, Ca2+ binding, and the C3 and

C4 domains that comprise the ATP- and substrate-binding regions fo the kinase.

Members of each isoform subfamily are listed on the left.

PKC δ

The δ isoform of PKC is a ubiquitously expressed isoform that is developmentally regulated in tissues such as the brain and epidermis (Gschwendt, 1999). PKCδ has been primarily implicated in regulation of cell growth and survival (Kilpatrick et al., 2002;

Zemskov et al., 2003). In particular, PKCδ has been shown to be important in the regulation of apoptosis and at least three experimental observations support this role.

First, a caspase 3 cleavage site was found in the V3 domain of PKCδ and cleavage at this spot resulted in a catalytically active fragment of PKCδ (Ghayer et al., 1996). The importance of this fragment was confirmed via overexpression studies in which it was found that overexpression of the cleaved catalytic fragment of PKCδ was capable of inducing apoptosis while overexpressing wild type PKCδ or a kinase dead catalytic domain mutant were not (Ghayer et al., 1996). Second, it has been shown that PKCδ activity increases scramblase activity in response to known apoptotic stimuli such as FAS ligand (Frasch et al., 2000). Scramblase, in turn, is known to be the mediator of phosphatidylserine translocation to the outer leaflet of the plasma membrane during apoptosis. Third, the mitochondria has been identified as a major target of PKCδ and

PKCδ has been shown to translocate to the mitochondria in response to diverse apoptotic stimuli such as Phorbol 12-myristate, 13-acetate (PMA) and oxidative stress (Majumder, et al., 2001; Li et al., 1999). Further, it was found that overexpression and activation of

PKCδ resulted in a decrease in the mitochondrial membrane potential and cytochrome c release (Majumder, et al., 2001; Li et al., 1999). A summation of the above evidence strongly implicates PKCδ as being important in cellular apoptosis.

PKCε

PKC epsilon (PKCε) has been implicated in cardioprotection against apoptosis, inflammation, and regulation of cellular proliferation (McJilton et al., 2003; Baines et al.,

2002; Petrovics et al., 2001). The majority of knowledge regarding PKCε’s function has

been derived from examining PKCε’s role in cardioprotection against ischemic damage.

Recently, studies have shown that PKCε‘s antiapoptotic effects occur in multiple cell

lines, not exclusively cardiac tissue (Wu et al., 2004; Rao et al, 2004). The mechanism

underlying PKCε’s anti-apoptotic/pro-proliferative effects are thought to involve the c-

Raf/MEK/ERK signaling pathway, specifically through PKCε’s activation of ERK1/2

(Rao et al., 2004). PKCε’s role in inflammation was first observed upon the generation of the first PKCε knockout (KO) mice (Castrillo et al., 2001). These mice had severe problems clearing both gram positive and gram negative bacterial infections. Further examination revealed that PKCε KO mice have an attenuated response to lipopolysaccharide, which did not result in the classic large-scale release of cytokines in vitro (Castrillo et al., 2001). Moreover, use of various PKC inhibitors showed that this effect is due to a non-redundant function of PKCε (Castrillo et al., 2001). The mechanism underlying these defects appears to be a vital role for PKCε in transmitting the LPS stimulus into NF-κB and MAPK activation (Aksoy et al., 2002). The mechanisms underlying the regulation of PKCε have not been examined to date.

PKCζ

The atypical PKC zeta (PKCζ) isoform, though not activated by diacylglycerol,

can be activated by lipids such as phosphatidylinositol, phosphatidic acid, arachidonic

acid, and ceramide (Nakanishi et al., 1993; Limatola et al., 1994; Muller et al., 1995).

Most of the experimental work done on PKCζ activation focused on the role of PI-3,4,5- triphosphate (PIP3) in activating PKCζ. The currently accepted mechanism of PKCζ activation is through PIP3 binding to and activating PDK1 which, in turn, binds to and activates PKCζ (Chou et al., 1998). The activation of PKCζ has been implicated in several diverse functions. The three major functions attributed to, or thought to involve,

PKCζ are; modulation of the mitogen activated protein kinase (MAPK) cascade (Berra et al., 1995; Monick et al., 2000), modulation of NF-κB activation through association with

IKK signaling (Lallena et al., 1999; Leitges et al., 2001), and control of cell polarity via maintenance of tight junction integrity (Lin et al., 2000; Suzuki et al., 2001).

3. PKC – Summary

Cellular homeostasis influenced by PKC. Interaction with many signaling pathways including TNFα and TMT-mediated pathways, provides the possibility of an interaction of one or more PKC isoforms with Snn, whether direct or indirect. Through the use of in silico modeling, Snn was found to have multiple potential phosphorylation sites. Given the known interactions between TMT, TNFα, and PKC we hypothesized that PKC may be mediating the observed increase in Snn mRNA observed after

TNFα stimulation.

Chapter III

Protein Kinase C Epsilon Regulates TNFα-Induced Stannin Gene

Expression

A. Introduction

Organotins are a class of chemicals used in multiple commercial applications including curing silicon rubber, forming certain industrial plastics, and as a biocide (Gunasekar et al., 2001; Toggas et al., 1992). Though organotins serve several useful functions, they are problematic to utilize since they are also potent neurotoxicants.

This toxicity resulted in the 1995 banning of organotin use in most commercial applications worldwide (Champ, 2000). The organotin trimethyltin (TMT), causes a highly specific, reproducible pattern of selective damage to human and vertebrate nervous systems (Balaban et al., 1988). Exposure to TMT produces a consistent constellation of symptoms including disorientation, increased aggression, and seizures

(Gunasekar et al., 2001). In addition, TMT exposure causes cellular degeneration in lymphocytic tissues, kidney, and spleen (Philbert el al., 2000; Balaban et al., 1988). The selective toxicity of TMT suggests a common factor underlying organotin sensitivity.

Subtractive hybridization studies indicated that this common factor was the protein Stannin (Snn); Snn was detected in a range of TMT sensitive tissues (Krady et al.,

1990; Dejneka et al., 1997). Previous studies have shown that Snn is necessary, but not sufficient, for TMT toxicity in vitro (Thompson et al., 1996). Snn is an 88 amino acid protein that is highly conserved throughout vertebrate evolution (Table 1; Dejneka et al.,

1998). Rat and mouse Snn amino acid sequences are 100% identical, and human Snn

Table 1 – Alignment and Conservation of Stannin Protein Sequences

Protein conservation of Stannin across vertebrates –

A) Alignment from the predicted and verified open reading frames of the

stannin peptide from multiple species. Residues which are not identical

are shown in bold.

B) The percentage identity of six vertebrate species to human stannin. Note

rat and human stannin are identical to each other.

differs by only two amino acids at the C-terminus. Further, mouse and human Snn nucleotide sequences are 90% identical (Dejneka et al., 1998). Such a highly conserved nature implies an important role for Snn in normal cellular function.

Tumor necrosis factor-alpha (TNFα) is a pleiotropic cytokine known to have diverse cellular actions including mediation of the inflammatory and immune responses

(Yoshizumi et al., 2004; Leeuwenberg et al, 1995). TNFα has also been linked to TMT toxicity in neuronal and mixed glial/neuronal cultures (Harry et al., 2003; Viviani et al.,

2003). Specifically, these groups showed that TNFα is upregulated after TMT administration in mixed glial/neuronal cultures. Further, Harry et al. observed some protection against TMT-induced cell death when cells were pre-treated with a neutralizing TNFα antibody (Harry et al., 2003). Using differential gene display,

Horrevoets, et al., showed that TNFα treatment of human umbilical vein endothelial cells

(HUVEC) induced several gene products including Snn, indicating a possible regulatory role of TNFα in the mediation of Snn expression in HUVECs (Horrevoets et al., 1999).

In addition, studies showed that both TNFα and Snn were present in similar tissues, including the nervous and immune systems of embryonic mice (Dejneka et al., 1997; Pan et al., 1997; Yeh et al. 1998). This pattern of expression raised the possibility that the two proteins are co-expressed in specific cell types and tissues, and that Snn may be a downstream component of a TNFα-mediated cell-signaling pathway.

In this study, we examined the signaling events involved in TMT toxicity. We demonstrate that endothelial cells are vulnerable to TMT damage and that this damage is mediated, in part, by TNFα and requires Snn. Further, we showed that TMT increases

Snn gene expression prior to inducing cell death. We also demonstrated that PKCε was a

critical modulator of TNFα-induced Snn mRNA expression in both HUVECs and Jurkat

T-cells. These data indicate that stannin can be induced by TNFα, and may further enhance cell sensitivity to TMT via PKC-mediated induction of the stannin gene product.

B. Methods

Cell Culture

Human umbilical vein endothelial cells (HUVEC) were obtained from Cambrex

(East Rutherford, NJ). HUVECs were maintained as recommended by Cambrex.

Briefly, cells were cultured in the chemically-defined EGM™ media (Cambrex, East

Rutherford, NJ) containing 2% fetal bovine serum (FBS). For all experiments, HUVECs were passage 3-5. In addition, HUVECs were allowed 24 hours of undisturbed growth prior to any experimental manipulation after plating. Jurkat T-cells were maintained in

RPMI 1640 from Mediatech (Herndon, VA) supplemented with 10% FBS.

Reagents

Tumor necrosis factor-α (TNFα) and interleukin-1β (IL-1β; Roche, Indianapolis,

IN) were dissolved in sterile PBS and administered at concentrations of 200 ng/ml and 10 ng/ml, respectively. The TNFα neutralizing antibody (BD PharMingen, San Diego, CA) was used at 3 µg/ml, a concentration expected to neutralize 60-70% of TNFα in culture

(per manufacturers literature). The α-synuclein antibody was also used at 3 µg/ml (Santa

Cruz, Santa Cruz, CA). Trimethyltin was a gift from James O’Callahan (NIOSH,

Morgantown, WV). Bisindolylmaleimide I (Calbiochem, San Diego, CA) was used at a

final concentration of 10 nM and was dissolved in dimethyl sulfoxide (DMSO). Gö6976

(Calbiochem, San Diego, CA) was used at a concentration of 8 nM and was dissolved in

DMSO. Phorbol 12-myristate, 13-acetate (PMA; Roche, Indianapolis, IN) was used at a final concentration of 100 nM and was dissolved in DMSO.

Cell Viability

The trypan blue exclusion test was used as a measure of cell viability. HUVECs were incubated in 0.2% trypan blue (Sigma, St. Louis, MO) diluted in phosphate-buffered saline (PBS) and then subsequently washed once with PBS. The number of normal and blue-stained, dead cells were counted in four independent microscopic fields per culture, with three independent cultures being used for each condition (12 total fields). The percentage of viable cells was compared in each treatment condition with that of the untreated control condition, which was considered 100% for the purposes of relative viability.

RNA Isolation/cDNA synthesis

RNA isolation was accomplished using the RNeasy kit, according to the protocol recommended by the manufacturer (Qiagen, Valencia, CA). Briefly, HUVECs were harvested using 0.05% trypsin-EDTA and pelleted at 2000 x g. Cells were then resuspended in 350 µl of RLT cell lysis buffer and homogenized using QIAshredder homogenization columns (Qiagen, Valencia, CA). The homogenized mixture was combined with an equal volume of 70% ethanol and added to an Rneasy RNA isolation column and spun at 8000 x g for 15 sec. The RNA bound on the column was washed

three times, and finally eluted with Rnase free water as indicated by the manufacturer.

The synthesis of cDNA was carried out using the First Strand cDNA Synthesis Kit (MBI

Fermentas, Hanover, MD). This kit employs a standard M-MLV reverse transcriptase reaction and was used according to the recommendations of the manufacturer.

Quantitative Real-Time PCR

The cDNA template from HUVECs were normalized based on their relative expression of β-Actin. To detect human Snn, the following primers and probe were used to amplify a 100 bp product corresponding to bases 222-322 of the mRNA:

Forward Primer: 5’ – TTG TCA TCC TCA TTG CCA TC – 3’

Reverse Primer: 5’ – GCT CTC CTC GTC CTC TGA CT – 3’

Probe: 5’ – CCT GGG CTG CTG GTG CTA CCT – 3’

β-Actin was detected using a pre-developed 20X primer-probe assay kit (Applied

Biosystems, Foster City, CA).

Reactions were carried out using a protocol from Qiagen (Valencia, CA). The

PCR program was as follows: step 1- 95oC for 15 minutes, step 2 – 95oC for 15 seconds,

step 3 – 60oC for 1 minute, with steps 2 and 3 repeated for 40 cycles. All reactions were carried out using the ABI Prism 7700 Lightcycler.

siRNA Construction

All siRNA except the siRNA specific to PKCζ was constructed using the

Silencer siRNA Construction Kit (Ambion, Austin, TX). The following

oligonucleotides were utilized to construct siRNA (only the sense strand is given, shown without T7 adapter sequence):

Control siRNA: 5’ – AAA GGC ACT TAG GAC CCA GGG – 3’

Snn siRNA 1: 5’ – AAG GAA CCC TTC CTG CTG GTG – 3’

Snn siRNA 2: 5’ – AAG GGA CCG TGC GTG GAG AGA – 3’

PKCε siRNA 1: 5’ – GCC CCT AAA GAC AAT GAA GTT – 3’

PKCε siRNA 2: 5’ – CTT CAT TGT CTT TAG GGG CTT – 3’

PKCδ siRNA 1: 5’- GAT GAA GGA GGC GCT CAG TT – 3’

PKCδ siRNA 2: 5’ – GGC TGA GTT CTG GCT GGA CTT – 3’

The procedure for contructing the Snn siRNA was as outlined by Ambion. In brief, sense and antisense DNA oligonucleotides, each containing an 8 nucleotide sequence complementary to the T7 promoter, were separately hybridized to a T7 promoter and made double-stranded with Exo-Klenow DNA polymerase. Each reaction was mixed with a T7 RNA polymerase in order to generate the siRNA templates. Both the sense and antisense reactions were combined and incubated to form dsRNA. Finally, each double stranded siRNA was purified and eluted into nuclease-free water. The sequences for the PKCε siRNA were obtained from Irie et al., 2002 and the sequences for the PKCδ specific siRNA were obtained from Yoshida et al., 2003. PKCζ was knocked

down using SMARTpool siRNA from Upstate Cell Signaling Solutions (Charlottesville,

VA).

Transfection of siRNA

All siRNA used in these studies was transfected into HUVECs using the siPORT

Lipid reagent (Ambion, Austin, TX). Briefly, siPORT Lipid reagent was diluted in Opti-

MEM I (Gibco, Carlsbad, CA) and allowed to incubate at room temperature for 20 minutes. Each siRNA was separately diluted in Opti-MEM I and allowed to incubate at room temperature for 5 minutes. The mixtures containing the siPORT Lipid and siRNA were then combined and allowed to incubate at room temperature for 15 minutes.

HUVECs were washed in Opti-MEM I, then fresh Opti-MEM I was added to the cells in place of EGM media. The siPORT Lipid/siRNA mixture was then overlaid dropwise onto the cells. In all figures, except Figure 8A, a 50:50 mix of two Snn specific siRNA were used.

Statistical Analysis

For all experiments, significance was determined using a Student’s T-test between treatment conditions and the control or vehicle condition, unless otherwise specified. A p-value of less than 0.05 was considered significant in all cases. Each experiment consisted of at least 3 replicates per condition.

C. Results

Tumor Necrosis Factor-α Contributes to Trimethyltin Toxicity. Previous work has implicated TNFα as a contributing factor in TMT toxicity in neurons and glia (Harry et al., 2003; Viviani et al., 2003). To determine if HUVECs were vulnerable to TMT, and whether TNFα played a role in such toxicity, HUVECs were exposed to 10 µM TMT

± 3 µg/ml of TNFα neutralizing antibodies or an α-synuclein antibody, as a control, for

48 hours. As shown in Figure 6, exposure of HUVECs to 10 µM TMT resulted in 60%

(± 4.01) cell death. In contrast, when HUVECs were pretreated with a neutralizing anti-

TNFα antibody, a significant decrease (20%) in cell death was observed, implicating

TNFα as a factor contributing to TMT toxicity in HUVECs. Although the magnitude of cell death was less (~20%), a similar trend was observed when HUVECs were exposed to

5 µM TMT (data not shown). This trend has also been observed in other studies involving mixed neuronal/glial co-cultures (Harrey et al, 2002). Together, these results indicate that TNFα may be an important factor in TMT-induced cytotoxicity.

Trimethyltin Induces Stannin Gene Expression in HUVECs. Trimethyltin- mediated cell death is known to require Snn (Thompson et al., 1996). However, the effect of TMT on the expression of Snn in vulnerable cell types has not been previously assayed. To characterize the genetic response of HUVECs to TMT exposure, a 5 µM dose of TMT was added to the culture medium for 1.5-24 hours and levels of Snn mRNA assayed using quantitative real-time PCR. A significant increase in Snn mRNA was found in HUVECs after 1.5 hours of exposure to 5 µM TMT, with a more robust induction of Snn observed at 6 hours (approximately 4 fold; Figure 7A). This induction

appears to be maintained for an extended period of time, as the levels of Snn mRNA were still 2 fold higher than the level of unexposed control cultures 24 hours after exposure to

TMT (Figure 7A). It is possible that the increase in Snn mRNA observed after TMT treatment is mediated by TNFα. To demonstrate that TMT administration does induce the expression of TNFα responsive genes, interleukin (IL)-6, a known TNFα-inducible gene, was examined for induction after 5 µM TMT administration. Significant upregulation of IL-6 mRNA was observed at 3, 6, and 9 hours of TMT exposure (Figure

7B), which corresponds to published literature examining IL-6 levels after TMT administration (Harry et al., 2002). These results indicate that TNFα-induced Snn mRNA expression may be an important component of TMT’s toxic activity in HUVECs.

Stannin Knockdown Rescues HUVECs from TMT Toxicity. To effectively knockdown Snn in HUVECs, we designed siRNA specific for human Snn. To validate the Snn siRNA, we transfected HUVECs with Snn siRNA using the siPORT reagent

(Ambion) and allowed the cells to incubate for 48 hours. The cells were then harvested and quantitative real-time PCR (QRT-PCR) was used to assess the expression levels of

Snn mRNA. As is shown in Figure 8A, each species of siRNA resulted in a significant level of knockdown (≥60%) of Snn mRNA expression after 48 hours. While this was taken as a positive indication of the siRNA’s ability to knock Snn mRNA expression down, a further verification was required to determine if this results in decreased Snn protein expression as well. Currently, there are no specific anti-Snn antibodies available.

Therefore, we validated our Snn siRNA by performing a functional assay. Earlier studies on Snn have shown that the Snn protein is required for TMT-induced cytotoxicity

(Thompson et al., 1996); if the Snn siRNA knocks down the Snn protein, a protection

Figure 6: TNFα is Important for TMT Toxicity. To determine the requirement for

TNFα in HUVEC response to TMT, a TNFα neutralizing antibody (nTNFα) was used.

HUVEC cells were exposed to a 10 µM dose of TMT ± 3 µg/ml of a mouse anti-human neutralizing TNFα antibody or a mouse anti-human α-synuclein antibody for 48 hours.

Blockade of TNFα with a neutralizing antibody resulted in a significant level of protection (60% survival vs. 40%) against TMT toxicity in HUVEC cells whereas the α-

synuclein antibody treatment resulted in no significant protection against TMT-induced

cytotoxicity. * p < 0.001, ** p < 0.05 (n = 3)

120 * o 100 ** 80

0 Control TMT TMT & nTNF αα TM T & - Antibody Synuclein Cell Survival as a Percent of Contr Antibody

Figure 7: TMT Exposure Results in Increased Stannin Gene Expression. (A) In order to assess the effect of TMT on HUVEC expression of Snn, HUVECs were exposed to 5 µM TMT for 1.5 - 24 hours. Quantitative real-time PCR was used to assess the levels of Snn mRNA in culture after TMT administration. A 5 µM dose of TMT significantly upregulates Snn expression at 1.5, 6, 9, and 24 hours of exposure. * p < 0.05

(n=3)

(B) As a control for the functional role of TNFα in TMT treatment, the gene expression level of IL-6, a known TNFα-inducible gene, was determined using QRT-

PCR after TMT administration. HUVECs exposed to 5µM TMT for 1.5 - 24 hours showed a significant upregulation of IL-6 at 3, 6, and 9 hours of exposure.

A) B)

Stannin 5 * 5 Interleukin-6 A 4.5 4.5 N * A R

4 N 4 R n m i 3.5 m 3.5 * 6 -

3 L ann 3 t 2.5 * 2.5 ** of S * on of I 2 i 2 1.5

ction 1.5 nduct u 1

d I 1 0.5 Fol 0.5 ld Ind 0 Fo 0 0 5 10 15 20 25 0 5 10 15 20 25 Hours of Exposure to Trimethyltin Hours of Exposure to Trimethyltin

from TMT toxicity should be observed. Indeed, this protective effect was observed in

HUVECs transfected with Snn siRNA (Figure 8B); a 20 nM mix of the two siRNA species (50:50 ratio) completely protected the cells from TMT-induced toxicity. These data indicate that Snn is required for the action of TMT in HUVECs.

Induction of Snn mRNA by TNFα. To determine the level of Snn mRNA

expression in HUVEC cells following treatment with TNFα, QRT-PCR was used.

Figure 9A illustrates the temporal characteristics of this induction, with significant

induction of Snn mRNA occurring by 1.5 hours of TNFα treatment, and reaching a

maximum of 8-fold (± 0.32) induction above control levels after 3 hours of TNFα

treatment. By 6 and 9 hours, the level of Snn mRNA expression was 5-fold (± 1.53) and

3.4-fold (± 1.14) above control levels, respectively. After 24 hours of TNFα exposure,

the level of Snn mRNA in HUVEC cells returned to basal levels. Figure 9B shows a

more direct representation of TNFα-mediated induction of Snn mRNA expression via

reverse transcription PCR (RT-PCR) followed by agarose gel electrophoresis. As a

control to determine if upregulation of Snn mRNA was a TNFα-specific effect, HUVECs

were also treated with IL-1β. The specificity of TNFα induced Snn gene expression was

examined using IL-1β since IL-1β and TNFα are both intimately involved in the

inflammatory and immune responses and are capable of upregulating some of the same

gene products (Apte and Voronov, 2002). As shown in Figure 9A, IL-1β did not

significantly upregulate Snn mRNA levels at any of the time points examined. Together,

these results show that TNFα specifically induces Snn mRNA expression in HUVEC

cells with a relatively rapid time of induction.

Figure 8: Knockdown of Stannin mRNA Expression in HUVECs. (A) HUVECs were

transfected with 20 nM of an inactive siRNA or one of two Stannin-specific siRNAs and allowed to incubate for 48 hours. Real-time PCR was used to determine the amount of stannin mRNA present in the HUVEC cultures. At 48 hours of exposure, both stannin- specific siRNA’s showed a knockdown of at least 65%, relative to both the untreated

HUVECs and HUVECs treated with 20 nM of an inactive siRNA species. * = p < 0.05

(n=3)

(B) HUVECs were transfected with stannin siRNA and allowed to incubate for 24 hours. After 24 hours of siRNA exposure, 5 µM TMT was administered to the culture vessels of both stannin siRNA transfected cells as well as previously naïve cells. After

24 hours, cultures were washed once in PBS, trypsinized, and cells counted on a hemacytometer using a trypan blue exclusion assay. Stannin knockdown, via stannin- specific siRNA, protected HUVECs from TMT-induced cytotoxicity. * = p < 0.05 (n=3)

(A) 140

A 120 N R 100

80 60 * * 40

Expression of Stannin m 20

0 Contr ol Contr ol s iRNA SNN s iRNA 1 SNN s iRNA 2

(B)

o r t

n 120 Co f 100

t o 80 * 60

le Cells, as a Percen 0

Viab Vehicle 5 uM TMT 20 nM siRNA 5 uM TMT & 20 nM siRNA Treatment * p < 0.05

Figure 9: Temporal pattern of Snn mRNA expression following treatment with

TNFα. (A) Expression of Snn mRNA was determined by QRT-PCR. A significant increase in Snn mRNA expression was observed at 1.5 hours and peaked at 3 hours after treatment with TNFα (200 ng/ml), as compared to controls. In contrast, IL-1β did not significantly increase Snn mRNA expression at any of the timepoints examined * = p <

0.05, ** = p < 0.01 (Student’s T-test, n = 4)

(B): Representative photograph of Snn mRNA RT-PCR product stained with

ethidium bromide and separated in a 2% agarose gel.

α F

α er F d N d s of T T

La s l

ur o p ur (B) r 0 b ank 5 Ho l ont 10 B C 1. 3 Ho (A) 300bp

TNF α 200bp 9 ** IL-1β –Snn 8 100bp 7 * mRNA –GAPDH 6

n 5 * sio es

r 4

Exp 3 2

Fold Induction of Stannin 0 0 5 10 15 20 25

Hours of Exposure

Protein Kinase C (PKC) is required for TNFα-induced Snn mRNA expression.

PKC is known to be vital for TMT cytotoxicity in some cell types and is activated in response to TNFα signaling in HUVECs (Basu et al., 2002; Pavlaković et al., 1995).

Based on these observations, we hypothesized that PKC might mediate TNFα-induced

Snn gene expression. To test this hypothesis, we first used pharmacological inhibitors of

PKC to determine the role of PKC in TNFα’s induction of Snn mRNA expression. For

all experiments, QRT-PCR was used to determine levels of Snn mRNA expression.

When HUVEC were pretreated with bisindolylmaleimide I, an inhibitor of the βI, βII, γ, δ,

and ε isoforms of PKC at 10 nM (Toullec et al., 1991; Gekeler et al., 1996), TNFα-

induced upregulation of Snn mRNA expression was completely blocked (Figure 10A).

In contrast, pretreatment of HUVECs with Gö6976, an inhibitor of the α and βI isoforms

of PKC at 8 nM (Martiny-Baron et al., 1993; Gschwendt et al., 1996), showed no

significant effect on TNFα−induced Snn mRNA expression (Figure 10A). These results may indicate that one or more of the βII, γ, δ, or ε isoforms of PKC may be responsible

for the observed induction of Snn mRNA by TNFα in HUVECs. To determine whether

activation of PKC is sufficient to induce Snn mRNA expression in HUVECs in the

absence of TNFα, phorbol-12-myristate-13-acetate (PMA) was used as a direct activator

of the classical and novel subfamilies of PKC. Exposure to 100 nM PMA for 30 minutes

resulted in a significant upregulation of Snn mRNA (5 fold ± 1.08) above control levels

(Figure 10B). Together, these results indicate that stimulation of one or more PKC

isoforms is sufficient to induce Snn mRNA expression.

Figure 10: Pharmacological PKC inhibitors block TNFα-induced Snn mRNA expression. (A) QRT-PCR was used to determine the levels of Snn mRNA in TNFα- stimulated (200 ng/ml) HUVEC cells pretreated for 3 hours with either 10 nM bisindolylmaleimide-I (an inhibitor of PKC βI, βII, γ, δ, and ε) or 8 nM Gö6976 (an inhibitor of PKC α and βI). * = p < 0.05 (Student’s T-test, n = 3)

(B) Phorbol 12-myristate, 13-acetate (PMA) stimulates Snn mRNA expression.

The level of SNN mRNA in HUVECs was examined using QRT-PCR. PMA (100nM) treatment for 30 minutes was sufficient to induce Snn mRNA expression in HUVECs. *

= p < 0.05 (Student’s T-test, n = 3)

Knockdown of PKC Epsilon via siRNA Prevents TNFα-Mediated Snn

Upregulation. A search of the literature to date indicates that only the α, δ, ε, and ζ isoforms of PKC are expressed in HUVECs and that TNFα activates the α, δ, and ε isoforms of PKC in HUVECs (Satoh et al., 2004; Basu et al., 2002; Mehta et al., 2001;

Ross and Joyner, 1997). Therefore, we next examined the role of specific PKC isotypes as putative mediators of TNFα-induced Snn mRNA expression. To examine the regulatory role of PKCδ, ε, and ζ we utilized isotype specific PKC siRNA to selectively knock down these enzymes in HUVECs. Figure 11A shows that siRNA targeted to

PKCε blocks TNFα-stimulated induction of Snn mRNA expression, with a maximal inhibition after 48 hours of exposure to 200 nM of siRNA. In contrast, treatment of cells with PKCδ siRNA, PKCζ siRNA, or a control siRNA, did not block TNFα-induced Snn mRNA expression. Together, these results demonstrate that PKCε, but not PKCδ or ζ, modulates TNFα-induced increased Snn mRNA expression in HUVECs. Figure 11B shows a representative western blot of PKCε knockdown after siRNA treatment in

HUVECs. In order to assess the cell-type specificity of PKCε-mediation of TNFα- induced Snn expression, we examined Jurkat T-cells, which have been shown to express

Snn (Thompson et al., 1996). As shown in Figure 12, Jurkat T-cells also showed an increase in Snn mRNA in response to TNFα treatment in a PKCε-dependent manner.

Based on these results, involving two different Snn expressing cell lines responding to

TNFα with a PKCε-mediated increase in Snn mRNA expression, it is suggested that this might be a common mechanism of TMT-induced toxicity in Snn expressing cell types.

Figure 11: PKCε is Required for TNFα-Mediated Upregulation of Snn mRNA. (A)

HUVECs were transfected with an inactive siRNA or a mix of two PKCε siRNAs and allowed to incubate under normal culture conditions for 45 hours. HUVECs were then exposed to a single 200 ng/ml dose of TNFα and allowed an additional incubation time of 3 hours. Real time PCR was utilized to determine the amount of stannin present in the culture. At 48 hours, 200 nM of PKCε siRNA completely blocked TNFα-mediated upregulation of stannin mRNA. * = p < 0.05 (Student’s T-test, n = 3)

(B) Western blot showing PKCε knockdown at 48 hours of 200 nM siRNA exposure. The same concentration and time of exposure to a control siRNA has no knockdown effect on the expression PKCε. (n=3)

B) l ro

t siRNA ε

led Con l siRNA o r M PK

ont 00n C Scramb 2

12 * 10 * * 8 * 6 4 2 0 TNFα: - + + + + + Fold Induction of Stannin mRNA Control TNFα PKC ε Scrambled PKC δζ PKC siRNA siRNA siRNA siRNA Control

siRNA = 200nM in all cases

Figure 12: Jurkat T-Cells Respond to TNFα with a PKCε-Mediated Increase in Snn mRNA. Jurkat cells were transfected with an inactive siRNA or a mix of two PKCε siRNAs and allowed to incubate under normal culture conditions for 45 hours. Jurkats were then exposed to a single 200 ng/ml dose of TNFα and allowed an additional incubation time of 3 hours. Real time PCR was utilized to determine the amount of stannin present in the culture. At 48 hours, 200 nM of PKCε siRNA completely blocked

TNFα-mediated upregulation of stannin mRNA.

* = p < 0.05 (Student’s T-test, n = 3)

A N

R 2.5 * m * n i n 2 n a t S f 1.5 o n o i t Expression 1 c u d n I

0.5 d l o F 0 + + TNFα - + Vehicle Vehicle PKC ε siRNA Control siRNA

D. Discussion

Currently, the normal cellular function of Snn is unknown. Snn was originally

characterized as a common factor present in cells that are sensitive to TMT, and was

found to be necessary, but not sufficient, for TMT toxicity (Toggas et al., 1992;

Thompson et al., 1996). Snn’s high degree of sequence conservation across diverse

vertebrate species suggests an important role in cell function (Table 1). As a result,

antibodies specific for Snn have been difficult to develop. To date, very few single

nucleotide polymorphisms (SNPs) have been characterized in the coding portion of the

Snn gene, suggesting that mutations to Snn are detrimental (www.ncbi.nlm.nih.gov).

Though the only known function for Snn involves mediating TMT toxicity, it is obvious

that Snn has not been conserved across vertebrates for this purpose. However, it is

possible that levels of Snn can lead to enhanced or reduced sensitivity to TMT toxicity.

In this report, we presented the first evidence of TMT toxicity in endothelial cells

and showed that Snn is required for TMT toxicity. Moreover, we demonstrated that

TNFα-induced Snn gene expression is regulated by protein kinase Cε, thus providing

insight as to how TMT exposure may result in cell death. Our work showed that while

Snn is required for TMT toxicity in HUVECs, blocking TNFα does not completely

rescue HUVECs. However, siRNA knockdown of Snn completely protects HUVECs

from TMT toxicity. There are at least two possible reasons for this observation. One, the

level of neutralization of TNFα expected using the anti-TNFα antibody was most likely

not complete; it is possible that the remaining unbound TNFα was able to effectively

upregulate Snn enough to allow TMT toxicity in some HUVECs. Two, it is possible that

Snn is regulated by additional signaling pathways. In either case, it is clear from our

60 work and that of others (Harry et al, 2003) that TNFα, though involved, is not a causal part of the TMT signaling cascade.

Since PKC has been implicated as having a regulatory role in both TNFα-initiated signaling events and TMT toxicity (Basu et al., 2002; Pavlaković et al, 1995), the functional role of PKC in Snn gene expression is not surprising. Studies have demonstrated that PKCε is involved in several cellular processes including cellular proliferation (Soh and Weinstein, 2003; Akita et al., 2002), cell death (Jung et al., 2004;

Comalada et al., 2003), and the immune response (Aksoy et al., 2004). The apparent regulation of stannin by PKCε is supported by the observation that stimulation of PKCε in vitro results in cell death in response to lipopolysaccharide (Camalada et al., 2003) and oxidative stress (Jung et al., 2004). In addition, PKCε is found in all the major TMT- sensitive tissues/cells including: neurons (Jung et al., 2004), glia (Xiao et al., 1994), B- and T-cells (Krappmann et al., 2001), and lung (Lang et al., 2004).

Although this study elucidated a signaling pathway leading to enhanced Snn gene expression, the specific transcription factor(s) underlying TNFα-induced Snn gene expression is not yet known. In our previous study, we have identified 26 putative transcription factor binding sites on the Snn gene promoter region (Dejneka et al., 1998).

However, functionality of these sites has not been determined. A review of the literature, together with the data presented in this report, implicates the human aryl hydrocarbon receptor nuclear translocator (hARNT) as a possible candidate for regulating Snn expression in HUVEC cells. Several characteristics of hARNT support this hypothesis.

First, an hARNT consensus binding sequence is present in the Snn promoter region

(Dejneka et al., 1998). Second, hARNT has been shown to be regulated by PKC (Long et al., 1999); this attribute provides a mechanism whereby Snn expression would be induced

61 by PKC. Third, hARNT is active in endothelial cells and Jurkat cells (Nagai et al., 2002;

Adelman et al., 1999).

In addition to transcription factors, it is possible that TNFα-induced mRNA upregulation is modulated by additional mechanisms. One such mechanism may involve mRNA stability. The Snn mRNA has a very long 3’-untranslated region (UTR, Dejneka et al., 1998); the 3’-UTR of mRNAs is a significant source of regulatory control

(Ostareck-Ledrer et al., 1994; Mendez et al., 2000) and the binding of regulatory elements to this region of the mRNA may well regulate the stability of the message and the rate of translation in the cell. Analysis of the Snn mRNA sequence has uncovered 9 candidate iron regulatory elements (IREs) in the Snn mRNA. Several papers have been published regarding TNFα’s role in iron homeostasis in the cell as well as the effect of

PKC on iron regulatory protein (IRP)-iron responsive element (IRE) binding in vitro

(Kwak et al., 1995; Xiong et al., 2003; Thomson et al., 2000). Thus, it is possible that the pattern of Snn induction resulting from stimulation by TNFα or PMA is due to a modulation of regulatory factors existing in the Snn mRNA’s 3’-UTR.

We have previously shown that Snn is primarily localized to the mitochondria

(Davidson et al., 2004). This localization implies that mitochondrial dysfunction may be an integral downstream event initiated by Snn in TMT toxicity. Supporting this, multiple studies have demonstrated that mitochondrial dysfunction occurs in vitro after TMT exposure (Skarning et al., 2002; Stine et al., 1988). In summary, we demonstrated that

Snn is a critical signaling molecule that mediates TMT-induced cytotoxicity in a TNFα-

PKCε-dependent manner. Based on evidence from our study and previous studies, we propose a model of TMT-induced cytotoxicity, which is illustrated in Figure 13.

Figure 13: A Proposed Model of TMT Cytotoxicity.

In response to TMT exposure, TNFα is released into the extracellular microenvironment. TNFα then may bind and stimulate TNFα receptors (TNFR) on the

cell surface in an autocrine manner. It is possible that other inflammatory cell types may

produce TNFα after TMT exposure in vivo, which can bind and stimulate HUVEC in a

paracrine manner. Stimulation of TNFR, results in PKCε activation, which in turn

induces stannin (Snn) expression. Stannin protein is then translocated to the

mitochondria, possibly resulting in disruption of mitochondrial function, resulting in cell

death.

Chapter IV

Microarray Analysis of Stannin Knockdown in Human Umbilical Vein Endothelial Cells in TNFα Response; Implications for Cell Cycle Control

A. Introduction

Tumor Necrosis Factor α (TNFα) is a pleiotropic cytokine capable of regulating

many disparate cellular processes in a dose- and cell type- specific manner.

Dysregulation of TNFα results in several pathological states including: rheumatoid

arthritis, asthma, septic shock, irritable bowel disorder, hemorrhagic fever, and cachexia

(Locksley et al, 2001). The diversity of cellular effects caused by TNFα results from

different combinations of: the form of TNFα, cellular receptors and internal cellular programs. TNFα has two known forms; the first is a 26 kDa transmembrane protein,

which can then be cleaved to form a second, a 17 kDa soluble form. TNFα has two

receptors; tumor necrosis factor receptor -1 and –2. The soluble form of TNFα

preferentially binds to the TNFR1 receptor over the TNFR2 receptor (Grell et al, 1995).

Other studies have shown that the membrane-bound form of TNFα has a greater

(Tartaglia et al, 1993) or lesser (Grell et al, 1998) affinity for TNFR2 relative to TNFR1.

The available permutations of two receptors and two forms of TNFα underscores the fact

that TNFα is capable of giving rise to a range of cellular effects. These effects include

but are not limited to: cellular differentiation, modulation of inflammatory and immune

responses, and mediation of cell growth and differentiation (Yoshizumi et al., 2004;

MacEwan DJ, 2001; Leeuwenberg et al, 1995). In endothelial cells, it has been shown

65 that TNFα halts cell growth and induces activation/differentiation (Fukushima et al.,

2005; Moldenhauer et al., 2004; Langeggen et al., 2001).

Snn is an 88 amino acid protein that is highly conserved throughout vertebrate evolution (Reese et al., 2005; Dejneka et al., 1998). Rat and mouse Snn amino acid sequences are 100% identical, and human Snn differs by only two amino acids at the C- terminus. Further, mouse and human Snn nucleotide sequences are 90% identical

(Dejneka et al., 1998). Such a highly conserved nature implies an important role for Snn in normal cellular function. Snn is known to be necessary, but not sufficient, for TMT toxicity (Thompson et al., 1996). Recently, we have shown that Snn mRNA expression is modulated by TNFα in a protein kinase Cε (PKCε)-dependent manner in Human

Umbilical Vein Endothelial Cells (HUVEC; Reese et al., 2005). Both TNFα and PKCε are known to be involved in many cellular processes including: apoptosis, inflammation, and regulation of cellular proliferation (Locksley et al., 2001; MacEwan DJ, 2001;

McJilton et al., 2003; Baines et al., 2002; Petrovics et al., 2001).

Because of the high degree of conservation, analysis of native Snn protein is difficult due to the lack of specific, high affinity antisera. Without this tool, alternative methods are required to determine how levels of Snn effect cellular processes. Thus, microarray technology was utilized in order to examine how Snn may be involved in the

HUVEC cellular milieu in response to TNFα exposure. This study examined the effects of Snn knockdown, via siRNA, on the HUVEC response to TNFα. Many genes were significantly altered between TNFα-exposed HUVECs and TNFα-exposed HUVECs that had Snn knocked down via siRNA. Several of these genes play roles in cell growth, transcription, and cytoskeletal integrity. Further, we show that knockdown of Snn via

66 siRNA results in further inhibition of cell growth beyond that observed with TNFα alone, potentially through an arrest of the cell cycle at the G1/S checkpoint.

B. Materials and Methods

Cell Culture

Human umbilical vein endothelial cells (HUVEC) were obtained from Cambrex

(East Rutherford, NJ). HUVECs were maintained as recommended by Cambrex.

Briefly, cells were cultured in the chemically-defined EGM™ media (Cambrex, East

Roche, Indianapolis, IN) was dissolved in sterile PBS and administered at 200 ng/ml.

Microarray Fabrication

A total of 9,998 amine-modified, human 50-mer oligonucleotide probes (MWG

Biotech, Germany) were printed onto epoxysilane slides (Schott-Nexterion) using a

MicroGrid II robotic microarrayer (Genomic Solutions, Ann Arbor, MI). The printed arrays were then incubated at 45% relative humidity for 8 hours to allow for optimal binding of the probe to the surface of the slide, then washed in 0.02% SDS for 2 minutes at room temperature. The slides were rinsed with several volumes of distilled, de-ionized water, immediately spun dry via centrifugation, then stored in a dessicator prior to their use.

Microarray cDNA probe synthesis and indirect labeling with AlexaFluor 555 and 647

A total of 10 ug total RNA from each sample or 10ug of universal reference RNA

(Stratagene, La Jolla, CA) was reverse transcribed to generate cDNA probes for

hybridization. Amino-modified dUTP and dATP were incorporated into each transcript

using SuperScript™ III RT (Invitrogen Corporation, Carlsbad, CA) per manufacturer’s protocol. After reverse transcription, template RNA was degraded by base hydrolysis, and the reaction neutralized with 1 N HCl. The reaction volume was brought up to 500 ul with water and transferred to Microcon® Centrifugal Filter Units (Millipore, Billerica,

MA) for removal of unincorporated nucleotides and primers.

The concentrated cDNA was then coupled to the active form of either Alexa

Fluor® 555 (reference group), or 647 (experimental groups) labeling conjugates

(Molecular Probes, Eugene ,OR). Dye-labeled cDNAs were then purified with a low- elution spin column to remove any unreacted dye molecules. Each labeled experimental condition sample was co-hybridized with the labeled exogenous universal human reference sample, in a 1:1 ratio of 1500 ng per sample. Directly comparing the experimental samples to the on-chip reference sample enabled indirect comparisons between experimental groups to be met with equal efficiency (Churchill, 2002).

Gene expression analysis

The scanned images of each microarray were quantitated with Scanarray

Express™ image analysis software (Perkin Elmer Life and Analytical Sciences, Inc.,

Boston). The raw data was then formatted and imported into Genespring 7.2 (Silicon

Genetics, Redwood City, CA) for normalization and comparative analyses. A Lowess curve was used to normalize the two-color gene expression dataset and adjust the control value (universal human reference) for each measurement. An additional normalization step was applied to pair each experimental condition against the average intensity of the

control condition array replicates. The measurement for each gene present in the

experimental samples was divided by the median of that gene's measurements in the

corresponding control sample. The control measurements had to be flagged as being

present, or the data was not reported.

Cell Growth

The trypan blue exclusion test was used as a measure of cell growth. HUVECs

were incubated in 0.2% trypan blue (Sigma, St. Louis, MO) diluted in phosphate-buffered

saline (PBS) and then subsequently washed once with PBS. The number of normal and

blue-stained, dead cells were counted in four independent microscopic fields per culture,

with three independent cultures being used for each condition (12 total fields). The

number of viable cells was compared in each treatment condition with the initial plating

density, which was considered 100% for the purposes of relative viability.

RNA Isolation/cDNA synthesis

RNA isolation was accomplished using the RNeasy kit, according to the protocol

recommended by the manufacturer (Qiagen, Valencia, CA). Briefly, HUVECs were

harvested using 0.05% trypsin-EDTA and pelleted at 2000 x g. Cells were then

resuspended in 350 µl of RLT cell lysis buffer and homogenized using QIAshredder

homogenization columns (Qiagen, Valencia, CA). The homogenized mixture was

combined with an equal volume of 70% ethanol and added to an Rneasy RNA isolation

column and spun at 8000 x g for 15 sec. The RNA bound on the column was washed

three times, and finally eluted with Rnase free water as indicated by the manufacturer.

The synthesis of cDNA for QRT-PCR was carried out using the First Strand cDNA

Synthesis Kit (MBI Fermentas, Hanover, MD). This kit employs a standard M-MLV

reverse transcriptase reaction and was used according to the recommendations of the

manufacturer.

Quantitative Real-Time PCR (QRT-PCR)

The cDNA template from HUVECs were normalized based on their relative

expression of β-Actin. To detect human Snn, the following primers and probe were used

to amplify a 100 bp product corresponding to bases 222-322 of the mRNA:

Forward Primer: 5’ – TTG TCA TCC TCA TTG CCA TC – 3’

Reverse Primer: 5’ – GCT CTC CTC GTC CTC TGA CT – 3’

Probe: 5’ – CCT GGG CTG CTG GTG CTA CCT – 3’

Pre-developed 20X primer-probe assay kits (Applied Biosystems, Foster City, CA) were used for the following genes: β-Actin, E-Selectin, cdc42BP, HRasLS, PRKC,

Phospholipase A2, GCIP, IL-4, and MDM4.

Reactions were carried out using a protocol from Qiagen (Valencia, CA). The

PCR program was as follows: step 1- 95oC for 15 minutes, step 2 – 95oC for 15 seconds,

step 3 – 60oC for 1 minute, with steps 2 and 3 repeated for 40 cycles. All reactions were carried out using the ABI Prism 7700 Lightcycler.

siRNA Construction

Snn siRNA was constructed using the Silencer siRNA Construction Kit

(Ambion, Austin, TX). The following oligonucleotides were utilized to construct siRNA

(only the sense strand is given, shown without T7 adapter sequence):

Snn siRNA 1: 5’ – AAG GAA CCC TTC CTG CTG GTG – 3’

Snn siRNA 2: 5’ – AAG GGA CCG TGC GTG GAG AGA – 3’

Transfection of siRNA

All siRNA used in these studies was transfected into HUVECs using the siPORT

Lipid reagent (Ambion, Austin, TX). Briefly, siPORT Lipid reagent was diluted in Opti-

HUVECs were washed in Opti-MEM I, then fresh Opti-MEM I was added to the cells in place of EGM media. The siPORT Lipid/siRNA mixture was then overlaid dropwise onto the cells. In all experiments, a 50:50 mix of two Snn-specific siRNA were used.

Statistical Analysis

After normalizations, the TNFα alone and Combo groups were compared using a

1-way ANOVA with a cutoff p-value of p < 0.001. The results of this ANOVA were then filtered to show only those genes that differed between the two groups by 2.5 fold or more. For the QRT-PCR and cell cycle experiments significance was determined using a

Student’s T-test between treatment conditions and the control or vehicle condition, unless otherwise specified. A p-value of less than 0.05 was considered significant in all cases.

Each experiment consisted of at least 3 replicates per condition. The growth curve presented in Figure 6 was analyzed using a two-way analysis of variance (ANOVA) followed by a Fisher’s least significant difference (LSD) post hoc test.

C. Results

Snn knockdown results in significantly altered HUVEC gene expression in response to TNFα. TNFα is capable of modulating Snn gene expression in a PKCε- dependent manner in HUVECs (Reese et al., 2005; Horrevoets et al., 1999). In an effort to better define a potential functional role for Snn in TNFα signaling in HUVECs, microarray technology was utilized.

Four independent groups were examined for differential gene expression: a control group of HUVECs growing in EGM™ complete media (Control), a group of

HUVECs that received a single 200 ng/ml dose of TNFα (TNFα alone), a group that was transfected with a 20 nM Snn siRNA mix (Snn siRNA alone), and a group that received both TNFα and Snn siRNA (Combo; see Figure 14 for experimental design). Results were normalized first in a LOWESS algorithm, then back to the average signal intensity of the control group. This second normalization allows for a more meaningful comparison of the other groups to “normal state” HUVECs. The control and Snn siRNA alone groups were compared using a one-way ANOVA with a p-value cutoff of p <0.001 and a 2.5 fold difference in expression filter. A total of 8 genes passed this combination of normalizations and statistical tests and, since the purpose of this study was to examine a potential role for Snn in TNFα signaling, these 8 genes were excluded from the final analysis. For clarity, since the significant changes between the control and Snn siRNA alone groups were excluded, the Snn siRNA alone group was not considered further in this study. This same combination of normalizations, filters, and statistical tests were used to compare the TNFα alone group to the Combo group. These groups proved much more divergent, with 100 genes passing these normalizations and tests. This group of

Figure 14: Experimental design. Four groups of HUVECs were plated (two biological replicates per group). The groups were: a control group grown in EGM complete media, a group which was grown as control but which received a 200ng/ml dose of TNFα 60 hours after plating, a group grown as control that received 20nM Snn siRNA 24 hours after plating, and a “Combo” group that was both transfected with Snn siRNA and dosed with TNFα at the appropriate timepoints. Before any treatments, HUVECs were allowed

24 hours of undisturbed growth subsequent to plating. The RNA isolated from each plate of HUVECs at the “Harvest” timepoint was split into 3 aliquots, each used on a separate microarray for a total of 6 technical replicates per group (1 array each from the control and combo groups did not process correctly and so these groups have n =5 rather than 6).

100 genes were deemed “significantly altered” for the purposes of the remainder of the

analysis. A hierarchical cluster diagram of the raw data is shown in Figure 15, with the

genes that are upregulated compared to the control condition highlighted in red and the

genes that were downregulated compared to the control condition highlighted in green.

Of the genes that were deemed significantly altered, an ontological classification is

shown in Figure 16. This classification was made using the Database for Annotation,

Visualization, and Integrated Discovery (DAVID; apps1.niaid.nih.gov/david/upload.asp).

Of the 100 genes to be entered into the DAVID software, 96 were successfully grouped into a biological function with 68.75% being categorized as involved in cell growth and maintenance, signal transduction, or nucleobase, nucleoside, nucleotide, and nucleic acid metabolism.

Snn Knockdown significantly affects several genes involved in cell growth. The cell growth and maintenance classification represents 33% of the total of significantly altered genes, with signal transduction and nucleobase, nucleoside, nucleotide, and nucleic acid metabolism representing 18.75% and 17.7%, respectively. Since these ontological classifications represented a large majority of the significantly altered genes, we chose to validate our array results by using QRT-PCR on 8 genes from these classifications. Three of the 8 genes have direct implications for cell cycle regulation at the G1/S checkpoint and 7 of the 8 have been described as important for control of cell

growth. Interleukin 4 (IL-4) is a soluble cytokine known to have anti-inflammatory

actions (Rocken et al., 1996). In addition, IL-4 has been shown to strongly inhibit the

progression of HUVECs in the G0 + G1 phase of the cell cycle by affecting p53, p21WAF, cyclin D1, and cyclin E (Kim et al., 2003). When Snn is knocked down, TNFα exposure

Figure 15: A cluster-based representation of the altered genes across the control,

TNFα only, and combo treatment groups. Genes that are upregulated relative to the control condition are colored green and genes that are downregulated are colored red.

Initial analysis comparing the control condition to the Snn siRNA alone condition resulted in 7 genes being statistically different out of 9,998. Since the two groups were so similar, this condition has been left out of the cluster diagram.

Figure 16: Gene ontologies of significantly different genes (TNFα vs. Combo groups). After LOWESS normalization of each condition, the two major groups of interest, TNFα alone and the combo groups, were normalized to the control condition according to signal intensity. Following these normalizations, a one-way ANOVA was run on the average signal intensity values for the TNFα alone vs. the combo group

(cutoff p-value was 0.001). From the genes that passed this test, those that were 2.5 fold

different between groups were assembled as meaningfully different genes. Of the 100

genes that met these criteria, 96 were successfully grouped according to ontology using

the Database for Annotation, Visualization, and Integrated Discovery (DAVID).

Gene Ontologies of Significantly Different Genes (TNF α vs Combo) 8

10 31 Cell Growth and Maintenance

Signal Transduction

12 Nucelobase, Nucleoside, Nucleotide, and Nucleic acid Metabolism Protein Metabolism

Biosynthesis

Cell-Cell Signaling 17 18

resulted in a significant increase in IL-4 mRNA expression (9.25 fold ±1.89), whereas

with no Snn knockdown, a significant, though less robust, increase was observed (3.21

fold ± 0.82; Figure 17A). The gene p29 (GCIP) has been implicated in regulation of the

G1/S transition through its association with cyclin D, with an increase in GCIP expression

correlating with a halt of the cell cycle at G1 (Chang et al., 2000). When Snn was

knocked down, TNFα exposure resulted in a significant increase in GCIP mRNA

expression (3.36 fold ± 1.44), whereas with no Snn knockdown, no significant increase

was observed (Figure 17B).

The gene MDM4 has been characterized as a p53 inhibitor and knockdown of

MDM4 results in G1 arrest in MDM4 deficient mouse embryos and mouse embryonic fibroblasts (Migliorini et al., 2002). When Snn was knocked down, TNFα exposure resulted in a significant increase in MDM4 mRNA expression (3.97 fold ± 0.94), whereas with no Snn knockdown, no significant increase was observed (Figure 17C). HRas-like suppresor (HRasLS) has not been extensively characterized, but multiple studies have implicated HRasLS expression with growth inhibition, though the exact nature of this inhibition is not known (Kaneda et al., 2004; Akiyama et al., 1999). Again, when Snn was knocked down, TNFα exposure resulted in a significant increase in HRasLS mRNA expression (4.63 fold ± 0.62), whereas with no Snn knockdown, no significant increase was observed (Figure 17D). PRKC/WT1 is known to be a transcriptional regulator, and has been shown to both activate and repress transcription depending upon other, currently, unknown factors in the cell (Johnstone et al., 1996). PRKC is involved in regulating cell growth via interaction with p53 (Maheswaran et al., 1995; Maheswaran et al., 1993). When Snn was knocked down, TNFα exposure resulted in a significant

Figure 17: Differential Gene Expression of known cell growth effector genes. To validate the results of the microarrays, quantitative real-time PCR (QRT-PCR) was used.

A) In response to 200 ng/ml TNFα, HUVECs normally respond with a 3-fold upregulation of IL-4 mRNA, when Snn is knocked down, however, HUVECs respond to this same dose of TNFα with a 10-fold upregulation of IL-4 mRNA compared to the control group. B) the levels of p29/GCIP mRNA do not normally change in HUVECs after exposure to TNFα; however, Snn knockdown results in a 3-fold increase in p29/GCIP mRNA in response to TNFα compared to the control condition. C) No significant difference was found in the levels of MDM4 mRNA between TNFα-exposed

HUVECs and control cells, however, under conditions of Snn knockdown, TNFα results in a 4-fold upregulation of MDM4 mRNA compared to the control condition. D)

HRasLS mRNA was upregulated 4-fold after both Snn knockdown and TNFα exposure, but no significant difference was found in HRasLS mRNA levels between control and

TNFα alone groups. * = P < 0.05 (Student’s T-test, n = 3)

83 increase in PRKC mRNA expression (2.42 fold ± 0.4), whereas with no Snn knockdown, no significant increase was observed (Figure 18A). Cdc42 binding protein (cdc42BP) is thought to act as a downstream effector of cdc42, mediating the formation of new actin filaments and promoting growth via cytoskeletal reorganization (Leung et al., 1998).

Upon stimulation with TNFα, HUVECs showed a significant decrease in cdcd42BP (0.62 fold ± 0.07; figure 18B); this decrease was significantly enhanced when HUVECs were treated with Snn siRNA prior to TNFα stimulation (0.31 fold ± 0.08; Figure 18B).

E-Selectin is a cell adhesion molecule know to be involved in tumor metastasis and atherosclerosis (Laferriere et al., 2004; Zhang et al., 2002). In addition E-Selectin levels increase in HUVECs after TNFα stimulation (ten Kate et al., 2004; Klein et al.,

1994). As expected, upon TNFα stimulation, a significant increase in E-Selectin mRNA was observed (195.86 ± 28.25; Figure 18C). This increase was partially attenuated under conditions of Snn knockdown (69.30 ± 6.65; Figure 18C). Phospholipase A2, group VII

(PLA2), also known as platelet activating factor (PAF) acetylhydrolase, has potent anti- inflammatory effects and is thought to have a protective function against oxidized phospholipids on low density lipoprotein (LDL; Six and Dennis, 2000). In addition,

PLA2-mediated arachidonic acid release inhibits the proliferation of renal proximal tubule cells (Han et al., 2004). When Snn was knocked down, TNFα exposure resulted in a significant increase in Phospholipase A2 mRNA expression (6.76 fold ± 0.78), whereas with no Snn knockdown, no significant increase was observed (Figure 18D).

Loss of Snn gene expression functionally affects HUVEC response to TNFα.

TNFα is known to inhibit the growth of HUVECs in culture (Jin et al., 2000; Lin et al.,

1992; Mauerhoff et al., 1994). Further, the results of the microarray study implicate Snn

Figure 18: Differential expression of genes involved in cell growth, transcription, and cell adhesion. A) No significant difference in WT1/PRKC mRNA was observed between the control and TNFα groups, a significant upregulation of WT1/PRKC mRNA was observed between the control and Combo groups. B) TNFα administration results in a significant downregulation of cdc42BP mRNA, while the levels of cdc42BP mRNA in

the Combo group are significantly decreased compared to both the control and TNFα

alone conditions. C) E-Selectin is known TNFα-responsive gene in HUVECs; a

significant increase in E-Selectin mRNA was observed betweent he control and

TNFα alone groups, this upregulation was dampened, though still significant, when the

control and Combo groups are compared. D) Phospholipase A2 mRNA levels are

unaltered by TNFα administration alone, but are significantly altered when TNFα is

administered under conditions of Snn knockdown. * = P < 0.05 (Student’s T-test, n = 3)

involvement in TNFα-mediated inhibition of HUVEC cell growth. To examine the functional role of Snn in TNFα inhibition of HUVEC proliferation, the effect of Snn knockdown using siRNA was examined.

Snn knockdown was achieved by using previously validated, Snn-specific siRNAs and cell growth was assessed in order to determine if knockdown of Snn would affect the TNFα-mediated inhibition of HUVEC growth. HUVECs were initially plated at a density of 2.0 x 105 cells per condition. After 24 hours of growth, HUVECs were then transfected with 20 nM Snn siRNA for 24 hours. Subsequently, cells were treated with 200 ng/ml TNFα for 24 or 48 hours. As shown in Figure 19, when Snn was knocked down via siRNA, we observed an additional impairment of HUVEC growth at both 24 and 48 hours post TNFα-exposure. Specifically, knockdown of Snn resulted in approximately 50% of the number of cells being present in culture after 24 and 48 hours of exposure to TNFα, relative to HUVEC cells treated with TNFα alone. Further, a significant decrease was observed in the number of cells treated with Snn siRNA alone, as compared to vehicle control. Finally, all conditions differed significantly in the number of cells present after 48 hours of TNFα exposure with/without Snn siRNA treatment.

Knockdown of Snn inhibits the ability of TNFα-treated HUVEC cells to progress through the cell cycle. In an effort to examine the mechanism underlying the differential growth of TNFα-treated HUVEC cells due to Snn knockdown, cell cycle analysis was performed using flow cytometry after cells were stained with propidium iodide. As shown in Figure 7, there was a significant increase (7% ± 0.13%) in the proportion of

Figure 19: Alteration of HUVEC cell growth by treatment with TNFα and/or Snn siRNA. Cell numbers were determined using a hemacytometer after treatment with

TNFα (200ng/ml) for up to 48 hours with/without exposure to Snn siRNA (20nM for up to 72 hours). Growth is presented as a percent of the plating density of 2.0 x 105 cells.

Treatment with TNFα alone resulted in significant inhibition of growth at both 24 and 48

hours of exposure (72 and 96 hours of growth). Treatment of TNFα-stimulated HUVECs

with Snn siRNA resulted in significant inhibition of cell growth compared with TNFα

alone or siRNA alone and both 24 and 48 hours of TNFα exposure. Treatment of Snn

siRNA alone resulted in a significant inhibition of cell growth compared to vehicle

treated controls by 72 hours of exposure. * = P < 0.05 from each other group at that

timepoint † = P < 0.05 from both vehicle and TNFα + Snn siRNA groups at that

timepoint (Two way ANOVA, Fishers LSD post hoc test, n = 3)

700 Treatment with Treatment * Snn siRNA with TNFα

600 * 500 Control 400 † Snn siRNA

th (Percent) * TNFα w 300 * * Snn siRNA & TNFα 200

Cell Gro 100

-48 -24 0 24 48 Hours

Figure 20: Co-treatment with TNFα and Snn siRNA alters HUVEC progression through the cell cycle. The proportion of cells in each phase of the cell cycle was analyzed via flow cytometry after propidium iodide staining. HUVECs exposed to both

TNFα and Snn siRNA showed significant changes in their cell cycle compared to those treated with either TNFα or Snn siRNA alone. HUVECs were transfected with Snn siRNA (20nM for 24 hours) and were then treated with TNFα (200ng/ml) and allowed to incubate for an additional 24 hours. Neither TNFα nor Snn siRNA alone had any effect on the progression of HUVECs through the cell cycle. * P < 0.05 (Student’s T-test, n =

80 70 * s 60

50 Control Snn siRNA 40 TNFα 30 * TNF & siRNA 20

Percent of Total Cell * 10 0 G1 S G2

HUVECs in the G1 phase of the cell cycle when they were treated with both TNFα and

Snn siRNA vs TNFα or Snn siRNA alone. Accordingly, co-treatment with TNFα and

Snn siRNA resulted in a significant decrease in the proportion of cells in the S (5-6% ±

0.08%) and G2 (3-4% ± 0.21%) phases compared to those treated with TNFα or Snn siRNA alone. These results indicate that Snn may play a role in progression of the cell cycle at the G1/S transition phase in HUVEC cells.

D. Discussion

The results of this study implicate Snn as an important part of the HUVEC response to

TNFα. Major shifts in gene expression were observed between the TNFα and Combo microarray groups (Figure 15). An unexpected result was the high degree of similarity between the control and the Snn siRNA only groups. This may imply that the effect of

Snn on cellular gene expression is normally passive and requires a stimulus, such as

TNFα, to activate. However, microarray technology is limited and that the results must be thoroughly validated and used as a guide for additional assays. With that in mind, we validated 8 significantly altered genes from our microarray study using QRT-PCR. The consistency between the microarray and QRT-PCR assays is evident in Figures 17 and

18. From these results, it appears that Snn regulates genes which play a role in regulating cell growth in response to TNFα. Though the exact mechanism is unknown, it has been shown that TNFα is capable of slowing or halting HUVEC growth in vitro (Fukushima et al., 2005; Moldenhauer et al., 2004; Langeggan et al., 2001; Jin et al., 2000). This decrease in cell number is significantly augmented when Snn is knocked down via siRNA (Figure 19).

There are two possibilities underlying the difference in cell number shown in

Figure 19: one is that cell death is occurring and two is that the cell cycle is being halted

at one of the cell cycle checkpoints. Cell death was ruled out because a terminal

deoxynucleotidyl transferase biotin-dUTP Nick End Labeling (TUNEL) assay revealed

no significant cell death in our experimental paradigm (data not shown). Further, the

growth curves were generated using a trypan blue exclusion assay so that only viable

(live) cells were counted. Analysis of the cell cycle was made using flow cytometry and

showed that a significant perturbation of the cell cycle is occurring at the G1/S

checkpoint (Figure 20), consistent with the results from the microarray experiments.

Protein kinase Cε is involved in cellular proliferation (Soh and Weinstein, 2003),

inflammation (Akita, 2002), and the immune response (Aksoy et al., 2004). The apparent

role of Snn in cellular proliferation via PKCε is supported by the observation that

stimulation of PKCε, in vitro, resulted in augmented cellular proliferation via increased

cyclin D1 and cyclin E expression (Soh and Weinstein, 2003). Thus, the knockdown of

Snn was predicted to inhibit cellular proliferation; results in figure 20 support this conclusion. The significant alteration of genes known to be present in cyclin D1 signaling pathways (IL-4, p29) further supports this hypothesis. Specifically, IL-4 has been shown to decrease the expression of cyclin D1 and E in HUVECs and p29 is known to inhibit the transcription factor E2F, which is induced by cyclin D1 and transactivates the expression of genes that are involved in S phase progression of the cell cycle (Kim et al., 2003; Chang et al., 2000).

One challenge to interpretation of the microarray data is the seemingly conflicting effects that may be occurring on p53. The upregulation of MDM4 (Figure 17C), a strong p53 inhibitor conflicts with the upregulation of IL-4, known to induce the expression of

p53 and with the upregulation of WT1/PRKC which is known to increase the effective

level of p53 in cells via stabilization of the protein (Kim et al., 2003; Migliorini et al.,

2002; Maheswaran et al., 1995). The p53 protein is a primary inhibitor of the cell cycle at the G1 checkpoint and limits the ability of the cell to advance through the cell cycle.

From the growth and cell cycle data, it appears that p53 is likely to be functional in our experimental design and that the observed MDM4 upregulation may be a transient compensatory response initiated by the cell to maintain homeostasis.

In summary, we have shown that knocking Snn down via targeted siRNA results in differential gene expression in TNFα exposed HUVECs vs HUVECs exposed to

TNFα with Snn levels decreased by siRNA. Using QRT-PCR, we validated 8 genes that were significantly altered on the microarrays, and found that many of these genes are suspected or known to play a role in the regulation of the cell cycle and/or cell growth.

Specifically, it appears that genes modulating the expression and activity of cyclin D1 and p53 are significantly perturbed by Snn knockdown in HUVECs exposed to TNFα.

Functional studies show that cell growth is inhibited over 48 hours of TNFα exposure

and that this inhibition is further augmented by Snn knockdown. The mechanism likely

responsible for this decrease in growth rate is an arrest of the cell cycle at the G1

checkpoint via disruption of cyclin D1 and/or p53. Future studies will need to examine

protein levels of these and other cell cycle regulating proteins in order to solidify a

potential role for Snn in the HUVEC response to TNFα. One limitation of the HUVEC

model is that survival for a substantial amount of time in serum free conditions is very

difficult; this may necessitate choosing a different cell line in which to better observe the

effect of Snn knockdown on the cell cycle.

Chapter V

Discussion

A. Stannin Expression is Highly Regulated in a Spatial and Temporal Manner

Stannin (Snn) is widely expressed in developing vertebrates; progressively more restricted expression of Snn was observed as the embryo progresses to, and reached, adulthood (Toggas et al., 1992). This observation implies that a complex level of control over Snn expression has developed in vertebrates. Analysis of the Snn gene shows 32 potential transcription factor-binding sites and two potential promoter regions (Dejneka et al., 1998). The presence of so many transcription factor binding sites, including those for

MZF-1 and CP2, transcription factors known to play significant roles in blood forming cells, and a potential binding site for the neuronal transcription factor N-myc may help to explain Snn’s tissue specificity in the adult vertebrate. Though the spatial regulation of

Snn may be explained by this analysis of the Snn gene promoters, the temporal regulation of Snn is not fully explained by these analyses.

In addition to transcriptional regulation, it is possible that Snn is temporally regulated in a post-transcriptional manner as well. One characteristic of the Snn mRNA in support of this notion is the large 3’ UTR, which composes approximately 2.5kb of the

3kb mRNA. Analysis of the Snn mRNA indicates that there are 6 putative iron responsive elements (IREs) on the Snn mRNA (consensus sequence cagagc). Iron responsive elements present in the 3’ UTR are known to stabilize mRNA, thereby increasing its halflife significantly (Thomson et al., 1999; Henderson and Kuhn, 1997.

Thus, the existence of functional IREs may be one way for the cell to regulate the amount of Snn present in the cell at any one time.

As described in chapter 3, Snn gene expression is induced by TNFα. TNFα is a powerful modulator during development (Ohazama and Sharp, 2004; Cheng et al., 2003;

Toder et al., 2002). Cellular differentiation, as well as life and death signals, are effects attributed to TNFα in specific tissues during development. Though TNFα has been shown to upregulate Snn gene expression, it is unlikely that this is the only signaling molecule capable of doing so. The observation that TNFα is not required for TMT toxicity strongly implies that there are additional cellular mechanisms capable of inducing Snn gene expression. Future work directed at determining the functionality of the putative transcription factor binding sites on the Snn gene, as well as surveying other cellular signaling molecules for their ability to modulate intracellular Snn levels will be useful in understanding the specific role of Snn in development.

B. Stannin as a mediator of Trimethyltin Toxicity

Stannin is necessary, but not sufficient, for trimethyltin (TMT) toxicity. The requirement for Snn has been shown conclusively, but a mechanism has not been identified. Recent work has shown that Snn binds to and demethylates TMT to dimethyltin (DMT; Buck et al., 2003). A Snn peptide corresponding to amino acids 31-

39 containing two vicinal cysteines, was used to model this interaction. Further, transient transfection of Snn can sensitize cells to both TMT and DMT toxicity, implying that Snn may interact with DMT as well as TMT to cause cell death (Davidson et al., 2004).

Assuming the demethylation of TMT by Snn occurs in vivo during classic TMT poisoning, the significance of this event and how it contributes to cell death must be

96 ascertained. One theory raised by the Snn-TMT binding and dealkyation data is that

DMT may actually be the toxic effector in cells following TMT exposure.

DMT is less hydrophobic than TMT, potentially accounting for the reduced toxicity observed in cells as compared to TMT. Studies using Lipofectamine2000 to allow DMT to cross the plasma membrane have failed to show that this is the case, however, as TMT is still approximately 5 times more toxic than DMT-

Lipofectamine2000. Thus, it is unlikely that the presence of DMT in the cell is sufficient for the level of cell death observed after TMT exposure. Upon exposure to TMT, mixed neuronal/glial cultures produce TNFα approximately 6 hours later (Harry et al., 2003).

The data shown in chapter 3 indicate TNFα will lead to an increase in Snn mRNA approximately 3 hours later and, presumably, Snn protein within a few hours of that. The release of TNFα followed by the increase in Snn could serve as a potential positive feedback loop where large amounts of DMT build up in the cells.

The question remains as to whether Snn-DMT, Snn-CH3, free methyl groups, or some combination thereof is responsible for the eventual cell death observed after TMT exposure. To date, there is no evidence that Snn retains its binding to DMT or the methyl group after the dealkylation of TMT or that binding is covalent. Thus, it is possible that the release of free methyl groups and/or generation of reactive species via demethylation is a primary source of the toxicity of TMT. Hypermethylation can result in apoptotic cell death (He et al., 2005). In addition, the dysregulation of methyl metabolism, as in cases of choline or folate deficiency, can overwhelm the cells primary methyl acceptors/donors

(Methylenetetrahydrofolate reductase and S-Adenosylmethionine, respectively) of the cell, resulting in cell death via gene silencing (Liu et al., 2005; Liu et al., 2005). Future studies attempting to characterize the activity of these enzymes after TMT exposure,

along with a more detailed understanding of the intermediate in the reactive

demethylation, would greatly advance our understanding of how TMT kills cells.

An alternative hypothesis as to the mechanism of Snn-mediated TMT toxicity

involves 14-3-3ζ, a scaffolding protein that binds the proapoptotic protein Bad (Yang et

al., 2001). Snn binds to 14-3-3ζ in PC12 cells and it is possible that the competition for

14-3-3ζ binding with other proteins, such as Bad, may result in cell death after TMT exposure. However, none of our studies, including the microarray studies, indicated any alteration of 14-3-3ζ or known direct binding partners. A second order relationship may exist, however, studies to date have not explained the role of 14-3-3ζ binding in Snn’s action in the cell.

C. Stannin as a Component of Cell Cycle Progression

The data presented in chapter 4 implicates Snn in the regulation of the HUVEC cell cycle. The nature of this influence may be via alteration of the function of p53 and/or cyclin D1. Several genes influencing these proteins were shown as differentially expressed in the microarray analysis. IL-4, in particular, is highly upregulated after

TNFα exposure under conditions of Snn knockdown (approximately 10 fold) while

TNFα stimulation alone results in a more modest increase (approximately 3 fold). Direct evidence for IL-4 halting HUVEC cell growth at the G1 checkpoint indicates alteration of

p53, p21WAF1 and cyclin D1 as the cause (Kim et al., 2003). In addition to IL-4,

p29(GCIP), HRasLS, PRKC/WT1, and MDM4 have direct effects upon cell cycle

progression by affecting p53 and/or cyclin D1 function (Chang et al., 2000; Migliorini et

al., 2002; Akiyama et al., 1999; Johnstone et al., 1996). The alterations of these genes,

98 coupled with the functional data presented in chapter 4, provide a strong case for Snn being involved in HUVEC cell cycle progression in response to TNFα. In addition to these genes, others involved in supportive functions important for cell cycle progression including phospholipase A2 (MAPK alterations) and cdc42BP (actin cytoskeleton reorganization) are also significantly altered between the Combo and TNFα only groups presented in chapter 4. In sum, the microarray and functional follow-up studies have provided the first evidence suggestive of a role for Snn in normal cellular function. It is important to note, however, that this work analyzed gene expression and not protein expression. While there is typically a strong correlation between gene expression and subsequent protein experession, that may not be the case with Snn. Additional work will be required to confirm a mechanism underlying Snn’s potential effect on cell growth, especially analyzing the production and function of those proteins outlined above.

Confirming these results in another cell line more amenable than HUVECs to serum starvation would allow more precise experiments to be performed, as serum starvation halts the growth of many cell types at the G1 checkpoint. Also, analyzing the expression of Snn in hyper-proliferative cells would help to determine the behavior of Snn in potentially pathogenic instances.

Snn is highly-conserved throughout vertebrate evolution, implying an important role for Snn in the vertebrate organism. Modulating the progression of the cell cycle would qualify as an important role, and several previous observations corroborate the data shown in chapter 4. First, Snn is highly expressed during early embryonic development (Dejneka et al., 1997). As an organism develops, many cells are reproducing, differentiating, and dying. Control over the cell cycle is integral in the coordination of these functions. Second, Snn is expressed in only specific tissues in the

99 adult (Dejneka et al., 1997). These tissues include immune cells (T- and B-cells) and the brain (primarily the hippocampus, a known site of neurogenesis in the adult animal).

Vertebrates are the only organisms to have fully functional, complex immune systems and the cells of the immune system are constantly changing as new antigens are presented and new memory and activated T- and B-cells are created (Weil and Israel, 2004). The adult brain is now accepted to be a place of new cell growth over the life cycle of an animal, and memories, in which the hippocampus plays a vital role, are retained in part through the formation of new connections and new cell growth (Schinder and Gage,

2004). The hippocampus is a site of near-constant cellular proliferation (Kemperman et al., 2004;Eriksson et al., 1998). In summary, Snn is expressed in tissues and at times during development that support a role for Snn in modulating cellular proliferation. A model of how stannin may be modulating cell cycle progression is outlined in figure 21.

Though currently unknown, the specific mechanism behind Snn’s role in the cell cycle may be discernable through an observation of TMT toxicity and why it requires

Snn. As described in section 2, Snn may be involved in the metabolism of methyl groups or in the usage thereof. Epigenetic silencing via methylation of DNA has been shown to regulate tumor development (Jubb et al., 2005), initiation of apoptosis (He et al., 2005), and development (Biron et al., 2004). Thus, Snn may affect gene expression through modulation of the methylation state of DNA. Snn’s localization to the mitochondria and other compartments, and not the nucleus, makes it unlikely that Snn is directly methylating genomic DNA, though future studies targeted toward assessing Snn’s intracellular localization in response to TMT exposure would help to clarify how Snn is mediating TMT toxicity. Further, Snn’s interaction with 14-3-3ζ, a scaffolding protein, provides a theoretical mechanism by which Snn could be targeted elsewhere in the cell.

100

Figure 21: Model of normal stannin function. TNFα is known to activate, via

one or both of its receptors, p38. The p38 protein is capable of activating p53. It is

known that p53 is activated by binding to one or more 14-3-3 isoforms, thereby greatly increasing its affinity for specific DNA sequences and initiating translocation of the 14-3-

3/p53 complex to the nucleus. TNFα also activates PKCε resulting in Snn gene expression. Snn has recently been shown to bind to 14-3-3ζ. The competition for 14-3-

3ζ binding may allow Snn to inhibit the activation of p53 and thus promote cell growth or inhibit cell cycle arrest normally mediated by p53.

101

102

More rigorous studies are required to conclude that Snn is in fact modulating methylation

of genomic DNA in PC12 cells or any other cell type. However, TMT toxicity may be

caused through a perturbation of Snn’s normal role in manipulation of intracellular

methyl groups and associated gene expression.

An alternative hypothesis is that Snn-14-3-3ζ binding may result in altered

progression through the cell cycle. The 14-3-3 proteins, and 14-3-3ζ in particular, participate in cell cycle control checkpoints (Ku et al., 2002; Xing et al., 2000).

Specifically, dimerization of 14-3-3 isoform ζ, and either the β or ε isoform, correlate

with progression of HeLa cells through the cell cycle (Alvarez et al., 2003). Thus, an

alternative hypothesis is that Snn competes for 14-3-3ζ binding with other isoforms of

14-3-3 or p53, indirectly inhibiting cell cycle progression. A model of this signaling

perturbation is outlined in figure 22. Additional studies probing the magnitude and

affinity of Snn-14-3-3ζ binding during the cell cycle would help to define the functional

consequence of this interaction in cell cycle regulation.

103

Figure 22: Model of stannin perturbation during trimethyltin exposure. Exposure to

TMT has been shown to result in upregulation of TNFα gene and protein expression.

This upregulation may initiate the sequence of events described in figure x, however, Snn is known to bind to and demethylate TMT. Thus, TMT may be blocking Snn from binding to 14-3-3z, allowing p 53 greater access. Presumably under the stressful condition of TMT exposure, this will result in enhanced levels of p53 binding to 14-3-3z, and thus becoming activated. This activated p53 can then proceed to the nucleus and initiate a signaling cascade culminating in cell death via apoptosis.

104

105

D. Overall Conclusions

We outlined a TNFα-mediated signaling pathway through which Snn gene expression is strongly induced in HUVECs and Jurkat T-cells. Using pharmacological inhibitors and siRNA, we showed that PKCε mediates TNFα-induced Snn gene expression. Our studies show for the first time that TMT administration results in increased Snn gene expression. In addition, we show that TNFα is not likely to be a required component of TMT toxicity. Since Snn is known to be required for TMT toxicity, it is likely that additional signaling molecules are capable of modulating Snn gene expression. Determining which other proteins are capable of this will shed light on what role Snn plays in the cellular milieu. Microarray studies indicated that Snn knockout results in perturbation of normal gene expression changes in HUVECs in response to TNFα. Specifically, several genes important for cell cycle regulation including IL-4, MDM4, and p29 are significantly altered by Snn knockdown prior to

TNFα administration. At this time, the mechanism of these gene expression changes is unknown. The models presented in this chapter explaining how Snn may function both in normal cells, and in the artifical case of TMT exposure, are a starting point for further study. Better characterizing the interactions of Snn with other proteins, such as 14-3-3ζ, will help to illuminate the specific mechanism underlying both TMT toxicity and Snn’s effect on cellular gene expression.

106

References

Adelman DM, Maltepe E, Celeste Simon M (1999). Multilineage embryonic hematopoiesis requires hypoxic ARNT activity. Genes Dev 13, 2478-2483. Akita Y (2002). Protein Kinase C-Epsilon (PKC-epsilon): Its Unique Structure and Function. J Biochem (Tokyo) 132(6), 847-852. Akiyama H, Hiraki Y, Noda M, Shigeno C, Ito H, Nakamura T (1999). Molecular Cloning and Biological Activity of a Novel Ha-Ras Suppressor Gene Predominantly Expressed in Skeletal Muscle, Heart, Brain, and Bone Marrow by Differential Display Using Clonal Mouse EC Cells, ATDC5. J Biol Chem 274(45), 32192-32197. Aksoy E, Goldman M, Willems F (2004). Protein Kinase C Epsilon: A New Target to Control Inflammation and Immune-Related Disorders. Int J Biochem Cell Biol 36(2), 183-188. Aksoy E, Amraoui Z, Goriely S, Goldman M, and Willems F (2002). Critical Role for Protein Kinase C Epsilon for Lipopolysaccharide-Induced IL-12 Synthesis in Monocyte-Derived Dendritic Cells. Eur J Exp Immunol 32, 3040-3049. Ali S, LeBel CP, and Bondy SC (1992). Reactive Oxygen Species Formation as a Biomarker of Methylmercury and Trimethyltin Toxicity. Neurotoxicology 13, 637-648. Alvarez D, Callejo M, Shoucri R, Boyer L, Price GB, Zannis-Hadjopoulos M (2003). Analysis of the Cruciform Binding Activity of Recombinant 14-3-3ζ-MBP Fusion Protein, Its Heterodimerization Profile with Endogenous 14-3-3 Isoforms, and Effect on Mammalian DNA Replication in Vitro. Biochem 42, 7205-7215. Apte RN, Voronov E (2002). Interleukin-1- a major pleiotropic cytokine in tumor- host interactions. Cancer Bio 12, 277-290. Arakawa Y, and Wada O (1993). Biological Properties of Alyltin Compounds. In Metal Ions in Biological Systems. Sigel, H, Sigel, A editor. Marcel Dekker, Inc., New York. 101-136. Arakawa Y, Wada O, and Yu TH (1981). Dealkylation and Distribution of Tin Compounds. Toxicol Appl Pharmacol 60, 1-7.

107

Baines CP, Zhang J, Wang GW, Zheng YT, Xiu JX, Cardwell EM, Bolli R, and Ping P (2002). Mitochondrial PKCε and MAPK Form Signaling Modules in the Murine Heart. Circ. Res. 90, 390-397. Bakalova R, Ohba H, Zhelev Z, Kubo T, Fujii M, Ishikawa M, Shinohara Y, and Baba Y (2004). Atypical Protein-Kinase Cζ, but Neither Conventional Ca2+ - dependent Protein-Kinase C Isoenzymes Nor Ca2+ -calmodulin, Participates in Regulation of Telomerase Activity in Burkitt’s Lymphoma Cells. Cancer Chemother Pharmacol 54, 161-172. Balaban CD, O’Callaghan JP, Billingsley ML (1988). Trimethyltin-Induced NeuronalDamage in the Rat Brain: Comparative Studies Using Silver Degeneration Stains, Immunocytochemistry and Immunoassay for Neuronotypic and Gliotypic Proteins. Neuroscience 26, 337-361. Basu A, Lu D, Sun B, Moor AN, Akkaraju GR, and Huang J (2002). Proteolytic Activation of Protein Kinase C-e by Caspase-Mediated Processing and Transduction of Antiapoptotic Signals. J Biol Chem 277, 41850-41856. Beg AA, and Baltimore D (1996). An Essential Role for NF-κB in Preventing TNF-α-Induce Cell Death. Science 274, 782-784. Berra E, Diaz-Meco MT, Lozano J, Frutos S, Municio MM, Sanchez P, Sanz L, and Moscat J (1995). Evidence for a Role of MEK and MAPK During Signal Transduction by Protein Kinase C ζ. EMBO 14, 6157-6163. Besser R, Kramer G, Thumler R, Bohl J, and Gutman L (1987). Acute Trimethyltin Limbic-Cerebellar Syndrome. Neurology 37, 94550. Beyaert R, Vanhaesebroeck B, Declerq W, Vanlint J, Vandenabeele P, Agostinis P, Vandenheede JR, and Fiers W (1995). Casein Kinase-1 Phosphorylates the p75 Tumor Necrosis Factor Receptor and Negatively Regulates Tumor Necrosis Factor Signaling for Apoptosis. J Biol Chem 270(40), 23293-23299. Biron VL, McManus KJ, Hu N, Hendzel MJ, Underhill DA (2004). Distinct Dynamics and Distribution of Histone Methyl-Lysine Derivatives in Mouse Development. Dev Biol 276(2), 337-351. Brodie ME, Opacka-Juffry J, Peterson DW, and Brown AW (1990). Neurochemical Changes in Hippocampal and Caudate Dialysates Associated with Early Trimethyltin Neurotoxicity in Rats. Neurotoxicology 11, 35-46.

108

Buck BA, Mascioni A, Cramer CJ, and Veglia G (2004). Interactions of Alkyltin Salts with Biological Dithiols: Dealkylation and Induction of a Regular β-Turn Structure in Peptides. J Am. Chem. Soc. 126, 14400-14410. Buck BA, Mascioni A, Que L, and Veglia G (2001). Dealkylation of Organotin Compounds by Biological Dithiols: Toward the Chemistry of Organotin Toxicity. J. Am. Chem. Soc. 125, 13316-13317. Castrillo A, Pennington DJ, Otto F, Parker PJ, Owen MJ, and Lisardo B (2001). Protein Kinase C Epsilon is Required for Macrophage Activation and Defense and Defense Against Bacterial Infection. J Exp Med 194, 1231-1242. Champ MA (2000). A Review of Organotin Regulatory Strategies, Pending Actions, Related Costs, and Benefits. Sci Total Environ 258, 21-71. Chan FKM, Chun HJ, Zheng L, Siegel RM, Bui KL, and Lenardo MJ (2000). A Domain in TNF Receptors That Mediates Ligand-Independent Receptor Assembly and Signaling. Science 288, 2351-2354. Chang MS, Chang CL, Huang CJ, Yang YC (2000). p29, A Novel GCIP- Interacting Protein, Localizes in the Nucleus. Biochem Biophys Res Comm 279, 732-737. Chang LW (1990). The Neurotoxicology and Pathology of Organomercury, Organolead, and Organotin. J Toxicol Sci 15(suppl 4), 125-151. Cheng X, Kinosaki M, Murali R, Greene ML (2003). The TNF Receptor Superfamily: Role in Immune Inflammation and Bone Formation. Immunol Res 27(2-3), 287-294. Chou MM, Hou W, Johnson J, Graham LK, Lee MH, Chen CS, Newton AC, Schaffhausen BS, and Toker A (1998). Regulation of Protein Kinase C ζ by PI 3- Kinase and PDK-1. Curr Biol 8, 1069-1077. Churchill GA (2002). Fundamentals of Experimental Design for cDNA Microarrays. Nature Genetics Suppl 32, 490-495. Comalada M, Xaus J, Valledor AF, Lopez-Lopez C, Pennington DJ, and Celada A (2003). PKCe is Involved in JNK Activation that Mediates LPS-Induced TNF- α, Which Induces Apoptosis in Macrophages. Am J Physiol Cell Physiol 285, C1235- C1245.

109

Cook LL, Heath SM, and O’Callaghan JP (1984). Distribution of Tin in Brain Subcellular Fractions Following the Administration of Trimethyl Tin and Triethyl Tin to the Rat. Toxicol and Appl Pharm 73, 564-568. Cuschieri J, Umanskiy K, and Solomkin J (2004). PKC-ζ Is Essential for Endotoxin-Induced Macrophage Activation. J Surg Res 121, 76-83. Darnay BG, Singh S, and Aggarwal BB (1997). The p80 TNF Receptor- Associated Kinase (p80TRAK) Associates with Residues 354-397 of the p80 Cytoplasmic Domain: Similarity to Casein Kinase. FEBS Lett 406, 101-105. Davidson CE, Reese BE, Billingsley ML, and Yun JK (2004). Stannin, A Protein That Localizes to the Mitochondria and Sensitizes NIH-3T3 Cells to Trimethyltin and Dimethyltin Toxicity. Mol Pharm 66, 855-863. Dejneka NS, Polavarapu R, Deng X, Martin-DeLeon PA, and Billingsley ML (1998). Chromosomal Localization and Characterization of the Stannin (Snn) Gene. Mamm Genome 9, 556-564. Dejneka NS, Patanow CM, Polavarapu R, Toggas SM, Krady JK, and Billingsley ML (1997). Localization and Characterization of Stannin: Relationship to Cellular Sensitivity to Organotin Compounds. Neurochem. Int. 31, 801-815. Deveraux QL, Stennicke HR, Salvesen GS, and Reed JC (1999). Endogenous Inhibitors of Caspases. J Clin Immunol 19, 350-364. Douni E, and Kollias G (1998). A Critical Role of the p75 Tumor Necrosis Factor Receptor (p75TNF-R) in Organ Inflammation Independent of TNF, Lymphotoxin Alpha, or the p55TNF-R. J Exp Med 188(7), 1343-1352. Dyer RS, Howell WE, and Wonderlin WF (1982). Visual System Dysfunction Following Acute Trimethyltin Exposure in Rats. Neurobehav Toxicol Teratol 4, 191- 195. Emoto Y, Kisaki H, Manome Y, Kharbanda S and Kufe D (1996). Activation of Protein Kinase Cdelta in Human Myeloid Leukemia Cell Treated with 1-beta-D- arabinofuranosylcytosine. Blood 87, 1990-1996. Eriksson PS, Perfilieva E, Bjork-Eriksson T, Alborn AM, Nordborg C, Peterson DA, Gage FH (1998). Neurogenesis in the Adult Hippocampus. Nature Med 4(11), 1313-1316.

110

Feldman RG, White RR, and Eriator II (1993). Trimethyltin Encephalopathy. Arch. Neurol 50, 1320-1324. Fent K (1996). Ecotoxicology of Organotin Compounds. Crit Rev Toxicol 26, 1- 117. Frasch SC, Henson PM, Kailet JM, et al. (2000). Regulation of Phospholipid Scramblase Activity During Apoptosis and Cell Activation by Protein Kinase Cδ. J Biol Chem 275, 23065-23073. Fukushima K, Miyamoto S, Tsukimori K, Kobayashi H, Seki H, Takeda S, Kensuke E, Ohtani K, Shibuya M, Nakano H (2005). Tumor Necrosis Factor and Vascular Endothelial Growth Factor Induce Endothelial Integrin Repertories, Regulating Endovascular Differentiation and Apoptosis in a Human Extravillous Trophoblast Cell Line. Biol Reprod Epub ahead of print. Ghayer T, Hugunin M, Talanian RV, et al (1996). Proteolytic Activation of Protein Kinase Cδ by an ICE/CED 3-like Protease Induces Characteristics of Apoptosis. J Exp Med 184, 2399-2404. Grell M, Becke FM, Wajant H, Mannel DN, and Scheurich P (1998). TNF Receptor Tpe 2 Mediates Thymocyte Proliferation Independently of TNF Receptor Type 1. Eur J Immunol 28(1), 257-263. Grell M, Douni E, Wajant H, Lohden M, Clauss M, Maxeiner B, Georgopoulos S, Lesslauer W, Kollias G, Pfizenmaier K, and Scheurich P (1995). The Transmembrane Form of Tumor Necrosis Factor is the Prime Activating Ligand of the 80 kDa Tumor Necrosis Factor Receptor. Cell 83, 793-802. Grell M, Zimmermann G, Hulser D, Pfizenmaier K, and Scheurich P (1994). TNF Receptors TR60 and TR80 Can Mediate Apoptosis Via Induction of Distinct Signal Pathways. J Immunol 153, 1963-1972. Gschwendt M (1999). Protein Kinase C δ. Eur J Biochem 259, 555-564. Gunasekar P, Li L, Prabhakaran K, Eybl V, Borowitz JL, and Isom GE (2001). Mechanisms of the Apoptotic and Necrotic Actions of Trimethyltin in Cerebellar Granule Cells. Toxicol Sci 64, 83-89. Gunasekar PG, Mickova V, Kotyzova D, Li L, Borowitz JL, Eybl V, Isom GE (2001) Role of Astrocytes in Trimethyltin Neurotoxicity. J Biochem Mol Tox 15:256-262.

111

Gupta S (2001). A Decision Between Life and Death During TNF-α-Induced Signaling. J Clin Immunol 22(4), 185-194. Han HJ, Lim MJ, Lee YJ (2004). Oxalate inhibits renal proximal tubule cell

proliferation via oxidative stress, p38 MAPK/JNK, and cPLA2 signaling pathways. Am J Physiol Cell Physiol 287, C1058-C1066. Harry JG, Bruccoleri A, d’Hellencourt LC (2003). Differential Modulation of Hippocampal Chemical-Induced Injury Response by Ebselen, Pentoxifylline, and TNFalpha-, IL-1alpha, and IL-6-Neutralizing Antibodies. J Neurosci Res 73(4), 526 -536. He B, Lee AY, Dadfarmay S, You L, Xu Z, Reguart N, Mazieres J, Mikami I, McCormick F, Jablons DM (2005). Secreted Frizzled-Related Protein 4 is Silenced by Hypermethylation and Induces Apoptosis in Beta-Catenin-Deficient Human Mesothelioma Cells. Cancer Res 65(3), 743-748. Henderson BR, and Kuhn, LC (1997). Interaction Between Iron-Regulatory Proteins and Their RNA Target Sequences, Iron-Responsive Elements. Prog Mol Subcell Biol. 18, 117-139. Hioe KM, and Jones JM (1984). Effects of Trimethyltin on the Immune System of Rats. Tox Lett 20(3), 317-323. Horrevoets AJG, Fontijn RD, Jan van Zonneveld A, de Vries CJM, Wouter ten Cate J, Pannekoek H (1999). Vascular Endothelial Genes That Are Responsive to Tumor Necrosis Factor-α In Vitro Are Expressed in Atherosclerotic Lesions, Including Inhibitor of Apoptosis Protein-1, Stannin, and Two Novel Genes. Blood 10, 3418- 3431. Ichijo N, Nishida E, Irie K, Ten Dijke P, Saitoh M, Moriguchi T, Tagaki M, Matsumoto K, Miyazono K, and Gotoh Y (1997). Induction of Apoptosis by ASK1, a Mammalian MAPKKK That Activates SAPK/JNK and p38 Signaling Pathways. Science 275, 90-94. Idriss HT, and Naismith JH (2000). TNFa and the TNFa Receptor Superfamily: Structure Function Relationships. Microsc Res Tech 50, 184-195. Irie N, Sakai N, Ueyama T, Kajimoto T, Shirai Y, Saito N (2002). Subtype- and Species-Specific Knockdown of PKC Using Short Interfering RNA. Biochem Biophys Res Commun 298, 738-743.

112

Jin Z, Pei L, Wang D (2000). Effects of tumor necrosis factor-alpha on proliferation and intracellular free calcium concentration in endothelium of human umbilical vein endothelial cells. Zhonghua Fu Chan Ke Za Zhi 35, 657-659. Johnstone RW, See RH, Sells SF, Wang J, Muthukkumar S, Englert C, Haber DA, Licht JD, Sugrue SP, Roberts T, Rangnekar VM, Shi Y (1996). A Novel Repressor, par-4, Modulates Transcription and Growth Suppresion Functions of the Wilms’ Tumor Suppressor WT1. Mol Cell Biol 16(12), 6945-6956. Jubb AM, Bell SM, Quirke P (2005). Methylation and Colorectal Cancer. J Pathol 195(1), 111-134. Jung YS, Ryu BR, Lee BK, Mook-Jung I, Kim SU, Lee SH, Baik EJ, and Moon CH (2004). Role for PKC-ε in Neuronal Death Induced by Oxidative Stress. Biochem Biophys Res Comm 320, 789-794. Karin M, Lin A (2002). NF-kB at the Crossroads of Life and Death. Nature Immunol 3, 221-227. Kaneda A, Wakazono K, Tsukamoto T, Watanabe N, Yagi Y, Tatematsu M, Kaminishi M, Sugimura T, Ushijima T (2004). Lysyl Oxidase is a Tumor Suppresor Gene Inactivated by Methylation and Loss of Heterozygosity in Human Gastric Cancers. Cancer Res 64(18), 6410-6415. Kempermann G, Jessberger S, Steiner B, Kronenberg G (2004). Milestones of Neuronal Development in the Adult Hippocampus. TRENDS Neurosci 27(8), 447-452. Kilpatrick LE, Lee JY, Haines KM, Campbell DE, Sullivan KE, and Korchak HM (2002). A Role for PKC-δ and PI 3-kinase in TNF-α-Mediated Antiapoptotic Signaling in the Human Neutrophil. Am J Physiol Cell Physiol 283, C48- C57. Kim J, Cheon IS, Won YJ, Na HJ, Kim; YM, Choe J (2003). IL-4 Inhibits Cell Cycle Progression of Human Umbilical Vein Endothelial Cells by Affecting p53, p21WAF1, Cyclin D1, and Cyclin E Expression. Mol Cells 16(1), 92-96. Klein CL, Kohler H, Bittinger F, Wagner M, Hermanns I, Grant K, Lewis JC, Kirkpatrick CJ (1994). Comaprative Studies on Vascular Endothelium In Vitro. I. Cytokine Effects on the Expression of Adhesion Molecules by Human Umbilical Vein, Saphenous Vein, and Femoral Artery Endothelial Cells. Pathobiology 62(4), 199-208.

113

Krady JK, Oyler GA, Balaban CD, Billingsley ML (1990). Use of avidin- biotin subtractive hybridization to characterize mRNA common to neurons destroyed by the selective neurotoxicant trimethyltin. Brain Res Mol Brain Res 7, 287-297. Krappmann D, Patke A, Heissmeyer V, and Scheidereit C (2001). B-Cell Receptor- and Phorbol Ester-Induced NF-kappaB and c-Jun N-Terminal Kinase Activation in B Cells Requires Novel Protein Kinase C’s. Mol Cell Biol 21, 6640-6650. Kreuz S, Siegmund D, Rumpf JJ, Samel D, Leverkus M, Janssen O, Hacker G, Dittriclh A, Breiholz O, Kracht M, Scheurich P and Wajant H (2004). NfkappaB Activation by Fas is Mediated Through FADD, Caspase-8, and RIP and is Inhibited by FLIP. J Cell Biol 166(3), 369-380. Ku NO, Michie S, Resurreccion EZ, Broome RL, Omary MB (2002). Keratin Binding to 14-3-3 Proteins Modulates Keratin Filaments and Hepatocyte Mitotic Progression. PNAS 99(7), 4373-4378. Kwak E, Larochelle DA, Beaumont C, Torti SV, Torti FM (1995). Role for NF-κB in the regulation of ferritin H by tumor necrosis factor-α. J Bio Chem 270, 15285-15293. Laferriere J, Houle F, Huot J (2004). Adhesion of HT-29 Colon Carcinoma Cells to Endothelial Cells Endothelial requires Sequential Events Involving E-Selctin and Integrin β4. Clin & Exp Metastasis 21, 257-264. Lallena MJ, Diaz-Meco MT, Bren G, Paya CV, and Moscat J (1999). Activation of IκB Kinase β by Protein Kinase C Isoforms. Mol Cell Biol 19, 2180-2188. Lang W, Wang H, Ding L, and Xiao L (2004). Cooperation Between PKC- Alpha and PKC-Epsilon in the Regulation of JNK Activation in Human Lung Cancer Cells. Cell Signal 16(4), 457-467. Langeggen H, Johnson E, Hetland G (2001). Effects of C5a and FMLP on Interleukin-8 Production and Proliferation of Human Umbilical Vein Endothelial Cells. Inflammation 25(2), 83-89. Leeuwenberg JFM, van Tits LJH, Jeunhomme TMAA, Buurman WA (1995). Evidence for Exclusive Role in Signaling of Tumor Necrosis Factor p55 Receptor and a Potentiating Function of p75 Receptor on Human Endothelial Cells. Cytokine 7, 457-462.

114

Leitges M, Sanz L, Martin P, Duran A, Braun U, Garcia JF, Camancho F, Diaz-Meco MT, Rennert PD, and Moscat J (2001). Targeted Disruption of the ζPKC Gene Results in the Impairment of the NF-κB Pathway. Mol Cell 8, 771-780. Leung T, Chen XQ, Tan I, Manser E, Lim L (1998). Myotonic Dystrophy Kinase-Related Cdc42-Binding Kinase Acts as a Cdc42 Effector in Promoting Cytoskeletal Reorganization. Mol Cell Biol 18(1), 130-140. Li L, Lorenzo PS, Bogi K, Blumberg PM, and Yuspa SH (1999). Protein Kinase Cδ Targets Mitochondria, Alters Mitochondrial Membrane Potential, and Induces Apoptosis in Normal and Neoplastic Keratinocytes When Overexpressed by an Adenoviral Vector. Mol Cell Biol 19, 8547-8558. Limatola, C, Schaap, D, Moolenaar, WH, and van Blitterswijk, WJ (1994). Phosphatidic Acid Activation of Protein Kinase C-ζ Overexpressed in COS Cells: Comparison with Other Protein KinaseC Isotypes and Other Acidic Lipids. Biochem J 304, 1001-1008. Lin D, Edwards AS, Fawcett JP, Mbamalu G, Scott JD, and Pawson T (2000). A Mammalian PAR-3-PAR-6 Complex Implicated in Cdc42/Rac1 and aPKC Signaling and Cell Polarity. Nat Cell Biol 2, 540-547. Lin PS, Ho KC, Sung SJ, Gladding J (1992). Effect of tumor necrosis factor, heat, and radiation on the viability and microfilament organization in cultured endothelial cells. Int J Hyperthermia 8, 667-677. Ling L, Cao Z, and Goeddel DV (1998). NF-κB-Inducing Kinase Activates IKK-α by Phosphorylation of Ser-176. Proc Natl Acad Sci 95, 3792-3797. Liu L, Vo A, McKeehan WL (2005). Specificity of the Methylation-Suppressed A Isoform of Candidate Tumor Suppressor RASSF1 for Microtubule Hyperstabilization is Determined by Cell Death Inducer C19ORF5. Cancer Res 65(5), 1830-1838. Liu DX, Nath N, Chellappan SP, Greene LA (2005). Regulation of Neuronal Survival and Death by p130 and Associated Chromatin Modifiers. Gen & Dev 19, 719- 732. Liu Y, Imai H, Sadamatsu M, Tsunashima K, and Kato N (2004). Cytokines Participate in Neuronal Death Induced by Trimethyltin in the Rat Hippocampus via Type II Glucocorticoid receptors. Neuro Res 50, 209-217.

115

Liu ZG, Hsu H, Goeddel DV, and Karin M (1996). Dissection of TNF Receptor 1 Functions: JNK Activation is Not Linked to Apoptosis While NF-kappa-B Activation Prevents Cell Death. Cell 87, 565-576. Locksley RM, Killeen N, and Lenardo MJ (2001). The TNF and TNF Receptor Superfamilies: Integrating Mammalian Biology. Cell 104, 487-501. Long WP, Chen X, Perdew GH (1999). Protein kinase C modulates aryl hydrocarbon receptor nuclear translocator protein-mediated transactivation potential in a dimmer context. J Biol Chem 274, 12391-12400. Lucas R, Garcia I, Donati YR, Hribar M, Mandriota SJ, Giroud C, Buurman WA, Fransen L, Suter PM, Nunez G, Pepper MS, and Grau GE (1998). Both TNF Receptors are Required for Direct TNF-Mediated Cytotoxicity in Microvascular Endothelial Cells. Eur J Immunol 28(11), 3577-3586. MacEwan DJ (2002). TNF Receptor Subtype Signalling: Differences and Cellular Consequences. Cell Signalling 14, 477-492. Maheswaran S, Englert C, Bennett P, Heinrich G, Haber AD (1995). The WT1 Gene Product Stabilizes p53 and Inhibits p53-Mediated Apoptosis. Genes Dev 9, 2143-2156. Maheswaran S, Park S, Bernard A, Morris JF, Rauscher FJ, Hill DE, Haber DA (1993). Physical and Functional Interaction Between WT1 and p53 Proteins. Proc Natl Acad Sci 90, 5100-5104. Majumder PK, Mishra NC, Sun X, et al. (2001). Targeting of Protein Kinase Cδ to Mitochondria in the Oxidative Stress Response. Cell Growth Differ 129, 465-470. Mauerhoff T, Belfiore A, Pujol-Borrell R, Bottazzo GF (1994). Growth inhibition of human endothelial cells by human recombinant tumor necrosis factor alpha and interferon-gamma. Tumori 80, 301-305. Mcjilton MA, Sikes CV, Wescott GG, Wu D, Foreman TL, Gregory CW, Weidner DA, Ford OH, Lasater AM, Mohler JL, and Terrian DM (2003). Protein Kinase Cε Interacts with Bax and Promotes Survival of Human Prostate Cancer Cells. Oncogene 22, 7958-7968. Mehta D, Rahman A, and Malik AB (2001). Protein Kinase C-a Signals Rho- Guanine Nucleotide Dissociation Inhibitor Phosphoryation and Rho Activation and Regulates the Endothelial Cell Barrier Function. J Biol Chem 276, 22614-22620.

116

Mendez R, Hake LE, Andresson T, Littlepage LE, Ruderman JV, Richter JD (2000). Phosphorylation of CPE binding factor by Eg2 regulates translation of c-mos mRNA. Nature (Lond) 404, 302-307. Migliorini D, Denchi EL, Danovi D, Jochemsen, Capillo M, Gobbi A, Helin K, Pelicci PG, Marine JC (2002). Mdm4 (Mdmx) Regulates p53-Induced Growth Arrest and Neuronal Cell Death During Early Embryonic Mouse Development. Mol Cell Biol 22(15), 5527-5538. Moldenhauer A, Nociari M, Lam G, Salama A, Rafu S, Moore MAS (2004). Tumor Necrosis Factor Alpha-Stimulated Endothelium: An Inducer of Dendritic Cell Development from Hematopoietic Progenitors and Myeloid Leukemic Cells. Stem Cells 22, 144-157. Monick MM, Carter AB, Flaherty DM, Peterson MW, and Hunninghake GW (2000). Protein Kinase C ζ Plays a Central Role in Activation of the p42/44 Mitogen-Activated Protein Kinase by Endotoxin in Alveolar Macrophages. J Immunol 165, 4632-4639. Muller G, Ayoub M, Storz P, Rennecke J, Fabbro D, and Pfizenmaier K (1995). PKCζ is a Molecular Switch in Signal Transduction of TNF-α, Bifunctionally Regulated by Ceramide and Arachidonic Acid. EMBO 14, 1961-1969. Munirathinam S, Rogers G, and Bahr BA (2002). Positive Modulation of α- Amino-3-Hydroxy-5-Methyl-4-Isoxazolepropionic Acid-Type Glutamate Receptors Elicits Neuroprotection After Trimethyltin Exposure in Hippocampus. Toxicol Appl Pharmacol 185, 111-118. Mushak P, Krigman MR, and Mailman RB (1982). Comparative Organotin Toxicity in the Developing Rat: Somatic and Morphological Changes and Relationship to Accumulation of Total Tin. Neurobehav Toxicol Teratol 4, 209-215. Nagai F, Hiyoshi Y, Sugimachi K, Tamura H (2002). Cytochrome P450 (CYP) Expression in Human Myeloblastic and Lymphoid Cell Lines. Biol Pharm Bull 25(3), 383-385. Nakanishi H, Brewer KA, and Exton JH (1993). Activation of the ζ Isozyme of Protien Kinase C by Phosphatidylinositol 3,4,5-Triphosphate. J Biol Chem 268, 13- 16.

117

Natoli G, Costanzo A, Ianni A, Templeton DJ, Woodgett JR, Balsano C and Levrero M (1997). Activation of SAPK/JNK by TNF Receptor 1 Through a Noncytotoxic TRAF2-Dependent Pathway. Science 275, 200-203. Nishimura T, Schwarzer C, Furtinger S, Imai H, Kato N, and Sperk G (2001). Changes in the GABA-ergic System Induced by Trimethyltin Application in the Rat. Mol Brain Res 97, 1-6. Noland EA, Taylor DH, and Bull RJ (1982). Monomethyl-and Trimethyltin Compounds Induce Learning Deficiencies in Young Rats. Neurobehav Toxicol Teratol 4, 539-544. Ohazama A, Sharpe PT (2004). TNF Signalling in Tooth Development. Curr Opin Genet Dev 14(5), 513-519. Old LJ (1985). Tumor Necrosis Factor (TNF). Science 230, 630-632. Ostareck-Ledrer A, Ostareck DH, Standard N, Thiele BJ (1994). Translation of 15-lipoxygenase mRNA is inhibited by a protein that binds to a repeated sequence in the3’ untranslated region. EMBO (Eur Mol Biol Organ) 13, 1476-1481. Pan W, Zadina JE, Harlan RE, Weber JT, Banks WA, Kastin AJ (1997). Tumor necrosis factor-alpha: a neuromodulator in the CNS. Neurosci Biobehav Rev 21, 603-613. Patanow CM, Day JR, and Billingsley ML (1997). Alterations in Hippocampal Expression of SNAP-25, GAP-43, Stannin, and Glial Fibrillary Acidic Protein Following Mechanical and Trimethyltin-Induced Injury in the Rat. Neuroscience 76(1), 187-202. Patel MN, Ardelt BK, Yim GKW, and Isom GE (1990). Interaction of Trimethyltin with Hippocampal Glutamate. Neurotoxicology 11, 601-608. Pavlaković G, Kane MD, Eyer CL, Kanthasamy A, and Isom GE (1995). Activation of Protein Kinase C by Trimethyltin: Relevance to Neurotoxicity. J Neurochem 65, 2338-2343. Petrovics G, Bird T, Lehel C, Oravecz T, and Anderson WB (2001). Protein Kinase Cε Mediates PMA-Induced Growth Inhibition of Low Population Density NIH 3T3 Fibroblasts. Arch Biochem Biophys 397 (2), 217-223. Philbert MA, Billingsley ML, Reuhl KR (2000). Mechanisms of Injury in the Central Nervous System. Tox Path 28, 43-53.

118

Pimentel-Muinos FX, and Seed B (1999). Regulated Commitment of TNF Receptor Signaling: A Molecular Switch for Death or Activation. Immunity 11, 783-793. Poole AW, Pula G, Hers I, Crosby D, and Jones ML (2004). PKC-Interacting Proteins: From Function to Pharmacology. TRENDS in Pharm Sci 25(10), 528-535. Posnick LM, Samson LD (1999). Influence of S-adenosylmethionine Pool Size on Spontaneous Mutation, Dam Methylation, and Cell Growth of Escherichia coli. J Bacteriol 181(21), 6756-6762. Rao VU, Shiraishi H, and McDermott PJ (2003). PKC-e Regulation of Extracellular Signal-Related Kinase; A Potential Role in Phenylephrine-Induced Cardiocyte Growth. Am J Physiol Circ Physiol 286, H2195-H2203. Reese BE, Davidson CE, Billingsley ML, Yun JK (2005). Protien Kinase C Epsilon Regulates TNFalpha-Induced Stannin Gene expression. J Pharmacol Exp Ther Epub ahead of print. Robertson DG, Kim SN, Gray RH, and de la Iglesia FA (1987). The Pathogenesis of Trimethyltin Chloride-Induced Nephrotoxicity. Fundam Appl Toxicol 8(2), 147-158. Rocken M, Rocke M, Shevach EM (1996). IL-4-Induced Immune Deviation as Antigen-Specific Therapy for Inflammatory Autoimmune Disease. Immunol Today 17, 225-231. Ross D, and Joyner WL (1997). Resting Distribution and Stimulated Translocation of Protein Kinase C Isoforms Alpha, Epsilon, and Zeta in Response to Bradykinin and TNF in Human Endothelial Cells. Endothelium 5(4), 321-332. Rothe M, Pan MG, Henzel WJ, Ayres TM, and Goeddel DV (1995). The TNFR2-TRAF Signaling Complex Contains Two Novel Proteins Related to Baculoviral Inhibitor of Apoptosis Proteins. Cell 83, 1243-1252. Rothe M, Wong SC, Henzel WJ, and Goeddel DV (1994). A Novel Family of Putative Signal Transducers Associated with the Cytoplasmic Domain of the 75kDa Tumor Necrosis Factor Receptor. Cell 78(4), 681-692. Schievella AR, Chen JH, Graham JR, and Lin LL (1997). MADD, a Novel Death Domain Protein That Interacts with the Type 1 Tumor Necrosis Factor Receptor and Activates Mitogen-Activated Protein Kinase. J Biol Chem 272, 12069-12075. Schinder AF, Gage FH (2004). A Hypothesis About the Role of Adult Neurogenesis in Hippocampal Function. Physiology Oct 19, 253-261.

119

Six DA, Dennis EA (2000). the Expanding Superfamily of Phospholiase A2 Enzymes: Classification and Characterization. Biochim Biophys Acta 1488, 1-19. Soh JW, and Weinstein IB (2003). Roles of Specific Isoforms of Protein Kinase C in the Transcriptional Control of Cyclin D1 and Related Genes. J Biol Chem 278(36), 34709-34716. Spitaler M, and Cantrell DA (2004). Protein Kinase C and beyond. Nat Immunology 5(8), 785-790. Stanger BZ, Leder P, Lee TH, Kim E, and Seed B (1995). RIP: a Novel Protein Containing a Death Domain The Interacts with FAS/APO-1 (CD95) in Yeast and Causes Cell Death. Cell 81, 513-523. Suzuki A, Yamanaka T, Hirose T, Manabe N, Mizuno K, Shimizu M, Akimoto K, Izumi Y, Ohnishi T, and Ohno S (2001). Atypical Protien Kinase C is Involved in the Evolutionarily Conserved PAR Protein Complex and Plays a Critical Role in Establishing Epithelia-Specific Junctional Structures. J Cell Biol 152, 1183- 1196. Swartzwelder HS, Hepler J, Holahan W, King SE, Leverenz HA, Miller PA, and Myers RD (1982). Impaired Maze Performance in the Rat Caused by Trimethyltin Treatment: Problem-Solving Deficits and Perseveration. Neurobehav Toxicol Teratol 4, 169-176. Tartaglia L, Pennica D, and Goeddel DV (1993). Ligand Passing: The 75-kDa Tumor Necrosis Factor (TNF) Receptor Recruits TNF for Signaling by the p55-kDa TNF Receptor. J Biol Chem 268, 18542-18548. ten Kate M, Hofland LJ, van Grevenstein WM, van Koetsveld PV, Jeekel J, van Eijck CH (2004). Influence of Proinflammatory Cytokines on the Adhesion of Human Colon Carcinoma Cells to Lung Microvascular Endothelium. Int J Cancer 112(6), 943-950. Thomson AM, Roger JT, Leedman PJ (1999). Iron-Regulatory Proteins, Iron- Responsive Elements and Ferritin mRNA Translation. Int J Biochem Cell Biol. 31(1): 1139-1152. Thompson TA, Lewis JM, Dejneka NS, Severs WB, Polavarapu R, Billingsley ML (1996). Induction of Apoptosis by Organotin Compounds In Vitro: Neuronal Protection with Antisense Oligonucleotides Directed Against Stannin. J Pharmacol Exp Ther 276:1201-1215.

120

Thomson AM, Rogers JT, Leedman PJ (2000) Thyrotropin-releasing hormone and epidermal growth factor regulate iron-regulatory protein binding in pituitary cells via proteinkinase C-dependent and –independent signaling pathways. J Biol Chem 275, 31609-31615. Toder V, Carp H, Fein A, Torchinsky A (2002). The Role of Pro- and Anti- Apoptotic Molecular Interactions in Embryonic Maldevelopment. Am J Reprod Immunol 48(4), 235-244. Toggas SM, Krady JK, and Billingsley ML (1992). Molecular Neurotoxicology of Trimethyltin: Identification of Stannin, a Novel Protein Expressed in Trimethyltin-Sensitive Cells. Mol Pharm 42, 44-56. Tracey KJ, and Cerami A (1993). Tumor Necrosis Factor, Other Cytokines and Disease. Ann Rev Cell Biol 9, 317-343. Urtreger AJ, Grossoni VC, Falbo KB, Kazanietz MG, and Bal de Kier Jaffe ED (2005). Atypical Protein Kinase C-ζ Modulates Clonogenicity, Motility, and Secretion of Proteolytic Enzymes in Murine Mammary Cells. Mol Carcinogenesis 42, 29-39. Viviani B, Corsini E, Pesenti M, Galli CL, Marinovich M (2003). Trimethyltin-Activated Cyclooxygenase Stimulates Tumor Necrosis Factor-Alpha Release from Glial Cells Through Reactive Oxygen Species. Toxicol Appl Pharmacol 172(2), 93-97. Viviani B, Corrado GL, Marinovich M (1998). Trimethyltin But Not Triethyltin Induces Specific Neual Death Through the Protein Stannin. Neurosci Res Comm 23(3), 139-149. Walsh TJ, McLamb RL, and Tilson HA (1984). Organometal-Induced Antinociception: a Time- and Dose-Response Comparison of Triethyl and Trimethyl lead and tin. Toxicol Appl Pharmacol 73, 295-299. Walsh TJ, Miller DB, and Dyer RS (1982). Trimethyltin, A Selective Limbic System Neurotoxicant, Impairs Radial-Arm Maze Performance. Neurobehav Toxicol Teratol 4, 177-183. Weil R, Israel A (2004). T-Cell Receptor- and B-Cell-Receptor-Mediated Activation of NF-kB in Lymphocytes. Curr Opin Immunol 16, 374-381.

121

Wenger GR, McMillan DE, Chang LW, Zitaglio T, and Hardwick W (1985). The Effects of Triethyltin and Trimethyltin in Rats Responding Under a DRL Schedule of Reinforcement. Toxicol Appl Pharmacol 78, 248-258. Wu D, Thakore CU, Wescott GG, McCubrey JA, and Terrian DM (2004). Integrin Signaling Links Protein Kinase Ce to the Protein Kinase B/Akt Survival Pathway in Recurrent Prostate Cancer Cells. Oncogene 23, 8659-8672. Xiao H, Goldthwait DA, and Mapstone T (1994). The Identification of Four Protein Kinase C Isoforms in Human Glioblastoma Cell Lines: PKC Alpha, Gamma, Epsilon, and Zeta. J Neurosurg 81(5), 734-740. Xiong S, She H, Takeuchi H, Han B, Engelhardt JF, Barton CH, Zandi E, Giuliviv C, Tsukamoto H (2003). Signaling role of intracellular iron in NF-kB activation. J Biol Chem 278, 17646-17654. Yang H, Masters SC, Wang H, Fu H (2001). The Proapoptotic Protein Bad Binds the Amphipathic Groove of 14-3-3ζ. Biochim Biophys Acta 1547, 313-319. Yeh WC, Luis de la Pompa J, McCurrach ME, Shu HB, Elia AJ, Shahinian A, Ng M, Wakeham A, Khoo W, Mitchell K, El-Deiry WS, Lowe SW, Goeddel DV, Mak TW (1998). FADD: Essential for Embryo Development and Signaling from Some, But Not All, Inducers of Apoptosis. Science (Wash DC) 279, 1954-1958. Yoshida K, Wang HG, Miki Y, Kufe D (2003). Protein Kinase Cδ is Responsible for Constitutive and DNA Damage-Induced Phosphorylation of Rad9. EMBO (Eur MolBiol Organ) 22(6), 1431-1441. Yoshizumi M, Fujita Y, Izawa Y, Suzaki Y, Kyaw M, Ali N, Tsuchiya K, Kagami S,Yano S, Sone S, Tamaki T (2004). Ebselen Inhibits Tumor Necrosis Factor- α-Induced c-Jun N-Terminal Kinase Activation and Adhesion Molecule Expression in Endothelial Cells. Exp Cell Res 292, 1-10. Zemskov EA, Jana NR, Kurosawa M, Miyazaki H, Sakamoto N, Nekooki M, and Nukina N (2003). Pro-apoptotic Protein Kinase Cδ is associated with Intranuclear Inclusions in a Transgenic Model of Huntington’s Disease. J Neurochem 87, 395-406. Zhang BH, Chen H, Yao XP, Cong WM, Wu MC (2002). E-Selectin and Its Ligand-sLeX in the Metastasis of Hepatocellular Carcinoma. Hepatobiliary Pancreat Dis Int 1(1), 80-82.

122

Appendix

Outlined in the tables on the following pages is every gene that was deemed

“significantly altered” according to the analysis outlined in chapter 4. The numbers for expression level in each table are in terms of the control condition (i.e. 2.00 means that condition expressed twice as much of that gene as the control condition did).

123

Cell Growth and Maintenance Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group ATPase, Na+/K+ NM_001677 0.82 2.06 transporting, beta 1 polypeptide ATP-binding cassette, NM_003742 0.98 3.48 sub-family B member 11 ATPase, Cu++ NM_000052 1.14 3.34 transporting, alpha polypeptide

ATP synthase, H+ NM_004889 1.90 0.70 transporting, mitochondrial F0 complex

CDC42 binding protein NM_014826 0.41 0.137 kinase alpha (DMPK- like) chromosome 22 open NM_002837 0.89 2.66 reading frame 1 chromosome NM_022346 1.28 0.38 condensation protein G component of NM_007357 0.77 1.99 oligomeric golgi complex 2 dachshund homolog NM_004392 1.06 3.97 (Drosophila) estrogen receptor NM_004215 0.21 0.63 binding site associated, antigen, 9 exostoses (multiple) 1 NM_000127 0.32 0.89

exportin 7 NM_015024 0.87 2.63 gamma-aminobutyric NM_000815 0.71 0.284 acid (GABA) A receptor, delta GCIP-interacting NM_015484 1.09 2.79 protein p29 HRAS-like suppressor NM_020386 1.48 3.94 interferon, omega 1 NM_002177 0.90 2.31 interleukin 4 NM_000589 2.68 10.02 major intrinsic protein NM_012064 0.66 1.72 of lens fiber MAX binding protein NM_020310 1.18 0.42 Mdm4, transformed NM_002393 0.73 3.95 3T3 cell double minute 4, p53 binding protein neuroglobin NM_021257 2.11 0.70 nuclear receptor NM_014071 1.22 4.45 coactivator 6 origin recognition NM_002553 1.48 4.13 complex, subunit 5-like oxysterol binding NM_024586 2.24 0.86 protein-like 9 124

Cell Growth and Maintenance Genes – cont’d

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group phospholipase A2, NM_005084 1.83 9.58 group VII POP4 (processing of NM_006627 2.11 6.50 precursor , S. cerevisiae) homolog

PRKC, apoptosis, WT1, NM_002583 1.04 3.03 regulator solute carrier family 38, NM_018018 0.47 1.31 member 4 solute carrier organic NM_013272 0.41 1.21 anion transporter family, member 3A1 RNA binding motif NM_016090 0.68 1.98 protein 7 transient receptor NM_021625 1.52 0.59 potential cation channel, subfamily V, member 4

125

Signal Transduction Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group CDC42 binding protein NM_014826 0.41 0.14 kinase alpha (DMPK- like)

exostoses (multiple) 1 NM_000127 0.32 0.89

gamma-aminobutyric NM_000815 0.71 0.29 acid (GABA) A receptor, delta GDNF family receptor NM_022139 0.98 0.38 alpha 4 guanine nucleotide NM_020988 0.56 1.49 binding protein (G protein), alpha activating activity polypeptide O

immunoglobulin NM_005849 1.28 3.28 superfamily, member 6

natural cytotoxicity NM_004828 1.32 4.54 triggering receptor 2

nuclear receptor NM_014071 1.23 4.45 coactivator 6

PDZ domain containing AL117397 1.33 3.47 guanine nucleotide exchange factor (GEF) 1 PHD finger protein 3 NM_015153 0.98 2.65

phosphodiesterase 6A, NM_000440 0.72 2.22 cGMP-specific, rod, alpha

phosphodiesterase 9A NM_002606 1.54 3.86 platelet factor 4 NM_002619 0.74 2.70 (chemokine (C-X-C motif) ligand 4) Ras homolog enriched NM_005614 0.77 1.98 in brain ring finger protein 110 NM_007144 2.61 0.83

secretagogin, EF-hand NM_006998 0.83 2.33 calcium binding protein supervillin NM_021738 1.26 4.49 xenotropic and NM_004736 1.26 3.66 polytropic retrovirus receptor

126

Nucleobase, Nucleoside, Nucleotide, and Nucleic Acid Metabolism

Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group 2',5'-oligoadenylate NM_002534 0.75 0.27 synthetase 1, 40/46kDa aldehyde dehydrogenase NM_005589 0.81 2.01 6 family, member A1

armadillo repeat NM_018076 0.82 2.75 containing 4

chromosome 22 open NM_002837 0.89 2.66 reading frame 1

D4, zinc and double NM_012074 0.69 0.26 PHD fingers, family 3 dachshund homolog NM_004392 1.06 3.97 MADS box transcription NM_002397 1.11 3.93 enhancer factor 2, polypeptide C (myocyte enhancer factor 2C)

MAX binding protein NM_020310 1.18 0.42 nuclear receptor NM_014071 1.23 4.45 coactivator 6 nuclear protein, ataxia- U58852 0.79 3.54 telangiectasia locus origin recognition NM_002553 1.47 4.13 complex, subunit 5-like polymerase (RNA) III NM_007055 0.39 0.99 (DNA directed) (155kD)

POP4 (processing of NM_006627 2.11 6.51 precursor , S. cerevisiae) homolog

PRKC, apoptosis, WT1, NM_002583 1.04 3.04 regulator ribosomal protein L6 NM_000970 1.39 0.40 thioredoxin domain NM_016616 0.83 2.18 containing 3 (spermatozoa)

zinc finger protein 215 NM_013250 1.41 3.63

127

Protein Metabolism Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group component of NM_007357 0.76 1.99 oligomeric golgi complex 2 eukaryotic translation NM_001958 2.41 0.68 elongation factor 1 alpha 2

exportin 7 NM_015024 0.87 2.63 gamma-glutamyl NM_000821 0.56 1.93 carboxylase kallikrein 15 NM_017509 0.98 3.07 mannosyl (alpha-1,6-)- NM_002410 0.69 0.27 glycoprotein beta-1,6-N- acetyl- glucosaminyltransferase mitochondrial ribosomal NM_016050 2.56 0.94 protein L11 occludin NM_002538 0.74 2.36 protein tyrosine NM_002837 0.89 2.66 phosphatase, receptor type, B ribosomal protein L6 NM_000970 1.39 0.40 STE20-like kinase NM_016281 2.02 6.09 SUMO1/sentrin specific NM_014554 1.91 4.89 protease 1

128

Biosynthesis Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group 1-acylglycerol-3- NM_032741 1.21 0.43 phosphate O- acyltransferase 1 (lysophosphatidic acid acyltransferase, alpha)

5- NM_024010 1.28 3.47 methyltetrahydrofolate- homocysteine methyltransferase reductase argininosuccinate NM_054012 0.13 0.62 synthetase

component of NM_007357 0.77 1.99 oligomeric golgi complex 2 eukaryotic translation NM_001958 2.40 0.68 elongation factor 1 alpha 2 exostoses (multiple) 1 NM_000127 0.32 0.89 histidine decarboxylase NM_002112 1.25 0.42

mannosyl (alpha-1,6-)- NM_002410 0.69 0.27 glycoprotein beta-1,6-N- acetyl- glucosaminyltransferase mitochondrial ribosomal NM_016050 2.56 0.94 protein L11 ribosomal protein L6 NM_000970 1.39 0.40

129

Cell-Cell Signaling Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group adhesion regulating NM_007002 0.89 4.00 molecule 1 catenin (cadherin- NM_001903 2.70 0.74 associated protein), alpha 1, 102kDa discs, large homolog 7 NM_014750 2.31 9.97 gamma-aminobutyric NM_000815 0.71 0.28 acid (GABA) A receptor, delta hypothetical protein NM_017639 1.14 3.78 FLJ20047 interferon, omega 1 NM_002177 0.90 2.31 major intrinsic protein NM_012064 0.66 1.73 of lens fiber selectin E (endothelial NM_000450 193.25 51.23 adhesion molecule 1)

130

Other Classifications or Unclassified Genes

Gene Name Accession Number Expression Level – Expression Level – TNFα Only Group Combo Group 3-oxoacid CoA NM_022120 1.51 0.60 transferase 2 acrosin NM_001097 0.80 2.14

aldolase A, fructose- NM_000034 0.47 0.19 bisphosphate

crystallin, gamma B NM_005210 0.68 2.80 cystatin B (stefin B) NM_000100 1.10 3.13

dachshund homolog 1 NM_004392 1.06 3.97 DEAD (Asp-Glu-Ala- NM_014314 1.40 3.57 Asp) box polypeptide 58 DnaJ (Hsp40) homolog, NM_018198 4.09 14.87 subfamily C, member 11 enoyl Coenzyme A NM_001398 0.59 1.72 hydratase 1, peroxisomal general transcription NM_001520 2.09 5.28 factor IIIC, polypeptide 1, alpha 220kDa

guanylate binding NM_002053 3.23 8.98 protein 1, interferon- inducible, 67kDa

interleukin 1 receptor NM_017416 0.51 1.46 accessory protein-like 2 ladinin 1 NM_005558 0.80 2.51

praja 1 NM_022368 1.01 0.40 programmed cell death 2 NM_002598 1.22 3.28 prominin 1 NM_006017 0.12 0.32 Purkinje cell protein 4 NM_006198 0.95 3.17

ribosomal protein S17 NM_001021 1.09 0.38 solute carrier family 40 NM_014585 0.77 1.96 (iron-regulated transporter), member 1 sulfotransferase family NM_014351 0.89 2.30 4A, member 1 sulfotransferase family, NM_001056 0.86 3.44 cytosolic, 1C, member 1 tetratricopeptide repeat NM_018259 1.55 0.61 domain 17 tropomodulin 2 NM_014548 0.91 2.80 (neuronal)

131

Vita

Brian Eric Reese

Education 08/2000-08/2005 Ph.D., Integrative Biosciences – Neuroscience Option, Penn State Univ., College of Medicine; GPA: 3.71

08/2000-08/2003 M.B.A., Penn State University, GPA: 3.89

08/1995-05/2000 B.S., Cellular Biochemistry, Cum Laude; SUNY at Plattsburgh; GPA: 3.6

Publications

Reese, BE, Davidson CE, Billingsley, ML, Yun J “TNFα Modulates Stannin Gene Expression in Human Umbilical Vein Endothelial Cells” 2004, In Press J Exp Therapeutics and Pharmacology.

Reese, BE, Krissinger, D, Davidson, CE, Billingsley, ML, Yun, JK “The Effects of Stannin Knockdown on HUVEC Response to Tumor Necrosis Factor-α” 2004, Manuscript submitted to Gene Expression

Davidson, CE, Reese, BE, Billingsley, ML, Yun J “Stannin, a Protein That Localizes to Mitochondria, Sensitizes NIH-3T3 Cells to Trimethyltin and Dimethyltin Toxicity” 2004, Molecular Pharmacology, Oct 66(4):14400-10.

Davidson, CE, Reese, BE, Billingsley, ML, Yun J “The Trimethyltin-Toxicity Mediating Protein Stannin Specifically Binds to 14-3-3ζ” 2004, Manuscript accepted in Molecular Brain Research.

Honors/Awards

-First place in Graduate Division of the First Annual College Venture Challenge Business Plan Competition. -Graduate Student Representative for the College of Medicine Information Technology Committee -Life Sciences Consortium Fellowship from Penn State University -Sudds and Munk Outstanding Senior Award -Chancellor’s Award for Student Excellence -Sankalchand Devjibhai Scholarship for Excellence in Biology - Redcay Honors Scholarship for Leadership - Vice President of Local Psi Chi Chapter

132