The Regulation of rat manganese superoxide dismutase gene

MISSING IMAGE

Material Information

Title:
The Regulation of rat manganese superoxide dismutase gene detection and characterization of trans-acting factors
Alternate title:
Detection and characterization of trans-acting factors
Physical Description:
ix, 142 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kuo, Shiuhyang, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Superoxide Dismutase -- genetics   ( mesh )
Superoxide Dismutase -- metabolism   ( mesh )
Genes, Structural   ( mesh )
Trans-Activators   ( mesh )
Binding Sites   ( mesh )
Enzyme Induction   ( mesh )
Gene Expression Regulation   ( mesh )
Rats   ( mesh )
Department of Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Biochemistry and Molecular Biology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Bibliography: leaves 122-141.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Shiuhyang Kuo.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029470505
oclc - 48668456
System ID:
AA00024972:00001

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
    Abstract
        Page viii
        Page ix
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
    Chapter 2. In vivo architecture of the rat MnSOD promoter: Basal transcription factors
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
    Chapter 3. In vivo architecture of the rat MnSOD promoter: LPS, TNF-a, and IL-1B-specific transcription factor
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
    Chapter 4. Library screening and cloning of the basal transcription factor
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
    Chapter 5. Conclusion and future directions
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
    References
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
    Biographical sketch
        Page 142
        Page 143
        Page 144
Full Text








THE REGULATION OF RAT MANGANESE SUPEROXIDE DISMUTASE GENE:
DETECTION AND CHARACTERIZATION OF TRANS-ACTING FACTORS












By

SHIUHYANG KUO























A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1998

















This work is dedicated to the memory of my beloved mother.




















I have a dream. In my dream,

ethics, religion, and the sciences are harmoniously mingled.

I have a dream. In my dream,

there is no black, brown, red, white, or yellow, but a rainbow.

I have a dream. In my dream,

there is no prokaryote, eukaryote, or Homo sapiens sapiens, but a biota.















ACKNOWLEDGMENTS

I would like to thank my mentor, Dr. Harry Nick for taking me as one of his

students and teaching me how to be a good scientist. I hope that I can be a good scientist one day. I would like to thank my committee members, Drs. Ferl, Kilberg, Purich, and Yang for their continuous support for these six years. I also would like to thank Dr. McGuire for his taking time to listen to my naive scientific opinions; I very much appreciated it.

Sallie offered her help when my daughter was hospitalized at three months old, which I will always remember. Maureen and Joan were always very considerate and patient to me. For Maureen, she was always a good senior student to me. Without Joan's good managements of experimental materials, I would not be able to do my experiments smoothly. Jane gave me a lot of suggestions when I was doing library screening experiments, which I value very much. Rich, Mike, Chris, and Vince are good colleagues to work with in the same laboratory. They taught me about American culture, and tried to shape my English. I knew that they have difficult time to do that due to my strong accent and poor English grammar, but I really enjoyed their education.



iii









At last, but not the least, I would like to thank my wife, Hsoumei, who walked

with me through my very difficult time for the last decade. We will walk together for the coming decades, and I believe we can make it.







































iv















TABLE OF CONTENTS

pM,e

ACKNOWLEDGMENTS ----------------------------------------------------------------- iii

ABSTRACT ---------------------------------------------------------------------------------- viii

CHAPTERS

1 INTRODUCTION ------------------------------------------------------------------------- 1

Free Radicals ------------------------------------------------------------------------------- I
Types and Physiological Significance of Superoxide Dismutases (SODs) ------- 2 Molecular Biology of MnSOD ----------------------------------------------------------- 8
Transcriptional Regulation of A TATA- and CAAT-Less Gene ------------------- 11

2 IN VIVO ARCHITECTURE OF THE RAT MnSOD PROMOTER: BASAL
TRANSCRIPTION FACTORS --------------------------------------------------------- 15

Introduction -------------------------------------------------------------------------------- 15
Materials and Methods ------------------------------------------------------------------ 20
Cell Culture ----------------------------------------------------------------------------- 20
In Vivo DMS Treatment --------------------------------------------------------------- 20
In Vitro A, C, G, T-Specific Chemical Reactions for Protein-Free DNA ------ 21 Ligation-Mediated Polymerase Chain Reaction (LMPCR) ----------------------- 23
Preparation of M13 Single-Stranded DNA Probe ---------------------------------- 25
Serum-Free Starvation of L2 Cells --------------------------------------------------- 26
Results -------------------------------------------------------------------------------------- 27
Identification of Ten Basal Transcription Factor Binding Sites ------------------ 27
The Relationship Between 5-Methyl Cytosine and The Binding Sites for
Potential Basal Transcription Factors ------------------------------------------------ 44
Cell Cycle Regulation of The Rat MnSOD Gene ---------------------------------- 51
Discussion ---------------------------------------------------------------------------------- 54
The Identify of Possible Transcription Factors That Bind to Basal
Binding Sites ----------------------------------------------------------------------------- 57


v










A Hypothesis for The Purpose of 5-Methyl Cytosine Residues Identified
on ThePromoter Region of The MnSOD Gene----------------------------- 63
The Biological Significance of The Enhanced Cytosine at Position +51 ---65

3 IN VIVO ARCHITECTURE OF THE RAT MnSOD PROMOTER: LPS,
TNF-a, AND IL-1If3-SPECIFJC TRANSCRIPTION FACTOR -------------- 68

Introduction---------------------------------------------------------------- 68
Biology of Lipopolysaccharide, Tumor Necrosis Factor-a, and Interleukin- 1 68
Materials and Methods ---------------------------------------------------- 72
Cell Culture-------------------------------------------------------------- 72
In Vivo DMS Treatment--------------------------------------------------73
In Vitro Guanine-Specific Chemical Reaction for Protein-Free DNA ------- 74 Ligation-Mediate Polymerase Chain Reaction (LMPCR) ------------------- 74
Preparation of M 13 S ingle-Stranded DNA Probe -------------------------- 75
LIP-cDNA Transient Transfection into L2 Cells --------------------------- 75
RNA Isolation and Northern Analysis------------------------------------- 76
Preparation of Random Primer Extension Probes-------------------------- 78
Results-------------------------------------------------------------------- 78
Identification of One Stimulus-Specific Binding Site ---------------------- 78
NE-icB Does Not Bind To The Rat MnSOD Promoter---------------------- 81
Expression of Liver-enriched Inhibitory Protein (LIP) in L2 Cells Does
Not Affect The Induced Expression of The Rat MnSOD Gene-------------- 83
Discussion-----------------------------------------------------------------87
A Model of The In Vivo Promoter Architecture of The Rat MnSOD Gene ---- 88
Is LAPNF-1L6 The Stimulus Specific Activator for The Induction of The
Rat MnSOD Gene?-------------------------------------------------------91

4 LIBRARY SCREENING AND CLONING OF THE BASAL
TRANSCRIPTION FACTOR ----------------------------------------------- 94

Introduction ---------------------------------------------------------------- 94
Materials and Methods -----------------------------------------------------95
Screening of A XZAP II Rat Lung Expression Library --------------------- 95
Preparation of Catenated Double-Stranded DNA Probe -------------------98
Results-------------------------------------------------------------------- 100
Screening of A XZAP II Rat Lung Expression Library--------------------100
In Vivo Excision for Cloning The Potential Positive Clone----------------- 108
Cloning Directly from Lambda Phage DNA------------------------------108
In Vitro DMS Footprinting to Verify The Potential Positive Clone ---------111I
Discussion ----------------------------------------------------------------113


vi









5 CONCLUSION AND FUTURE DIRECTIONS ------------------------------------ 115

REFERENCES ------------------------------------------------------------------------------ 122

BIOGRAPHICAL SKETCH -------------------------------------------------------------- 142











































vii












Abstract of Thesis Presented to the Graduate School of the Univeristy of Florida in Partial
Fulfillment of the Requirements for the Degree of Doctor of Philosophy

THE REGULATION OF RAT MANGANESE SUPEROXIDE DISMUTASE GENE: DETECTION AND CHARACTERIZATION OF TRANS-ACTING FACTORS By

SHIUHYANG KUO

August, 1998

Chairman: Dr. Harry S. Nick
Major Department: Biochemistry and Molecular Biology

Manganese superoxide dismutase (MnSOD), an enzyme of the mitochondrial matrix, is the primary cellular defense against superoxide radicals generated as a byproduct of aerobic metabolism and as a consequence of disease pathologies which involve an inflammatory response. It is well documented that elevated expression of this enzyme provides a potent cytoprotective advantage during acute inflammation. Mammalian organisms have therefore evolved endogenous cytoprotective mechanisms to elevate the cellular levels of MnSOD through induction of MnSOD mRNA by proinflammatory mediators including lipopolysaccharide (LPS), tumor necrosis factor(TNF-a), and interleukin-l (IL-1). The nuclear encoded MnSOD gene contains a GCrich and TATA/CAAT-less promoter which falls into the category of a house-keeping gene, however, in contrast to most housekeeping genes, this gene is not constitutively expressed but rather has a basal expression level which can be dramatically induced in a viii









variety of cells by numerous proinflammatory mediators. To understand the underlying regulatory mechanisms for basal and induced transcription of the MnSOD gene, I have employed dimethyl sulfate in vivo footprinting coupled with ligation-mediated polymerase chain reaction to reveal the protein-DNA contacts at single nucleotide resolution. I have identified eleven potential binding sites in the MnSOD proximal promoter region. One of these binding sites is LPS, TNF-a, and IL- 13-specific, whereas the remaining ten binding sites are always present in control cells, and stimuli treated cells. I have thus identified an in vivo promoter architecture of an inducible TATA/CAAT-less gene. I have also performed transient transfection of L2 cells with a LIP expression vector. The overexpression of LIP in L2 cells suggested that NF-IL6/LAP is not involved in the induced expression of the rat MnSOD gene. I then further screened a rat lung lambda cDNA expression library to identify and clone one of the proteins bound to a basal binding site. I have identified a potential positive clone which may constitute a novel family of transcription factors.


















ix














CHAPTER 1
INTRODUCTION

Free Radicals

Free radicals, which are defined as atoms or molecules with one or more unpaired electrons in the outer orbital, are very unstable, and thus very reactive. Examples of free radicals are the superoxide anion radical, hydroxyl radical, and hydroperoxyl radical. As a group, the so called reactive oxygen species (ROS) include hydrogen peroxide, hypochlorous acid, and the above free radicals. The production of reactive oxygen species is found in most cell types, including fibroblasts, epithelial cells, endothelial cells, adipocytes, and tumor cells (Janssen et al. 1993). Formation of these ROS is widely distributed within cells in the mitochondrial electron transport chain (Bandy and Davison 1990), the cyclooxygenase pathway, and by cellular enzymes including P450 oxidase, xanthine oxidase and NADPH oxidase (Bandy and Davison 1990; Trush and Kensler 1991). Phagocytic leukocytes make use of oxygen molecules (oxidative burst) to produce various ROS during phagocytosis. The metabolic pathway of ROS can be summarized as follows: Oxygen molecules are transformed into superoxide anion radical (02 ') by NADPH oxidase, xanthine oxidase, P450 oxidase, or redox active compounds. Superoxide anion radical (02 ') can spontaneously dismutate or through the action of superoxide dismutases (SOD) into H202, which can then be converted into HOC1 by

1









2

myeloperoxidase. 02- and H202 can be transformed into OH* by divalent cations, such as Fe +. NO *, one of the cellular metabolic products of arginine, can react with superoxide anion radical to generate peroxynitrite (ONOO).

Reactive oxygen species can be very harmful to cells (Janssen et al. 1993),

causing peroxidation of polyunsaturated fatty acids leading to alterations in the integrity and permeability of cell membranes. They will inactivate certain cellular proteins such as glutamate synthetase (Olivier 1987) and SOD (Sharonov and Churilova 1990). Furthermore, ROS will oxidize bases of DNA, cause single and double strand breaks, crosslinking of DNA, and cell death at high enough concentrations (Fridovich 1978; Imlay and Linn 1988; Halliwell and Aruoma 1991). Due to these metabolic reactions, ROSs have been associated with a large number of diseases. Reactive oxygen species have been shown to be associated with aging, cancer, immune complex-mediated disease, and pulmonary disorders (Farmer and Sohal 1989; Farber et al. 1990; Sun 1990; Trush and Kensler 1991).

Types and Physiological Significance of Superoxide Dismutases (SODs)

The composition of the atmosphere changed dramatically three times after the formation of Earth. The atmosphere contained little or no free oxygen initially, then oxygen increased to about 80% nearly 2.0 billion years ago, followed by a drop to about 15%, and gradually elevated to the present level of 20% oxygen (Kasting 1993). The rise in atmospheric oxygen just before the emergence of multicelluar organisms during the Cambrian period correlates with the views of the importance of oxygen levels to









3

biological evolution. To take advantage of oxygen, aerobic systems thus evolved mechanisms to generate energy efficiently from oxygen consumption; however, they also suffered from the toxicity of reactive oxygen species as by-products of aerobic metabolism. About 1-2% of the oxygen used in resting respiration is released as reactive oxygen species (Boveris and Chance 1973), which are too toxic to be tolerated by these living systems. To survive successfully, these living systems evolved a detoxification scheme to remove these reactive oxygen species. The first line of defense is the superoxide dismutases (SODs). The major function of SODs is to detoxify 02-, produced as the by-product of aerobic metabolism, via the following reaction: 02- + 2H'

-3 H202 + 02. Hydrogen peroxide is then converted into water and molecular oxygen by catalase and glutathione peroxidase (Bannister et al. 1987; Fridovich 1986). The SODs, catalase, and glutathione peroxidase form this mutually supportive protective chain to help aerobic systems survive in an aerobic environment, and thus enjoy the advantage of energy generation through oxygen consumption.

Depending on the metals found in their active site, SODs are classified into three types: the predominantly eukaryotic copper- and zinc-containing SODs (Cu/ZnSODs), including a cytoplasmic and an extracellular form; a prokaryotic iron-containing SOD (FeSOD); and manganese-containing SOD (MnSOD), found in both prokaryotic cells and eukaryotic mitochondria. In fact, these three types of SODs are widespread among archaebacteria, eubacteria, and eukaryotes, with no clear border to define which kind of SOD existed first in prokaryotic or eukaryotic cells (Bannister et al. 1987). For example,









4

eukaryotic algae do not have the Cu/ZnSOD, which was found in two bacterial species, Photobacterium leiognathi and Caulobacter crescentus (Bannister et al. 1987). On the other hand, eukaryotic algae contain FeSOD, which has also been identified in the leaves of lemon trees. This raises a very interesting question, whether eukaryotic SODs were derived from prokaryotic cells via endosymbiogenesis, or prokaryotic SODs were from eukaryotic cells via horizontal gene transfer such as in the case of P. leiognathi, which is a symbiont of the ponyfish Leiognathus (Bannister et al. 1987). Superoxide dismutase probably evolved after the appearance of cyanobacteria, since it serves as a defense against oxygen toxicity. Superoxide dismutase in aerobic prokaryotic cells was then most likely passed on to the eukaryotic cells. This argument can be supported by the similarity found between bacterial and mitochondrial SODs (Steinman and Hill 1973). In some special cases (for example, P. leiognathi), SOD was horizontally transferred to prokaryotic cells from eukaryotes as a consequence of environmental changes. Based on the sequence and structural homology between Fe and MnSOD (Stallings et al. 1984), these two enzymes were proposed to evolve from the same common ancestor; however, Cu/ZnSOD evolved independently (Smith and Doolittle 1992). Interestingly, the number of Cu/ZnSOD genes were increased from simple to complex live beings; however, there is only one copy of Fe/MnSOD gene among all species examined to date.

Cu/ZnSOD constitutes about 85% 90% of the total eukaryotic cellular SOD activity. It is located in the cytosol and extracellular matrix, and is also found in chloroplasts (Bannister et al. 1987; Fridovich 1986). Evidence also suggests that the









5

Cu/ZnSOD may be located in peroxisomes (Keller et al. 1991). The human Cu/ZnSOD gene extends 11 kb, and contains 5 exons on chromosome 21 (Bannister et al. 1987). The cDNAs of human (Sherman et al. 1984) and rat (Delabar et al. 1987) have been sequenced. Cu/ZnSOD protein is a homodimer with a molecular weight of 32,000 daltons (Fridovich 1975); however, the extracellular Cu/ZnSOD is tetrametic and has a molecular weight of 135,000 daltons (Fridovich, 1986). The three dimensional structure of cytosol Cu/ZnSOD is similar to a cylinder whose wall is composed of eight antiparallel p sheets (P barrel) (Tainer et al. 1983).

Bacterial MnSOD is a dimer with a molecular weight of 40,000 daltons. In

eukaryotic cells, MnSOD is found in the matrix of mitochondria (Bannister et al. 1987). The mitochondrial MnSOD is tetrameric with a molecular weight of 80,000 daltons (Bannister et al. 1987). The structure of human mitochondrial MnSOD exhibits two identical 4-helix bundles, which form tetrameric interfaces that stabilize the active sites neighbored by metal, Mn+3 (Borgstahl et al. 1992). The human MnSOD gene is located on chromosome 6 (Bannister et al. 1987). The comparison of sequence of human MnSOD protein with that of the cDNA shows that there is a 24 amino acid mitochondrial signal sequence which is removed after the processing of MnSOD protein (Ho and Crapo 1988). A similar situation occurs in rat (Ho and Crapo 1987) and mouse (Hallewell et al. 1986). Basically, all three types of SODs catalyze the same chemical reaction. However, the rate of nucleotide mutation is higher for Cu/ZnSOD than for MnSOD (Smith and Doolittle 1992), which leads us to suspect that they may play different roles in different









6

physiological states, since the rate of nucleotide mutation reflects the needs of the environment.

Rats preexposed to 85% oxygen became tolerant to high doses of oxygen (Frank 1982), and the pulmonary level of SOD is increased in rats exposed to 85% -90% of oxygen (Tsan 1993). This suggests the important role of SOD in protecting living systems from the damage of oxygen. Moreover, the levels of reactive oxygen species parallel the level of oxidative stress, which induces apoptosis, a process of programmed cell death. Since SOD can balance the level of reactive oxygen species, SOD may have an important effect on apoptosis (Sandstrom and Buttke 1994). Recently, the Cu/ZnSOD was shown to associate with familial amyotrophic lateral sclerosis (Rosen et al. 1993), play an important role in Parkinson's disease (Sandler et al. 1993), and was implicated as an important factor in the life-span of Drosophia melanogaster (Orr and Sohal 1994).

Scientists had not paid much attention to MnSOD's role in the protection of

pulmonary cells from oxygen toxicity due to its small percentage of total cellular SOD activity and because its activity is difficult to measure. However, a report by Massaro et al. (Massaro et al. 1992) suggests that the MnSOD but not Cu/ZnSOD plays an important role in the protection of cells from oxygen toxicity. Moreover, adult rats exposed to 85% hyperoxia for 3-5 days showed increased MnSOD mRNA levels but no changes in Cu/ZnSOD or catalase mRNA levels in lung (Ho et al. 1990). MnSOD also has been shown to offer cells resistance to cytotoxicity mediated by TNF-a (Wong et al. 1989), or paraquat (Clair et al. 1991). TNF-cc and paraquat are known to mediate cytotoxicity via









7

oxygen free radicals or superoxide anion radicals. Furthermore, the survival rate of heterozygotic transgenic mice, which overexpress the human MnSOD, was shown to be higher than that of normal mice after they were exposed to 95% oxygen (Wisp6 et al. 1992). The above evidences support the important role of MnSOD in protecting cells from oxygen toxicity.

Recent data have decisively demonstrated the critical cellular importance of

MnSOD in a variety of different tissues. For example, homozygous mutant MnSOD mice die within 10 days of birth exhibiting severe dilated cardiomyopathy, an accumulation of lipid in liver and skeletal muscle, metabolic acidosis, and decreased activities of aconitase, succinate dehydrogenase, and cytochrome c oxidase, enzymes which are all extremely sensitive to alterations in the cellular redox state (Li et al. 1995). Additionally, transgenic mice expressing elevated levels of human MnSOD under the control of a surfactant promoter were highly protected from lung injury during exposure to 95% oxygen and thus survived longer than nontransgenic littermates (Wisp6 et al. 1992). Overexpression of MnSOD has also been implicated in the suppression of tumorigenicity in human melanoma cells (Church et al. 1993), breast cancer cells (Li et al. 1995), glioma cells (Zhong et al. 1997), oral squamous carcinoma cells (Liu et al. 1997) and SV40transformed human fibroblast cells (Yan et al. 1996). The increased MnSOD gene expression and protein levels in whole lung was shown to be related to the degree of lung inflammation (Holley et al. 1992). Alterations in MnSOD levels have also been associated with a number of neurodegenerative diseases, including Parkinson's disease









8

(Eggers et al. 1994), Duchenne muscular dystrophy, Charcot-Marie-Tooth disease, and Kennedy-Alter-Sung syndrome (Yahara et al. 1991).

Molecular Biology of MnSOD

Our laboratory has previously characterized the rat MnSOD cDNA (Dougall

1990). The genomic locus for the rat MnSOD gene was first sequenced by Ho et al. (Ho et al. 1991). The promoter region of this gene contains neither a "CAAT box" nor a "TATA box." Our laboratory has also identified and characterized the rat MnSOD gene. The rat MnSOD gene contains five exons. Exon one encodes the 5' untranslated leader sequence, the mitochondrial signal sequence, and the N-terminus of the rat MnSOD protein. Exon 2, 3, 4, and 5 encode the mature MnSOD protein. Exon 5 contains the stop codon, TGA, and the 3' untranslated region (Dougal 1990). Primer extension analysis was used to locate the transcription initiation site at between 70 and 74 nucleotides 5' to the initiation site of translation (Hurt et al. 1992). There are five species of MnSOD mRNA identified by Northern analysis. Our laboratory has demonstrated that these five species of MnSOD mRNA are caused by differential polyadenylation (Hurt et al. 1992).

The regulation of MnSOD biosynthesis in E coli is under rigorous control. The induction of this enzyme in E. coli is in response to the cellular environmental redox state. E. coli grown in iron-poor medium or in the presence of chelating agents for iron results in an induction of the bacterial MnSOD gene. On the other hand, cells grown in iron-enriched medium leads to an inhibition of MnSOD gene expression. All of these









9

observations lead Fridovich (1986) to suggest that E. coli MnSOD gene is controlled by an iron-containing repressor (Fridovich 1986). More recently, Hassan and Sun (1992), and Privalle and Fridovich (1993) identified Fnr, Fur, and Arc transcriptional regulators, which negatively regulate the expression of MnSOD in E. coli. Unlike bacteria, MnSOD synthesis in eukaryotic cells is upregulated dramatically by proinflammatory mediators including lipopolysaccharide (LPS), tumor necrosis factor alpha (TNF-a), interleukins-1 and -6 (IL-1, IL-6), and interferon gamma (IFN-y) (Wong and Goeddel 1988; Shaffer et al. 1990; Del-Vecchio and Shaffer 1991; Dougall and Nick 1991; Borg et al. 1992; Eddy et al. 1992; Gibbs et al. 1992; Valentine and Nick 1992; Visner et al. 1992; Whitsett et al. 1992; Eastgate et al. 1993; Melendez and Baglioni 1993; Bigdeli et al. 1994; Jacoby and Choi 1994; Akashi et al. 1995; Gwinner et al. 1995; Jones et al. 1995; Lontz et al. 1995; Stephanz et al. 1996). In L2 cells, a rat pulmonary epithelial-like cell line, MnSOD mRNA levels show an 18 23 fold induction after stimulation with lipopolysaccharide (LPS) (Visner et al. 1990), a mediator of the immune response and a component of cell wall of all gram-negative bacteria. Cells treated with TNF-a or IL-1 showed similar results.

To evaluate the importance of on-going protein synthesis and de novo

transcription, studies with cycloheximide, an inhibitor of protein synthesis, showed no effect on LPS, TNF-a or IL-1-dependent induction of MnSOD mRNA level. On the other hand, L2 cells co-treated with stimulant and actinomycin, an inhibitor of mRNA









10

transcription, inhibited the stimulus-dependent induction of MnSOD mRNA level (Visner et al. 1990). Furthermore, nuclear run-on data showed a 9 fold induction in MnSOD mRNA level (Hsu 1993). The above evidences suggest that the regulation of MnSOD gene expression is, at least, partly transcriptionally dependent. The difference between nuclear run-on analysis and in vivo data on mRNA level following LPS treatment may be caused by the stability of the mRNA or the loss of some transcription factors during the preparation of nuclei for nuclear run-on experiments. Examinations of other cell types treated with LPS, TNE-a., or IL-i1 also showed similar results at the mRNA level, including rat pulmonary artery endothelial cells (Visner et al. 1992), porcine pulmonary artery endothelial cells (Visner et al. 1991), and intestinal epithelial cells (Valentine and Nick 1992). Interestingly, though Cu/ZnSOD contributes the major part of the total cellular SOD activity, its mRNA level is not regulated by any known stimulant to any large degree.

Dr. Jan-Ling Hsu in our laboratory has identified seven DNase I hypersensitive sites within and near the rat MnSOD gene. Six of them are located within the MnSOD gene, the other one is located in the promoter region (Hsu, 1993). Furthermore, high resolution DNase I hypersensitive site analysis shows that there is a single LPS, TNF-a or IL- 1 -specific hypersensitive subsite, which appears in the promoter region following stimulus treatment (Hsu, 1993). Promoter deletion analysis data also shows that important cis-acting elements exist within this promoter region (Rogers et al. 1998 submitted for publication). The above observations led to the proposal that there are









11

trans-acting binding site(s), which control the basal and induced expression of MnSOD gene, in the promoter region.

Transcriptional Regulation of A TATA- and CAAT-Less Gene

Two types of DNA sequence elements are associated with the regulation of

transcription in higher eukaryotes: promoters and enhancers. Promoters play critical roles in the architecture of a functional transcriptional initiation complex; enhancers increase the efficiency and rate of transcription. Most of the metazoan protein-coding genes contain promoters with a transcription initiation consensus element, known as the TATA box, which is located at 25-30 base pairs (bp) upstream of transcription start site. In addition, most genes contain a CAAT box located at 70-80 bp upstream of transcription start site (Zawel and Reinberg 1995). TBP (TATA binding protein), one of the components of TFII D protein complex (also contains TAF, TBP-associated factor), has been shown to recognize the TATA box and start the nucleation of the transcription initiation complex including TFIIB, TFITD, TIE, TFIIF, TFIIH, TFIIJ, as well as DNAdependent RNA polymerase II (RNAPII). This complex then interacts with activators, which are either recruited by the transcription initiation complex via protein-protein interaction or find their own way to specific binding sequences. Together, they activate the transcriptional machinery (Ptashne and Gann 1997).

On the other hand, some of protein-coding genes contain promoters without

TATA and/or CAAT boxes. How do they initiate transcription? It turns out that these promoters contain different core regulatory elements either being utilized in









12

transcriptional initiation or facilitating the binding of the transcription initiation complex (Weis and Reinberg, 1992; Smale 1997). Among these core regulatory elements, initiator (Inr) is the most highly studied. The Inr element functions similarly to the TATA box. Both elements can direct accurate transcription initiation by RNAPII and a high level of transcription when stimulated by other trans-acting factors. The Inr element usually extends from -6 to +11 (+1 at transcription start site), and contains the consensus sequence, PyPyA+IN(T/A)PyPy (Py=pyrimidine) (Smale 1997). There are four proteins, which have been shown to specifically recognize the Inr element, including TFIID (Kaufmann and Smale, 1994), TFII-I (Roy et al., 1991), RNAPII (Carcamo et al., 1991) and YY1 (Seto et al., 1991). However, it is not clear yet whether one or more of these proteins are required for the activity of the Inr element.

The second core regulatory element of a TATA-less promoter is the downstream promoter element (DPE) with the consensus sequence of (A/G)G(A/T)CGTG, which is located about 30 bp downstream of transcription start site (Burke and Kadonaga, 1996). Recently, it has been implicated that the DPE is recognized by TAF11 60 (Burke and Kadonaga, 1997). Interestingly, the spacing between Inr and DPE is extremely important, which implies that both core elements cooperate with each other to activate the transcription of a TATA-less promoter if Inr and DPE existed at the same time in the same promoter (Burke and Kadonaga, 1997). Only about 20% of TATA-less promoters contain DPE.









13

Recently, Lagrange et al. (1998) identified a new transcriptional core element, in addition to the TATA/CAAT box, Inr, and DPE, termed TFIIB recognition element (BRE) with the consensus sequence 5'- (G/C)(G/C)(G/A)CGCC -3'. This core element is specifically recognized by TFIIB. This element may play a role in determining the overall strength of a promoter, the upstream to downstream directionality of the transcription preinitiation complex assembly or possibly in transcription initiation (Lagrange et al. 1998). BRE was proposed as a possible candidate for the core element of TATA-less promoters. However, its in vivo role and relevance has yet to be verified.

Most of the studies on TATA-less promoters have focused on core elements and their binding proteins by using naked or constructed DNA templates as experimental systems, which may not reflect the physiological situations. Furthermore, without the help of trans-acting factors, the regulation of transcription would not be possible. For example, three clustered transcription factor Spl sites were reported to be required for efficient transcription of a TATA-less insulin-like growth factor-binding protein-2 promoter (Boisclair et al., 1993). Furthermore, Sp 1-like sites are found in the transforming growth factor-alpha promoter (Chen et al., 1994), as well as in the promoter of human DIA dopamine receptor gene (Minowa et al., 1993). Moreover, it was shown that there were multiple transcription factor binding sites including GATA- 1, Sp 1, IgNFA, Lva, bicoidQ9, NF-icB, HNF-5, WAP5, and ADH on the TATA-less promoter of the human pyruvate dehydrogenase beta gene (Madhusudhan et al., 1995). Therefore, delineating the architecture of TATA-less promoters is a prerequisite to understanding









14

how this type of promoter regulates transcription and how these regions interact with either upstream or downstream enhancers.

For billions years, DNA molecules have evolved and formed functional operative units called genes driven by promoters consisting of different combinations of core elements (TATA/Inr, TATA alone, Inr/DPE, Inr alone, or Inr-/DPE-), which, with assistance of trans-acting factors, regulate the transcriptional machinery. Why do organisms need such a variety of combinations? How are these promoters regulated? In this thesis, I studied the promoter of the manganese superoxide dismutase gene, and tried to answer some of the questions. The manganese superoxide dismutase gene has a TATA- and CAAT-less promoter, which, therefore, can be used as a good system to study, in vivo, the regulation of an inducible TATA/CAAT-less promoter. Understanding the underlying regulatory transcriptional mechanisms of MnSOD gene is not just clinically important but may also shed some light on the nature of a TATA/CAAT-less promoter. I hope this study can further our understanding of transcriptional regulatory mechanisms.














CHAPTER 2
IN VIVO ARCHITECTURE OF THE RAT MnSOD PROMOTER: BASAL TRANSCRIPTION FACTORS

Introduction

Reactive oxygen species (ROS) produced during both normal cellular function, and most importantly, as a consequence of the inflammatory response, have been implicated in the initiation and propagation of a variety of pathological states (McCord and Roy 1982; Weiss et al. 1983; Ward et al. 1988). The superoxide dismutases (SODs) are the primary cellular defense that has evolved to protect cells from the deleterious effects of oxygen free radicals (Bannister et al. 1987; Fridovich 1989). Three forms of SOD have been identified in eukaryotic cells: the cytoplasmic copper/zinc SOD (Cu/ZnSOD), the extracellular Cu/Zn SOD (EC-Cu/ZnSOD), and the mitochondrial manganese SOD (MnSOD). In contrast to the cytoplasmic Cu/ZnSOD, which is expressed constituitively in most cases, MnSOD gene expression is highly regulated by proinflammatory mediators in a variety of tissues including intestinal epithelial cells (Grisham et al. 1990; Valentine and Nick, 1992), hepatocytes (Dougall and Nick 1991; Czaja et al. 1994), pulmonary epithelial (Wong and Goeddel 1988; Visner et al. 1990) and endothelial cells (Visner et al. 1992), as well as in neurons and astrocytes derived from the central nervous system (Kifle et al. 1996).

15









16

Recent data have decisively demonstrated the critical cellular importance of

MnSOD in a variety of different tissues. For example, homozygous mutant MnSOD mice die within 10 days of birth exhibiting severe dilated cardiomyopathy, an accumulation of lipid in liver and skeletal muscle, metabolic acidosis, and decreased activities of aconitase, succinate dehydrogenase, and cytochrome c oxidase, enzymes which are all extremely sensitive to alterations in the cellular redox state (Li et al. 1995). Additionally, transgenic mice expressing elevated levels of human MnSOD under the control of a surfactant promoter were highly protected from lung injury during exposure to 95% oxygen and thus survived longer than nontransgenic littermates (Wisp6 et al. 1992). Overexpression of MnSOD has also been implicated in the suppression of tumorigenicity of human melanoma cells (Church et al. 1993), breast cancer cells (Li et al. 1995), glioma cells (Zhong et al. 1997), oral squamous carcinoma cells (Liu et al. 1997) and SV40transformed human fibroblast cells (Yan et al. 1996). Alterations in MnSOD levels have also been associated with a number of neurodegenerative diseases, including Parkinson's disease (Eggers et al. 1994), Duchenne muscular dystrophy, Charcot-Marie-Tooth disease, and Kennedy-Alter-Sung syndrome (Yahara et al. 1991).

MnSOD synthesis in eukaryotic cells is upregulated markedly by proinflammatory mediators including lipopolysaccharide (LPS), tumor necrosis factor alpha (TNF-a), interleukins-1 and -6 (IL-1, IL-6), and interferon gamma (IFN-y) (Wong and Goeddel, 1988; Shaffer et al., 1990; Dougall and Nick, 1991; Borg et al., 1992; Valentine and Nick, 1992; Whitsett et al., 1992; Jacoby and Choi, 1994; Akashi et al., 1995). This









17

induction is blocked completely by actinomycin D suggesting that the increase in MnSOD mRNA in response to LPS, TNF-ct, or IL-1 may result from an increase in the rate of transcription of the MnSOD gene (Wong and Goeddel 1988; Visner et al. 1990; Borg et al. 1992; Valentine and Nick 1992; Visner et al. 1992; Bigdeli et al. 1994; Stephanz et al. 1996), results confirmed by nuclear run-on studies (Hsu, 1993). Although highly inducible to levels which often exceed the basal expression by 50-100 fold, the rat MnSOD gene contains a GC-rich promoter lacking a TATA and CAAT box. This promoter architecture was originally associated with housekeeping genes that are constituitively expressed (Dynan 1986). The additional layer of transcriptional regulation of this gene differentiates it from most housekeeping genes. Unfortunately, current knowledge about the molecular mechanisms controlling transcriptional regulation from promoters which lack a TATA- and CAAT-box is limited. Most of the studies addressing regulation of TATA- or CAAT- less promoters have focused on either the initiator (Inr), an element which controls transcriptional initiation (Smale and Baltimore 1989), or the general transcription machinery, especially TFIID (Pugh and Tjian 1991; Colgan and Manley 1992; Wiley et al. 1992; Burke and Kadonaga 1996) and, most recently, TFII-I (Johansson et al. 1995). In addition, most studies have analyzed transcription from TATA- and CAAT-less promoters by employing naked DNA templates in vitro, a model system which may not adequately reflect the physiological situation.









18

Hsu (1993) has employed DNase I hypersensitive (DNase I HS ) site analysis to map DNase I HS sites along the MnSOD gene and promoter region upstream transcription start site to 5 kb. She has observed seven HS sites, including one located in the promoter region, along the MnSOD gene. Following a high resolution DNase I analysis of the HS site in the promoter region, she observed four HS subsites (1-1 to 1-4 in Figure 2-1) responsible for constitutive expression of MnSOD gene, and a 5' most subsite specific for stimulus treated samples. Her results are summarized in Figure 2-1. DNase I HS sites in chromatin are generally free of nucleosomes, however, analysis of HS sites at higher resolution has demonstrated that while such sites may include segments of unbound DNA, they also contain internal regions associated with non-histone DNAbinding proteins such as RNA polymerase II and, most importantly, various transcription and regulatory factors (Pauli et al. 1988). To further delineate the binding of specific transcription factor(s), at single nucleotide resolution, in the proximal promoter region of the rat MnSOD gene, I used genomic in vivo footprinting coupled with ligation-mediated polymerase chain reaction (LMPCR) to screen the region surrounding the prominent promoter HS site and the transcription start site in L2 cells.









19





E B P P E
I I III IV V VI VII




/ exon 1 2 3" - 4 5
/ ~P S 1-41-31-2 1-1


I ]

B R
B=BamH I E=EcoR I P=Pst I R=Rsa I

represents DNase I Hypersensitive Site




Figure 2-1. Summary of DNase I hypersensitive (HS) site data (Hsu 1993). HS sites are numbered by Roman numerals. The stimulus-specific HS subsite is marked by S. The arrow represents the transcription start site.









20

Materials and Methods

Cell Culture

The L2 rat pulmonary epithelial-like cell line (ATCC CCL 149) was grown as a monolayer in 150 mm cell culture dishes containing Ham's modified F12K medium (GIBCO) supplemented with 10% fetal bovine serum, 10 Ig/ml penicillin G, 0.1 mg/ml streptomycin, and 0.25 .tg/ml amphotericin B at 370C in humidified air with 5% CO2. At approximately 90% confluence, cells were treated with 0.5 ptg/ml Escherischia coli (E. coli) LPS (E. coli serotype 055:B5, Sigma), 10 ng/ml TNF-a (kindly provided by the Genentech Corp.), or 2 ng/ml IL-13 (kindly provided by the National Cancer Institute) for

0.5 to 8 hr to induce MnSOD gene expression. Untreated cells were used as controls. In Vivo DMS Treatment

L2 cells were cultured as described above in 150 mm plates. The medium was removed and cells washed with room temperature phosphate buffered saline (PBS, 10 mM sodium phosphate, pH 7.4 and 150 mM NaCl). The PBS buffer was removed and replaced with room temperature PBS containing 0.5%-0.25% dimethyl sulfate (DMS, Aldrich) for 1-2 min at room temperature. The PBS containing DMS was rapidly removed, and the cell monolayer washed with 40C PBS to quench the DMS reaction. The cells were lysed in 5 ml of lysis solution containing 66.7 mM EDTA pH 8.0, 1% SDS, and 0.6 mg/ml proteinase K, followed by incubation overnight at room temperature. Genomic DNA was then purified by phenol/chloroform extractions. Each sample was extracted once with an equal volume of Tris-equilibrated phenol followed by two









21

extractions with a 24:24:1 (v/v/v) mixture of Trisphenol-chloroform-isoamyl alcohol, and finally by one extraction with a 24:1 (v/v) mixture of chloroform-isoamyl alcohol. The aqueous phase collected each time by centrifugation at 14,000 g for 10 min at room temperature and ethanol precipitated. Samples were then treated with 100 gg/ml RNase A, organic extracted as above, precipitated and suspended in TE (10 mM Tris pH 8.0, and

1 mM EDTA). The DNA samples were digested with BamH I, and strand cleavage at modified guanine residues was achieved by treatment with IM piperidine (Fisher) at 900C for 30 min. Naked genomic DNA was harvested and purified from cells without any DMS treatment and restricted with BamH I. In Vitro A, C, G, T-Specific Chemical Reactions for Protein-Free DNA

I used 25-30 pag BamH I restricted purified genomic protein-free DNA for each chemical reaction. The samples were lypholized and resuspended in appropriate amount of H20.

Adenine/guanine-specific chemical reaction. Genomic DNA was resuspended in 20 pl H20 followed by the addition of 50 [d formic acid (Fisher). The final formic acid concentration is 63% (40 g1 formic acid from Sigma can be used alternatively, in this case the final formic acid concentration will be 66%). Samples were then incubated at room temperature for 10 min. The reaction was stopped by adding 200 gl cold stop solution (2.53 M NH4OAc, 0.0675 pgg/p1 E. coli tRNA), and 750 p.1 cold 100% ethanol. Samples were immediately incubated in dry ice-ethanol bath for at least 5 min followed by centrifugation at 40C for 15 min. I then added 250 p.l common reagent (1.875 M









22

NH4OAc and 0.1 mM EDTA) and 750 jtl cold 100% ethanol followed by the incubation in dry ice-ethanol bath for at least 5 min. Each sample was centrifuged at 40C for 15 min and then lypholized and resuspended in 90 pl H20. Piperidine cleavage (final concentration = 1 M) was performed at 900C for 30 min. Ethanol precipitation of each sample was done after the sample was cooled down to room temperature. The final lypholized sample was ready for ligation-mediated PCR as described below.

Guanine-specific chemical reaction. Each DNA sample was resuspended in 10 p1 H20 followed by the addition of 190 ptl dimethyl sulfate (DMS) buffer (50 mM sodium cacodylate and 0.1 mM EDTA) and DMS (final concentration, 0.25%). Each sample was incubated at room temperature for 30 sec. The reaction was quenched by adding 68.1 p.1 cold DMS stop solution (7.35 M NI-OAc and 0.2 p g/pl E. coli tRNA) and cold 100% ethanol, and the sample was immediately incubated on a dry ice-ethanol bath for at least 5 min followed by centrifugation at 40C for 15 min. The following procedures (common reagent addition and piperidine cleavage) are the same as described in adenine/guaninespecific chemical reaction.

Cytosine/thymine-specific chemical reaction. Each DNA sample was

resuspended in 20 ptl H20 followed by the addition of 20 pl hydrazine (Aldrich), and was incubated at room temperature for 4 min. The reaction was stopped by adding 200 P1 cold pyrimidine stop solution (0.1 mM EDTA, 2.34 M NH4OAc, and 0.063 [tg/pl E. coli tRNA) and 750 pl cold 100% ethanol followed by incubation on a dry ice-ethanol bath for at least 5 min. Each sample was then centrifuged at 40C for 15 min. The following









23

procedures (common reagent addition and piperidine cleavage) are the same as described in adenine/guanine-specific chemical reaction.

Thymine-specific chemical reaction. Each DNA sample was resuspended in 20 tl H20 and boiled for 5 min, and was immediately incubated at ice-water bath. Twenty j.d of freshly prepared 0.1 mM KMnO4 in cold 20 mM Tris-HCI, pH 7.0 was added to the denatured DNA (final KMnO4 concentration = 0.05 mM). Each sample was then incubated at 800C for 1 min. Chemical reaction was stopped by adding 200 jtl cold pyrimidine stop solution and 750 jil cold 100% ethanol followed by incubation at dry iceethanol bath for at least 5 min. Each sample was then centrifuged at 40C for 15 min. The following procedures (common reagent addition and piperidine cleavage) are the same as described in adenine/guanine-specific chemical reaction. Ligation-Mediated Polymerase Chain Reaction (LMPCR)

The LMPCR was performed as described previously (Garrity and Wold 1992).

Briefly, 6 pmole of a promoter specific primer one was annealed to 2 jig DMS/piperidine cleaved DNA for each sample in lx Vent buffer (New England BioLabs), with denaturation at 950C for 5 min, followed by primer annealing at 450C for 30 min. The primer extension was performed in lx Vent buffer with dATP, dCTP, dGTP, and dTTP at

0.25 mM, and 2 U Vent DNA polymerase (New England BioLabs). The samples were incubated for 1 min each at 530C, 550C, 570C, 600C, 620C, and 660C, followed by 680C, and 760C for 3 min each. The extension reaction was stopped by addition of 20 gl of a 40C solution containing 50 mM DTT, 18 mM MgC12, 0.125 mg/mI BSA, and 110 mM









24

Tris pH 7.5. Twenty five pl of a ligation solution (20 mM DTT, 10 mM MgCl2, 0.05 mg/ml BSA, 3 mM ATP), 4 pmole annealed common linker (5'-GAATTCAGATC-3', and 5'-GCGGTGACCCGGGAGATCTGAATTC-3'), and 4.5 units (U) T4 DNA ligase were combined and incubated overnight at 160C followed by ethanol precipitation. Following the ligation, PCR amplification (Coy ThermoCycler II) was performed for 25 cycles in lx Vent buffer, 3 mM MgSO4, 0.25 mM dNTP, 25 pmole promoter specific primer two, 20 pmole common primer, and 3 U Vent DNA polymerase. For the first cycle, the DNA is denatured at 950C for 3 min, annealed for 2 min at a temperature specific for each primer, and then extended at 760C for 3 min. The remaining cycles were

1 min at 950C, 2 min at the specific annealing temperature, and 3 min plus a 5 sec extension for each cycle at 760C. The reaction was terminated with 38 tl cold stop solution containing 6.8 M NH4OAc, 27 mM Tris pH 7.5, 11 mM EDTA pH 7.7 and 0.26 p.g/pl E. coli tRNA, followed by organic extraction of the amplified DNA products and ethanol precipitated. The following primers were used for LMPCR (see Figure 2-2 for their positions) : for the top strand primer sets: A. primer one 5'-TTGTGCCGCTCTGTTACAAG-3', primer two 5'-GTGTCGCGGTCCTCCCCTCCGTTGATG-3'; B. primer one 5'-ATTGTAGCTCACAGGCAGAG-3', primer two 5'-GGGCCTAGTCTGAGGGTGGAGCATA-3'; C. primer one 5'-TGATTACGCCATGGCTCTGA-3', primer two 5'-TCTGACCAGCAGCAGGGCCCTGGCTT-3'; for the bottom strand primer sets: G. primer one 5'-CATAGTCGTAAGGCAGGTCA-3', primer two 5'GTCAGGGAGGCTGTGCTTGTGCCG-3'; H. primer one 5'-GCCGAGACCAA-








25

CCAAA-3', primer two 5'-GCCGCCCGACACAACATTGCTGAGG-3'; I. primer one 5'-CTGCTCTCCTCAGAACA-3', primer two 5'-AACACGGCCGTTCGCTAGCAGCC-3'; J. primer one 5'-ATCAACGGAGGGGAGGA-3', primer two 5'-CGGCCCAGCTTGTAACAGAGCGGCAC-3'. The PCR products were size fractionated on a 6% denaturing polyacrylamide gel, electrotransferred to a noncharged nylon membrane (Cuno) and covalently cross-linked to the membrane by UV irradiation. The membrane was prehybridized in a buffer containing 0.76 M sodium phosphate (NaHPO4), pH 7.4, 7% (w/v) SDS, 1% (w/v) BSA (Sigma A-7511), and 1 mM EDTA at 650C for 15 min and hybridized with an M13 single-stranded probe over night. After overnight hybridization, the membranes were washed 3-4 times with 1 mM EDTA, 40 mM sodium phosphate, pH

7.4, and 1% (w/v) SDS at the appropriate temperature for 10 min each time followed by exposure to X-ray film (Amersham).

Preparation of M 13 Single-Stranded DNA Probe

An M13 clone with the MnSOD promoter insert was originally isolated and

cloned by Dougal (1990). This promoter insert contains a 5.5 kb EcoR I/Pst I fragment which was used as the template for generating a single-stranded M13 DNA probe. The ratio of template to each oligo primer (primer two in each individual LMPCR primer set) for primer extension was optimized beforehand. The appropriate amounts of M 13 template and oligo primer were mixed together in annealing buffer (200 mM NaCl and 50 mM Tris, pH 8.0), and the total volume was brought to 20 ptl with H20. The above mixture was boiled for 3 min, and incubated at 500C for 45 min. At the end of the









26

incubation, 3.3 mM each for dGTP, dCTP, and dTTP and 100 pCi [ot-32p]-dATP were mixed with extension buffer (final 5 mM MgCl2, 7.5 mM DTT). Ten units of the large fragment of E. coli DNA polymerase (New England BioLab) were added in a total reaction volume of 40 pl and incubated at room temperature for 45 min. The reaction was stopped by adding 90 pl formaldehyde dye (10 mM EDTA, 0.00003% (w/v) of bromophenol blue and xylene cyanol, each, in deionized formamide). The mixture was then boiled for 5 min and loaded onto a prerun minigel (6% denaturing polyacrylamide gel) for about 10-15 min allowing the bromophenol blue and xylene cyanol dyes to be well separated. The glass plates were separated, and the polyacrylamide gel was wrapped in plastic wrap. The position of the probe was detected by Polaroid photography. The probe was cut out of the gel, ground into a paste, and was eluted in hybridization solution containing 0.76 M sodium phosphate (NaHPO4), pH 7.4, 7% (w/v) SDS, and 1% (w/v) BSA.

Serum-Free Starvation of L2 Cells

L2 cells were grown as described above. Cells were washed twice with prewarmed PBS, and changed into Ham's modified Fl2K medium (GIBCO) supplemented with 10 p.g/ml penicillin G, 0.1 mg/ml streptomycin, and 0.25 pg/ml amphotericin B, and 0% FBS. Cells were grown at 370C in humidified air with 5% CO2 for 48 hr. Cells were then washed with pre-warmed PBS, and then refed the same medium containing 10% FBS for another 1, 2, 4, 8, or 24 hr. Cells without 10% FBS refeeding were used as control. The samples were then subject to in vivo DMS treatment.









27

Results

Identification of Ten Basal Transcription Factor Binding Sites

I employed genomic in vivo footprinting using dimethyl sulfate (DMS) as a

molecular probe coupled with ligation-mediated PCR (LMPCR) to resolve possible cisacting elements at single nucleotide resolution and thus display the in vivo protein-DNA contacts. DMS is a small hydrophobic chemical probe which can enter intact cells and react predominately by methylating the N-7 atom of guanine and, to a lesser extent, the N-3 atom of adenine in duplexed DNA. Amino acid side chains of trans-acting factors which contact guanine residues can protect these bases from methylation by DMS. Alternatively, amino acid side chains can create a hydrophobic pocket around specific guanine residues which increases the DMS solubility and results in enhanced reactivity. Ultimately protein side chains produce a footprint composed of protections and/or enhancements which appear as diminished or more intense bands as compared to the corresponding band in the naked DNA guanine ladder on the final sequencing gel autoradiograph (Nick and Gilbert 1985).

The relative positions of LMPCR primer sets used in this study are shown in Figure 2-2. In order to verify that the kinetics of basal transcription factor binding are stable throughout the whole period, control samples without stimuli treatment were










28














F
0, E
D +1
C




-1000 +200
4-4 G
4H




L L
M 100 bp










Figure 2-2. Primer sets used in LMPCR. A total of 13 primer sets were used to screen 720 bp upstream, and 180 bp downstream with respect to the transcriptional initiation site. #A, #B, #C, #D, #E, #F are top strand primer sets, which were used to screen bottom strand sequences. #G, #H, #I, #J, #K, #L, #M are bottom strand primer sets, which were used to display top strand sequences. With the exception of primer sets D and K, the other primer sets were used to identify basal transcription factor binding sites. The sequences for primer sets are detailed in the Materials and Methods. The directions of arrowheads represent the 5'-->3' orientation.









29

compared with samples induced with stimuli such as lipopolysaccharide (LPS). Both control and stimulated samples were sampled for testing after 0.5, 4, and 8 hr of treatment. These experiments demonstrated that the observed protein-DNA contacts are detectable as early as 0.5 hr and as late as 8 hr after the addition of LPS.

Illustrated in Figures 2-3, 2-4, 2-5 and 2-6 are representative examples from each time point. Figure 2-3 illustrates in vivo footprinting and LMPCR results for control and 0.5 hr LPS treated samples for the top strand of the promoter from position -166 to -286 relative to the transcriptional initiation site. Figure 2-4 illustrates control and 4 hr LPS treated samples for the bottom strand. As depicted in Figures 2-3 and 2-4, numerous guanine residues exhibited altered DMS reactivity which appeared as either diminished or enhanced hybridization signal relative to the in vitro DMS-treated DNA lanes. I have summarized this in vivo footprinting data by postulating the existence of protein binding sites at obviously clustered residues and through symmetry in the contacts and in the DNA sequence. Figures 2-3 illustrates binding sites for proteins I V on the top strand, while Figure 2-4 shows binding sites for proteins from II to VI on the bottom strand. Binding site I has guanine residues protected from DMS methylation at positions -273,

-271, -270, -268, -266, and -265, and an enhanced guanine at position -267 on the top strand, but no contacts on the bottom strand. Binding site H1 has protected guanines at

-254, -253, -252, -250, and -247 on the top strand, and at -246 on the bottom strand. Binding site III is delineated by protected guanine residues at positions -234, -233, -232,



























Figure 2-3. Identification of basal transcription factor binding sties I to V on the top strand (-286 to -166) of the MnSOD promoter. In vivo DMS footprinting primer set J was used for LMPCR. Control or LPS treated cells (30 min) were exposed to DMS in vivo and DNA isolated and fractionated as described in the materials and methods. The same results were observed in LPS 4 hr treated samples. Lanes G, guanine sequence derived from DMS treated purified genomic DNA; lanes C, guanine sequence from in vivo DMS treated control cells; and lanes L, guanine sequence from in vivo DMS/LPS treated cells. Each C and L lane represents individual plates of cells. Open circles, 0, represent protected guanine residues, whereas filled circles, @, represent enhanced guanine residues. The arrowheads represent enhanced adenine residues. Each bar represents an individual binding site with Roman numeral designation. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.








31



Top Strand
in vivo
GGCCCLL
-286 8Aaaa












II








-oV 4166 -





























Figure 2-4. Identification of basal transcription factor binding sties from HI to VI on the bottom strand (-258 to -134) of the MnSOD promoter. Primer set C was used for LMPCR. Control or LPS treated cells (4 hr) were exposed to DMS in vivo and DNA isolated and fractionated as described in the materials and methods. Lanes G, guanine sequence derived from DMS treated purified genomic DNA; lanes C, guanine sequence from in vivo DMS treated control cells; and lanes L, guanine sequence from in vivo DMSILPS treated cells. Each C and L lane represents individual plates of cells. Open circles, 0, represent protected guanine residues, whereas filled circles, 0, represent enhanced guanine residues. The arrowheads represent enhanced adenine residues. Each bar represents an individual binding site with Roman numeral designation. The nuclteotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.






33

Bottom Strand
in vivo
GGCCL L
-134 <8I




.4oIV










-011


-258









34

-230, -228, and -226, an enhancement at -227 on the top strand, and no contacts on the bottornstrand. Protected guanine residues at positions -195, -194, -192,-191, -189, -187, and 186 on the top strand, and at 185 on the bottom strand define binding site IV. Binding site V has protected guanines at positions -177, -176, -175, -174, -172, -170, and

- 169 on the top strand, 173, and 168 on the bottom strand, and an enhanced guanine at position 166 on the top strand.

Figures 2-5 and 2-6 illustrate the protein-DNA contacts seen in cells stimulated for 8 and 4 hr of LPS on the top and bottom strands, respectively. Binding sites 11 through VII are shown in Figure 2-5, while Figure 2-6 illustrates binding sites VI, VIII, IX, and X. Binding site VI exhibits symmetrically protected guanine residues at positions

-152 and -151 on the top strand, and -145 and -144 on the bottom strand. Binding site VII has five continuous guanines protected on the top strand from position 13 3 to 129 and no contacts on the bottom strand. A single protected guanine at position 115 has been postulated to define binding site VM based on the distance and isolation from sites VII and IX Consecutive protected guanines at positions -68, -67, -66, and -65 delineate binding site IX, and protected guanine residues at positions -47 and -46 define site X. No contacts were observed on the top strand for sites VHI-X. Interestingly, in addition to the guanine contact sites, I also observed consistently reproducible enhanced adenines marked by arrowheads from Figure 2-3 to 2-6, which are also clustered near specific binding sites.






























Figure 2-5. Identification of basal binding sites il-VII on the top strand (-254 to -121) in the promoter of the MnSOD gene. Primer set I was used for LMPCR. For in vivo samples, cells were treated with LPS for 8 hr, identical results were obtained for 30 min and 4 hr LPS-treatment. As in Figures 2-3 and 2-4, lanes G, C, and L reflect in vitro DMS-treated DNA, in vivo DMS-treated control or LPS exposed cells, respectively. Each C and L lane represents an individual plates of cells. Open circles represent protected guanine residues, whereas filled circles represent enhanced guanine residues. The arrowheads represent enhanced adenine residues. Each bar represents an individual binding site with Roman numeral designation. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.







36




Top Strand
in vivo

-254 amamlo





OwIw.1W- /8


40 04








-l21SS8""




























Figure 2-6. Identification of binding sites for basal transcription factors VI to X on the bottom strand (- 150 to -3 1) of the MnSOD promoter. Primer set B was used for LMPCR. Cells were either non-treated or treated for 4 hr with LPS. All of the symbols are identical to those used in Figure 2-3, lanes G, C, and L are in vitro DMS-treated DNA, in vivo DMS-treated control or LPS exposed cells, respectively. Each C and L lane represents an individual plate of cells. Open circles represent protected guanine residues, whereas filled circles represent enhanced guanine residues. The arrowheads represent enhanced adenine residues. Each bar represents an individual binding site with Roman numeral designation. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.






38
Bottom Strand
in vivo
GGCCLL
-31 0
-NOO N 81x

~I x







- Vill






-150o VI
-150









39

The initiator element (Inr), (PyPyA+iN(T/A)PyPy) (Javahery et al. 1994) and

downstream promoter element (DPE), (A/G)G(A/F)CGTG, (Burke and Kadonaga 1996) located at -+30, were shown to be important core elements for the regulation of TATAless promoter genes. To determine whether these regulatory elements existed near or downstream to the transcriptional start site of the rat MnSOD gene, I used computer analysis, and [ located a reverse Inr-like sequence at -40 and a DPE-like sequence at +56 of the MnSOD promoter. As a result of the reported significance of these elements, I employed genomic in vivo footprinting to further examine the region downstream to the transcription start site. But I did not observe any protected or enhanced guanine residues on either strand as far 3' as +180 bp. Interestingly, I did observe an enhanced adenine residue at position -38 on the bottom strand within the Inr-like sequence as shown in Figure 2-6, and an enhanced cytosine residue at position +51, also on the bottom strand, upstream to the DPE-like sequence. Most interestingly, the intensity of this cytosine residue can be dramatically increased following by stimulation by LPS. This intriguing phenomenon is shown in Figure 2-7 (A). However, this enhancement is not always reproducible in every sample. In Figure 2-7 (B), I showed another set of experiments as an example to demonstrate the problem of reproducibility. In summary, this enhanced cytosine residue never appeared in any protein-free genomic DNA sample, and it only appeared in about 45% of total in vivo DMS treated samples. However, 70% of stimuli treated samples showed a more intense signal for this cytosine residue compared in vivo control samples. I hypothesized that this enhanced cytosine residue may be involved in






























Figure 2-7. Identification of an enhanced cytosine residue at +51 position on the bottom strand of the rat MnSOD gene. Primer set A was used for LMPCR. (A). Lanes G, C, and L are in vitro DMS-treated DNA, in vivo DMS-treated control or LPS exposed (30 min) cells, respectively. Each C and L lane represents an individual plates of cells. The enhanced cytosine residue was marked by a star. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure. (B). The same symbols are used as in (A), except lanes I represent IL- 1 P exposed cells for 4 hr.








41





Bottom Strand
in vivo
G G C C L L +65





+51








Bottom Strand
in vivo
G G G C C C I I I I I +55 +51 C


+42



























Figure 2-8. Summary of the in vivo DMS footprinting for ten potential basal binding sites and the enhanced cytosine residue at positions +5 1. The MnSOD promoter sequence is depicted from position -339 to +62 relative to the transcriptional initiation site (+ 1). HS S 1- 1 to HS S 1-4 represent hypersensitive (HS) subsites 1 -1 to 1-4 within HS site I defined by the high resolution DNase I HS site studies (Figure 2- 1). The position of each HS site was defined by the fragments migration relative to molecular markers within an accuracy of 50 base pairs. Open circles, 0, represent protected guanine residues, filled circles, 0, represent enhanced guanine residues. The arrows represent enhanced adenine residues, the star designates an enhanced cytosine residue. Each bar represents an individual binding site with Roman numeral designation. The Inrlike and DPE-like sequences are boxed. The sequences differ from published sequences (Ho et al. 1991) are underlined.



















I
HSS1-4 0 00 oew
-339 CCAGGAATGGAAAAGGAGTGGAGACATTGTAGCTCACAGGCAGAGGTGGCCAAGGCGGCCCGAGAAGAGGCGGGGCCTAG -260
GGTCCTTACCTTTTCCTCACCTCTIGTAACATCGAGTGTCCGTCTCCACCGGTTCCGCCGGGCTCTTCTCCGCCCCGGATC


II III IV
0000 0 000 0 000 HSSI-3 00 00 0 00
-259 TCTGAGGGTGGAGCATAGCCACACCGGG TGCGGGCACGAGCGGGCCGAGGCCAAGGCCGGTGATGGAGGCGTGGCCACAC -180
AGACTCCCACCTCGTATCGGTGTGGCCCACGCCCGTGCTCGCCCGGCTCCGGTTCCGGCCACTACCTCCGCACCGGTGTG
0 0

V vi vii viii
0000000 9 00 00000 HSS1-2
-179 TAGGGGCGTGGCCGTGGCAAGCCCGCGGGCTCTACCAACTCGGCGCGGGGGAGACGCGGCCTTCCCT-GTGTGCCGCTCTG -100
ATCCCCGCACCGGCACCGTTCGGGCGCCCGAGATGGTTGAGCCGCGCCCCCTCTGCGCCGGAAGGGA IACACGGCGAGAC
04 0 00 0

Ix x
HSS1-1 Inr-like
-99 TTACAAGCTGGGCCGTCCGTGTCGCGGTCCTCCCCTCCGTTGATGGGCGCTGCCGGCAGFGTT-CA GC CCTAGCTGTGTCC -20
AATGTTCGACCCGGCAGGCACAGCGCCAGGAGGGGAGGCAACTACCCGCGACGGCCGTCI GGATCGACACAGG
0000 00

DPE-like
-19 TTGCGGACGCCGGGCGGACGCCGCAGAGCAGACGCGCGGCTGCTAGCGAACGGCCGTGTTCTGAG GAGAGCAG+7GTFGgGTI +62
TTCGCCTGCGGCCCGCCTGCGGCGTCTCGTCTGCGCGCCGACGATCGCTTGCCGGCACAAGACTCCTCTCGTCGICCACCACI









44

the regulation of basal as well as induced transcription of the rat MnSOD gene or is a result caused by the stalling of transcription by RNA polymerase II or related protein(s). What I meant by that is the footprinting of a dynamic protein moving along the DNA molecule is not detectable if the time frame of DMS reactivity is slower than the rate of the moving protein. Thus this enhanced cytosine residue at the position +51 may be caused by RNA polymerase II or the entire transcription complex.

The genomic footprinting data obtained were summarized in Figure 2-8. The DNase I HS subsites obtained at high resolution analysis were also approximately mapped along the promoter region.

The Relationship Between 5-Methyl Cytosine and The Binding Sites for Potential Basal Transcription Factors

Three to six percents of cytosine residues are methylated in mammalian cells. The biological significance of 5-methyl cytosines (m5Cs) must be important, otherwise the unstable m5Cs (m5C can be oxidized to thymine) would have vanished through natural selection. These 5-methyl cytosines (m5Cs) are predominantly in CpG sites in the 5' end of genes indicating that m5Cs may be involved in the regulation of the expression of a specific gene through DNA-transcription factor(s) interactions. It has been well documented that the existence of m5Cs relates with the inactivation of genes, the affinity of protein-DNA interaction, and chromatin structure (Razin and Riggs 1980; Chomet 1991). Pfeifer et al. (1990) have reported that every cytosine of all CpGs is methylated on inactive X-chromosomes, while not on active X-chromosome for human phosphoglycerate kinase I gene. The silencing of corresponding genes were related to a









45

closed chromatin structure, a high density of methylated cytosine residues on CpG islands, and the lack of detectable transcription factor on their corresponding regulatory elements (Selker 1990). On the other hand, the repressor MeCP2, which binds to methylCpG sequences, may aid in the recruitment of other co-repressors and/or deacetylases thus further strengthening the inactivation of the genes in a specific chromosomal region (Kass et al. 1997). The DNA binding ability of some transcription factors, such as E2F (Kovesdi et al. 1987), and cAMP responsive element binding protein (Iguchi-Ariga and Schaffner 1989) were found to be reduced by cytosine methylation; however, the DNAbinding affinity of methylated DNA-binding protein (Huang et al. 1984) was found to be enhanced by cytosine methylation. Interestingly, the binding of the transcription factor Sp 1 to its specific sequence was found not to be affected by the methylation status of the binding sequence (Harrington et al. 1988). It seems that m 5Cs may play different roles in various situations, and in some cases leading to opposite outcomes.

I have shown that there are ten potential binding sites for basal transcription factors in the proximal promoter region of rat MnSOD gene. This raises the question regarding how these proteins identify their specific binding sites in such a crowded region (about 270 bp distributed for potential ten proteins) especially since some of the binding sequences are so similar to each other. Is there a common structure or binding sequence, which can be used as a "landmark" for these proteins to reach their "homes"? Does secondary or tertiary DNA structure play a role? Another possibility is that the









46

methylation state of specific cytosine residues may affect protein-DNA interactions (Razin and Riggs 1980; Molloy and Watt 1990).

Protein-free genomic DNA was isolated, purified, and subjected to

adenine/guanine-, guanine-, cytosine/thymine-, and thymine-specific chemical reaction. An original Maxam-Gilbert DNA sequencing reaction was employed to locate m5C. Methylated cytosine residues have been shown to be less reactive to hydrazine than are cytosine and thymine, so that bands corresponding to m5Cs will not appear in the pyrimidine cleavage ladders (Ohmori et al. 1978). Comparing the A+G, G only, C+T, and T only sequence patterns, I then can define the positions of m5Cs. I chose to examine the region covering binding sites II VII on the top strand, since I found a lot of proteinguanine contacts on this strand, and the protein binding sites are closely clustered to each other. Figure 2-9 shows the positions of m5Cs are at -96, -125, -136, -141, -156, -162,

-180, and -208, which flank the binding sites VIII, VII, VI, V, and IV. In Figure 2-10, I summarize the positional relationship between ten potential basal binding sites and the m5Cs. m5Cs at -208, and -180 flank binding site IV. Binding site V has m5Cs at -180, and -162 flanking both sides; m5Cs at -156, and -141 flank binding site VI. Binding site VII is surrounded by m5Cs at -136, and -125; m5Cs at -125, and -96 flank binding site VIII. m5Cs were found predominantly existing on CpG dinucleotides in mammalian cells. Interestingly, I observed some m5Cs on CpT, and CpA, in addition to CpG dinucleotides. Toth et al. (1990) have also reported that m5Cs on CpT, and CpA in the promoter region of late E2A promoter integrated into cell line HE2.





























Figure 2-9. Identification of m5C flanking the binding sites II VII on the top strand of the MnSOD promoter. Primer set I was used for LMPCR. Cells were either non-treated or treated for 4 hr with LPS. Lanes -, and + are in vivo DMS-treated control or LPS exposed cells, respectively. Each and + lane represents an individual plates of cells. Lanes G are guanine-specific DMS reaction, lanes A/G are formic acid depurination reaction, lane T/C is pyrimidine-specific hydrazine reaction, and lane T is thyminespecific potassium permanganate reaction. The positions of 5-methyl cytosines are marked by m5C. Each bar represents an individual binding site with Roman numeral designation. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.









48






T A A LPS
TCGGGG-.----+++++
mop


-229
'Af
00 do
-208-C 6.0 ro
gA
11W liv

-180n'C
al JV
wo
-162"'C
-156"'C IVI

-141"'C
-136"'C Vil

-125"Ic


4 d .



-96-C



AW


-73--


























Figure 2-10. Summary of the relative positions for the ten potential basal binding sites and the m5Cs.
(A). The MnSOD promoter sequence is depicted from position -339 to +62 relative to the transcriptional initiation site (+1). Each bar represents an individual binding site with Roman numeral designation. The bold and shadowed Cs marked by arrows represent 5methyl cytosines. The Inr-like and DPE-like sequences are boxed. The sequences differ from published sequences (Ho et al. 1991) are underlined.
(B). A model shows that the m5Cs of CpG dinucleotides only appear outside the potential binding sites, but not within the binding sites. m5CNs represent methylated CpN dinucleotides; CNs represent unmethylated CpN dinucleotides. DNA molecule is illustrated by the straight line, and the ovals represent the potential binding sites.
















I

(A)-339 CCAGGAATGGAAAAGGAGTGGAGACATTGTAGCTCACAGGCAGAGGTGGCCAAGGCGGCCCGAGAAGAGGCGGGGCCTAG -260
GGTCCTTACCTTTTCCTCACCTCTGTAACATCGAGTGTCCGTCTCCACCGGTTCCGCCGGGCTCTTCTCCGCCCCGGATC


II 111 4 IV 4
-259 TCTGAGGGTGGAGCATAGCCACACCGGGTGCGGGCACGAGCGGGCCGAGGCCAAGGCCGGTGATGGAGGCGTGGCCACAC -180
AGACTCCCACCTCGTATCGGTGTGGCCCACGCCCGTGCTCGCCCGGCTCCGGTTCCGGCCACTACCTCCGCACCGGTGTG



V VI VII VIII
4 4 4 4 4
-179 TAGGGGCGTGGCCGTGGCAAGCCCGCGGGCTCTACCAACTCGGCGCGGGGGAGACGCGGCCTTCCCTGTGTGCCGCTCTG -100
ATCCCCGCACCGGCACCGTTCGGGCGCCCGAGATGGTTGAGCCGCGCCCCCTCTGCGCCGGAAGGGACACACGGCGAGAC


Ix x Inr-like

-99 TTACAAGCTGGGCCGTCCGTGTCGCGGTCCTCCCCTCCGTTGATGGGCGCTGCCGGCAGp5TUA--G-(l3CCTAGCTGTGTCC -20
AATGTTCGACCCGGCAGGCACAGCGCCAGGAGGGGAGGCAACTACCCGCGACGGCCGTCLCAQTf.tGGATCGACACAGG


.1 DPE-like
-19 TTGCGGACGCCGGGCGGACGCCGCAGAGCAGACGCGCGGCTGCTAGCGAACGGCCGTGTTCTGAGGAGAGCAGCG9- GG-T--G1+62
TTCGCCTGCGGCCCGCCTGCGGCGTCTCGTCTGCGCGCCGACGATCGCTTGCCGGCACAAGACTCCTCTCGTCGL2LLCAC




(B)









51

Cell Cycle Regulation of The Rat MnSOD Gene

Previously, I illustrated the existence of ten potential basal binding sites, which had been identified in unsynchronously growing populations of L2 cells. In other words, these ten basal binding sites may be occupied by transcription factors throughout the cell cycle. In order to test the possibility that some of these proteins may occupy the promoter at specific times in the cell cycle, I synchronized cells by starving L2 cells with medium containing 0% FBS for 48 hr followed by either no, or 1, 4, 8, or 24 hr refeeding medium containing 10% FBS. Cells were then subject to in vivo DMS treatment. I observed that all of the ten potential basal binding sites were continuously occupied in synchronized cells as was seen in an asynchronously population of growing L2 cells. I concluded that the transcription factors are bound to these ten binding sites all the time, and are not associated with cell cycle regulated-transcription. Interestingly, I did observe the appearance of the same enhanced cytosine at +51 as in Figure 2-7. The intensity of this enhanced cytosine was strongest in synchronized cells (starvation without refeeding cells) as shown in Figure 2-11. The result of this experiment may support my hypothesis, which is that this enhanced cytosine may be caused by the stalling of transcription by RNA polymerase HI or related protein(s). I will discuss my hypothesis in the Discussion section.




























Figure 2-11. In vivo DMS footprinting of the synchronized L2 cells. Primer set A was used for LMPCR. Lane G represents naked genomic DNA ladder. Unsynchronized cells were marked -, and used as a control. Plus signs (+) represent synchronized cells. Lane +/0 represents the synchronized cells without refeeding medium containing 10% FBS. The time periods for incubation with 10% FBS after synchronization and before in vivo DMS treatment were 1, 4, 8, or 24 hr.






53




in vivo
+ + ++ starvation
G 0 1 4 8 24 hr10% FBS +73











+28









54

Discussion

The MnSOD gene has characteristics similar to most housekeeping genes, such as a GC-rich promoter which lacks both TATA and CAAT boxes. In contrast to most housekeeping genes, however, MnSOD is not constituitively expressed but rather has a basal expression level which can be dramatically induced in a variety of cells by numerous proinflammatory stimuli. There are a few examples of other housekeeping genes which can be regulated or induced by nutrients or hormones, such as the dihydrofolate reductase, HMGCoA reductase (Dynan, 1986), pyruvate dehydrogenase (Madhusudhan et al. 1995), and insulin-like growth factor-I receptor (IGF-I-R) genes (Werner et al. 1993). GC-rich promoters lacking both a TATA and CAAT box have also been associated with other inducible and tissue-specific genes, such as the urokinase-type plasminogen activator receptor (uPAR) (Soravia et al. 1995), Pim-1 (Meeker et al. 1990), CD7 (Schanberg et al. 1991), and MAL genes (Tugores et al. 1997).

Recently, significant progress has been made on the structure and function of TATA-less promoters (Weis and Reinberg 1992; Smale 1997). Most of these studies involved identification of initiator (Inr) elements and characterization of general transcription factor(s) by using in vitro systems. For example, TFIID, TFII-I, YY1, or the core RNA polymerase II was found to bind to the Inr element thus aiding in the nucleation of the pre-initiation complex. This pre-initiation complex is thought to interact with upstream activators, such as Spl, and/or enhancer elements to facilitate transcriptional initiation at TATA-less promoters. Furthermore, Burke and Kadonaga









55

(1996) have identified a downstream promoter element (DPE), (A/G)G(AIT)CGTG, located at +30, which was shown to be important for the regulation of TATA-less promoter genes. Another potential core element possibly associated with TATA-less promoters was localized to both sides of the transcription initiation site of the rat catalase gene (Toda et al. 1997).

To date, however, we have limited information on the general machinery involved in the transcription of genes lacking both a TATA and CAAT box. Specific transcription factors such as SpI and the Wilms' tumor suppressor (WT1) have been associated with the developmental and neoplastic down regulation of the TATA/CAAT-less IGF-I-R gene, respectively (Werner et al. 1993). Spl has also been associated with the regulation of numerous TATA/CAAT-less genes including the uPAR (Soravia et al. 1995), T-cellspecific MAL (Tugores et al. 1997), and the human Pim-1 genes (Meeker et al. 1990). Unfortunately, our knowledge of the molecular architecture of an inducible TATA/CAAT-less promoters is quite limited. I believe, therefore, that my in vivo footprinting studies on the transcriptional regulation of the rat MnSOD gene have defined a collection of constitutive basal binding sites, and may delineate the general architecture of an inducible TATA/CAAT-less promoter. My data on in vivo DMS footprinting are summarized in Figure 2-8. HS subsitesl-I to 1-4 defined by Dr. Jan-ling Hsu are constitutive HS sites which are present in both control and stimulated cells. Their relative positions, mapped to within +/- 50 bp, and flank the location of three large cis-acting protein binding regions (binding sites I-Ill, IV-VIII, and IX-X). In Figure 2-12, 1 have









56












BASAL
11 III IVV VIVIIVIII IX X










Figure 2-12. A model for the basal transcription of the rat MnSOD gene. The spacing between each binding site is approximately scaled. The arrow represents the transcription start site of the MnSOD gene.









57

drawn a model indicating the locations of the ten potential basal binding sites relative to the transcriptional start site.

The Identity of Possible Transcription Factors That Bind to Basal Binding Sites

I have defined the exact position of each potential protein-DNA interaction using the in vivo accessibility of guanine residues to dimethyl sulfate methylation. Unlike some enzymatic probes used for in vivo footprinting (such as DNase I) which can define the borders of each DNA-protein binding site, DMS typically defines guanine contacts internal to the complete binding site. DMS, therefore, is not an ideal probe to delineate the entire protein binding sequence, however, an examination of my in vivo footprinting data has allowed me to define clustered protein-DNA contacts as individual binding sites by utilizing the symmetry of the guanine contacts as well as any two fold symmetry in the DNA sequence defined by the contacts. In addition, I hypothesize that each protein has a unique guanine protection or enhancement pattern not unlike a "signature," with the entire consensus binding sequence defining the "address" on the DNA. Having the right "signature" and "address," I can then attempt to predict the possible identity of the protein. Based on these arguments, I have compared the vast transcription factor literature including existing consensus DNA binding sequences as well as available DMS in vivo/in vitro footprinting or methylation interference data with my in vivo footprinting results. In Table 2-1, 1 list the putative transcription factors obtained through both literature searches and computer analysis of transcription factor databases (Prestridge





























Table 2-1. Comparison of Transcription Factor Consensus Binding Sequences with each In Vivo Binding Site. Potential transcription factor consensus binding sequences are shown in the first column, whereas the in vivo contact sites are summarized in the right column. Only transcription factors with consensus binding sequences less than or equal to 2 bp mismatches to the in vivo contact elements were shown. The arrows represent the 5' to 3' orientation of each transcription factor consensus binding sequences, open circles represents protected guanine residues, filled circles represent enhanced guanine residues, and the numbers within the parentheses indicate the matching base pairs out of the total base pairs in the consensus binding sequence.











59








Transcription Factor Consensus In Vivo Contact Elements Binding Sequence
0 00 000
Spi 5' (G/T) (G/A)GG(C/A)G I5'-273GAGGCGGGGC
(G/T) (G/A) (G/A) (C/T) CTCCGCCCCG
'(l0/10)
(Courey and Tjian, 1992) 000 0 0
II 5'-255AGGGTGGAGC TCCCACCTCG
04(7/10)
000 0 00
III 5' -23 5CGGGTGCGGG GCCCACGCCC
owi6/10)
00 00 0 00
HBP-1 CCACGTCACC IV 5'-195GGAGGCGTGGC
(Tabata et al., 1989) CCTCCGCACCG
k -0(8/10)
0000 0 00 0 V 5' -17 7GGGGCGTGGCCG CCCCGCACCGGC
0 0
00
VI 5'-5GGCTCTACC CCGAGATGG
____ ____ ____ ____ ____ ___00 00000
GCF 5'NN(G/C)CG(G/C) (G/C) VII 5'-138GGCGCGGGGGAGA
(G/C)CN (Kageyama (j/0CCGCGCCCCCTCT
and Pastan, 1989) 1/0
EGR-1 5'GCG(C/G)GGGCG(89
(Lemaire et al., 1990) (/)
AP-2 5' (T/C)C(C/G) CC (A/C)
N(G/C) (C/G) (G/C) (9/10)
(Imagawa et al., 1987)

MYB 5(T/CAAC(/T)G VIII 5'-115CCTGTG
MYB 5(T/CAAC(/T)GGGACAC
(Biedenkapp et al., 0
1988) k (4/6)
IX 5' -73GTCCTCCCCTCCG CAGGAGGGGAGGC GCF 0000 (9/10)
AP-2 (/0
MZF-1 5'AGTGGGGA (9/10)8
(Morris et al., 1994) ___________X 5' -55GGGCGCTGCCGG LVc 5'CCTGC (Speck CCCGCGTCGGCC
00
and Baltimore, 1987) (4/5)









60

1991; TFSEARCH, Ver. 1.3, GenomeNet WWW Server) whose consensus DNA binding sequences overlap my footprinting data. The computer analysis used a window of less than or equal to a 2 bp mismatch to define the potential identity of these protein binding sequences. For example, the SpI consensus binding sequence has been defined as 5'(G/T)(G/A)GG(C/A)G(.r/T)(_CJA)(G/A)(C/T) (Courey and Tjian 1992). Li et al.(1991), as part of the bovine papillomavirus (BPV) promoter, have also defined a possible exception to the consensus binding sequence for Sp 1 in which the center base is a T rather than C/A. Sp 1 has three DNA-binding zinc finger motifs where each zinc finger motif contacts 3 bp, with a total binding site size of 10 bp. The distance between the center of two adjacent Spl proteins has also been determined to be no less than 10 bp. Moreover, DMS in vitro footprinting has demonstrated that the guanines at the second, third, fourth, and sixth positions are usually protected (single underlined), and the seventh, and eighth positions are usually enhanced (double underlined) within the consensus binding sequence. The literature has clearly demonstrated that there are no contacts, either protections or enhancements observed on the bottom strand. Considering the GC-rich nature of the MnSOD promoter sequence and after examining our first five binding sequences and their respective guanine contacts, we propose that Sp1 is bound to site I, and that the proteins occupying binding sites II and I are Sp I-like proteins. As supporting evidence, Boisclair (1993) reported that three clustered SpI sites are essential for efficient transcription of the rat insulin-like growth factor-binding protein-2 (IGFBP-









61

2) gene which contains a TATA-less promoter. My data may illustrate a similar situation in which Sp I interacts with site I and two Sp I-like proteins occupy sites II and III. The guanine footprinting patterns of binding sites IV and V differ from that of Sp 1 in that the binding site sizes, defined by my guanine contacts, extend to eleven and twelve bp respectively. I also detected in vivo contacts on both strands which is not indicative of Spl (Courey and Tjian 1992). Interestingly, binding site IV differs by only 2 bp from the reverse binding sequence (Table 2-1) previously shown as the binding site for the wheat histone DNA binding protein-1 (HBP-1) (Tabata et al. 1989; Mikami et al. 1994). Moreover, in vitro methylation interference data (Tabata et al. 1989) has demonstrated that all guanine residues within the HBP-1 binding sequence are important for its binding, a result which is consistent with our in vivo DMS footprinting data (Figure 2-8 and Table 2-1). Wheat HBP- 1 is a basal transcription factor that specifically binds to a hexameric motif (ACGTCA) associated with a number of plant regulatory elements. A purportedly related mammalian transcription factor, ATF, also contains a hexameric motif in its consensus DNA binding sequence, 5'(T/G)(A/T)CGTCA (Hurst and Jones 1987). Methylation interference data from the DNA-ATF complex, however, showed that the middle guanine residue is critical for protein binding; a residue not present on the bottom strand of site IV. Hurst and Jones's (1987) have also reported that when the third thymine residue was mutated to cytosine, ATF lost its affinity for that binding sequence. Therefore a comparison of the site-specific mutational analysis of the ATF binding site (Hurst and Jones 1987) with my in vivo footprinting data for site IV indicates that the site









62

IV binding protein may belong to the ACGTCA family but is most likely not occupied by ATF.

In the case of site VI, my computer analysis revealed no matching or similar transcription factor binding sequences, leading to the postulate that it is occupied by a novel protein. The binding sequences for sites VII and IX have five and four sequentially protected guanine residues on the top and bottom strands, respectively. Based on known consensus DNA binding sequences, the transcription factors AP-2 (Imagawa et al. 1987) and GC factor (GCF) (Kageyama and Pastan 1989) have overlapping consensus sequences with both sites VII and IX (Table 2-1). I also determined that the consensus binding sequences for EGR-1 (Lemaire et al. 1990) and MZF-1 (Morris et al. 1994; Hromas et al. 1996) overlap with sites VII and IX, respectively (Table 2-1).

In vitro DMS protection or interference data for EGR-1 (Christy and Nathans 1989) and AP-2 (Courtois et al. 1990; Williams and Tjian 1991) implicates specific guanine residues on both strands. These in vitro patterns are not similar to my in vivo footprinting data for sites VII or IX. Unfortunately, no information on in vitro guanine contacts is currently available for GCF or MZF- 1. In addition, the documented physiological functions for GCF (Kageyama and Pastan 1989), MZF-1 (Hromas et al. 1996), EGR-! (Cao et al. 1990), and AP-2 (Williams et al. 1988) are not consistent with that of the basal transcription factors interacting with binding sites VII and IX.

In the case of binding sites VIII and X, I identified guanine contacts at positions

- 115 and -46/47, respectively, and propose potential binding sites based solely on their









63

separation from adjacent contacts and possible sequence dyad symmetry. I have included potential consensus sequences for MYB (Tanikawa et al. 1993) and leukemia virus factor c (LVc) (Speck and Baltimore 1987) in Table 2-1. However, methylation interference data for these proteins identifies different specific guanine residues from my in vivo DMS footprinting data for binding sites VIII and X. A Hypothesis for The Purpose of 5-Methyl Cytosine Residues Identified on The Promoter Region of The Rat MnSOD Gene

In addition to the ten basal binding sites, I also observed 5-methyl cytosines

(m5C), whose positions flank these binding sites. Is there a functional purpose for these m5Cs? My hypothesis is that these m5Cs serve as a mechanism to increase the specificity of binding of transcription factors.

von Hippel and his colleagues have proposed a model for the specificity of

protein-DNA interactions (von Hippel and Berg 1986,1989). Like the other proteins the surfaces of regulatory DNA-binding proteins are negatively charged except that their DNA-binding domains are positively charged. Once a positively charged DNA-binding domain contacts the negatively charged surface of DNA molecule, this protein may sit on the DNA. In most cases, the first contact is nonspecific. In other words, this protein does not find its target DNA binding site. It was known that nonspecific protein-DNA interactions can be predominantly attributed to electrostatic forces. The positively charged monovalent ion concentration within the nucleus is so high that the competition between these ions and the protein for DNA will reduce the overall binding through nonspecific contacts. The free protein will then continue to search for a specific contact in a









64

different region of DNA molecule and continue to compete with positively-charged ions for DNA binding. This kind of three dimensional diffusion movement will go over and over among intra- or inter-domains of DNA, which is formed by coiling a long stretch of a DNA molecule, until the correct protein-DNA interaction is found. However, the experimental value (5x1Ol0M1'S- for lac repressor) of this three dimensional diffusion movement was surprisingly found to be much higher than that of theoretical value (108 M_1S-1) (von Hippel and Berg, 1989). In other words, the regulatory DNA-binding protein can find its target site much quicker than what we expected based on a theoretical calculation of three dimensional diffusion. To solve this problem, von Hippel and his colleagues then proposed another protein-DNA interaction in addition to the above three dimensional diffusion movement. Since the DNA-binding domain of a protein is oppositively charged to the surface of the DNA molecule, the protein can take advantage of this electrostatic affinity, and the high positively charged monovalent ion concentration environment to "slide" along the DNA molecule. This "sliding" movement can be considered as one dimensional diffusion movement caused by the microcollisions between a protein and the DNA. The speed of this movement was calculated to be 10 bp/sec for lac repressor. This speed alone is still not high enough to explain the location of the target site on DNA by a DNA-binding protein, von Hippel and his colleagues then proposed that a combination of the above three and one dimensional diffusion movements ultimately facilitate specific protein-DNA interaction. However, there are









65

many combinations of sequences, which may be identical to the specific binding sequence. How do they distinguish specific from non-specific? Furthermore, the movement of regulatory proteins along DNA molecule in this model between either specific or non-specific binding site may not be efficient enough. Considering the existence of binding sites along 230 bp in the rat MnSOD promoter region and the fact that some of them have very similar binding sequences, an additional layer of specificity may be required to result in specific protein-DNA interactions. Bestor (1990) has proposed that methylated eukaryotic sequences are used as a signal to sector the dramatically expanded eukaryotic genome to facilitate gene regulation. Based on Bestor (1990), 1 hypothesize that 5-methyl cytosine modifications I observed in the MnSOD promoter region may perform similar function to promote the efficient identification of the target site by the regulatory DNA-binding protein, and can also serve as markers to increase the specificity of protein-DNA binding. In Figure 2- 10 (B), I have proposed a model to illustrated this situation. The 5-methyl cytosine residues in CpG dinucleotides (mCG) were only found between potential binding sites, but not within binding sites. These mCGs might be used as "landmarks" for specific protein-DNA interactions. The Biological Significance of The Enhanced Cytosine at Position +51

DMS-dependent methylation of cytosine residues has been associated with singlestranded DNA (Kirkegaard et al. 1983). I have observed what appears to be an enhanced cytosine residue at +5 1. The intensity of this enhanced cytosine residue was much stronger in the LPS treated cells than control cells, and no enhanced cytosine was









66

observed on naked DNA (Figure 2-7). Moreover, this enhanced cytosine residue appeared strongest in synchronized cells (Figure 2-11).

In this original article, Kirkegaard et al. (1983) demonstrated that DMS-dependent methylation of cytosine residues was associated with the existence of a single-stranded DNA region, which appeared in E. coli RNA polymerase-promoter complexes. The single-stranded DNA region formed in RNA polymerase-promoter complex presumably is a transcription bubble. The basis for the existence of single-stranded DNA stems from the fact that methylation at the N-3 position of the cytosine residue by DMS is blocked because this position is involved in H-bonding in a double helix. It is possible that the enhanced cytosine residue (+51) that I identified is within a transcription bubble. In that case, an alternative chemical probe, such as potassium permanganate, should be used to identify thymine residues within the single-stranded region of transcription bubble to test this possibility. Recently, Orphanides et al. (1998) identified a regulatory protein termed FACT (facilitates chromatin transcription) from HeLa cell nuclear extract. This protein can facilitate transcript elongation by releasing RNA polymerase II from a obstacle caused by nucleosomes. The addition of purified FACT protein into a constructed chromatin DNA template can promote the elongation of RNAs, which usually stall before the synthesis of 40 nucleotides (+40). It is possible that FACT is associated with or plays a role in appearance of the enhanced cytosine residue at position +51. By computer analysis, I also found a downstream promoter element (DPE)-like element 3' downstream to this enhanced cytosine. It is also possible that this enhancement may result from









67

structural alterations specific to the DPE-like element. In this case, further confirmation will require alternative molecular probes.














CHAPTER 3
IN VIVO ARCHITECTURE OF THE RAT MnSOD PROMOTER: LPS, TNF-a, AND IL-113-SPECIFIC TRANSCRIPTION FACTOR Introduction

Biology of Lipopolysaccharide, Tumor Necrosis Factor-a, and Interleukin-1

Lipopolysaccharide (LPS). Lipopolysaccharide (LPS) is a component of the outer membrane of all gram-negative bacteria (Rietschel and Brade 1992). Lipopolysaccharide is composed of a polysaccharide region including O-antigen, hydrophilic inner and outer core, and a lipid region, hydrophobic lipid A, which contributes to the biological activities of LPS (Sweet and Hume 1996). Two types of LPS receptors have been identified. One of them is CD 14, found on cells of the myeloid lineage (Wright et al. 1990), whereas the other receptor is a soluble form of CD14 (sCD 14), which is employed to activate nonmyeloid cells (Pugin et al. 1993). For nonmyeloid cells, for example, endothelial or epithelial cells, a serum glycoprotein, LPS binding protein (LBP), will bind to LPS via lipid A, followed by the replacement of LBP by sCD14. This LPS-sCD14 complex will presumably bind to a receptor then trigger the activation of cells through a series of signal transduction pathways. The proposed signal transduction pathways for the activation of LPS include mitogen-activated protein kinase (MAPK), protein kinase C (PKC), sphingomelin derived ceramide-activated protein kinase (CAK), or G proteins related protein kinase A (PKA) pathway.

68









69

The above signaling pathway(s) will activate transcription factor(s), which then bind to the specific binding site(s) and regulate the expression of genes induced by LPS. The transcription factors found to be associated with LPS activation are two Ets family proteins, Ets-2 (Boulukos et al. 1990) and Elk-1 (Reimann et al. 1994), which are macrophage-specific, LPS-responsive factor (LR1) (Williams and Maizels 1991), Egr-1 (Coleman et al. 1992), AP-1 (Mackman et al. 1991; Fujihara et al. 1993), NF-icB (Mackman et al. 1991; Lowenstein et al. 1993; Zhang et al. 1994), and NF-IL6 (Bretz et al. 1994; Zhang et al. 1994).

Tumor Necrosis Factor-a (TNF-a). Tumor necrosis factor-a performs its

biological activities via two receptors, p60 TNF receptor (TNFR-I, p55) and p80 TNF receptor (TNFR-H, p75). Once TNF-a binds to its receptor, a variety of TNFR-associated proteins will react with the cytoplasmic domain of TNFR and trigger downstream signal transduction pathways (Damay and Aggarwal 1997). A number of signal transduction pathways have been proposed to mediate TNF-a. For example, TNF-a can activate sphingomyelinase and generate ceramide from sphingomyelin (Wiegmann et al. 1994). Ceramide can then serve as a second messenger and trigger the downstream signal pathway via MAPK (Kolesnick and Golde 1994). The intracellular C-terminal region of TNFR-I has been found to be homologous to the intracellular domain of Fas. This homologous domain can initiate the signal and presumably perform a similar function to Fas, namely, programmed cell death. This region of the protein was referred to as a death domain (Tartaglia et al. 1993). All these signaling pathways are proposed to activate a









70

transcription factor, NF-iB, via a kinase (R6gnier et al. 1997). Indeed, a great deal of literature exists showing the tight relationship between TNF-a and NF-KB. For example, Beg et al. (1993) have demonstrated that TNF-a can trigger a signal and lead to phosphorylation of IKBa, an inhibitor of NF-KB. Once IBa is phosphorylated, it will dissociate from the NF-KB heterodimer (p50 and p65 monomers). NF-KB is then activated and can enter the nucleus and activate transcription. Bierhaus et al. (1995) have also suggested that AP- 1 in addition to NF-KB is required for the induction of human tissue factor gene by TNF-a, as has been shown for the collagenase gene (Brenner et al. 1989).

Interleukin-1 (IL-1). The Interleukin-1 family consists of IL-la, IL-1 3, and IL-1 receptor antagonist (IL-ira). Two types of IL-1 receptors (IL-1R) were identified, type I IL-lR (IL-1RI) and type II IL-1R (IL-1RII) (Sims et al. 1989; McMahan et al. 1991). All three members of the IL-1 family can bind to both IL-1Rs, but the type II IL-1R preferentially binds IL-103. However, only type I IL-1R can trigger a signal in response to IL-1 (Sims et al. 1993). It was then proposed that type II IL-1R functions as a "decoy" receptor to regulate the activities of IL-1 3 (Colotta et al. 1994). Once IL-1 binds to its receptor, it will trigger a series of signal pathways and that orchestrate its activities on cells. Almost all the identified signal transduction pathways have been found to be associated with IL-1 activities. G proteins and GTPase, sphingomyelin-ceramide pathway, prostaglandin E2 (PGE2), MAPK, cAMP-dependent kinase (PKA), protein kinase C (PKC), and other kinases have all been reported or suggested to be utilized as









71

signal transduction pathway for IL-1 (Bankers-Fulbright et al. 1996). On the other hand, it was also reported that IL-1 along with the type I receptor can be internalized via endocytosis and accumulated in the nucleus (Mizel et al. 1987; Solari et al. 1994). Furthermore, the internalized IL-1 was still bound to its receptor and the internalized ILiR correlated with increased signal transduction (Curtis et al. 1990). In addition, three major regulatory transcription factors AP-1, NF-B, and/or NF-IL6 are believed to be activated in response to IL-1 stimulation (Banders-Fulbright et al. 1996), and thus regulate IL-1 targeted genes.

It is obvious that LPS, TNF-u, and IL-1 utilize many common signal transduction pathways and regulatory transcription factors. This phenomena may reflect the evolutionary benefit of cell stress and its conservation through common signal. It would be very interesting to examine which regulatory DNA-binding protein(s) are responsible for the induction of the rat MnSOD gene by these three proinflammatory mediators.

As described previously in Chapter 2, MnSOD mRNA levels show an 18 23 fold induction after stimulation of L2 cells with LPS (Visner et al. 1990), similar results were observed on cells treated with TNF-a or IL-1. To evaluate the importance of on-going protein synthesis and de novo transcription, studies with cycloheximide, an inhibitor of protein synthesis, showed no effect on LPS, TNF-ax or IL-I-dependent induction of MnSOD mRNA level. On the other hand, L2 cells co-treated with stimulant and actinomycin, an inhibitor of mRNA transcription, inhibited the stimulus-dependent









72

induction of MnSOD mRNA level (Visner et al. 1990). The above evidence suggests that the regulation of MnSOD gene expression is, at least, partly transcriptionally dependent. This was confirmed by nuclear run-on studies, which demonstrated a 3-9 fold increase in nascent RNA transcription in response to these pro-inflammatory mediators. Furthermore, Dr. Jan-Ling Hsu in our laboratory has identified a single LPS, TNF-a, or IL--specific hypersensitive subsite by using high resolution DNase I hypersensitive (HS) site analysis (Hsu, 1993). To further explore the stimuli-specific cis-acting element at single nucleotide resolution. I then employed genomic in vivo footprinting coupled with ligation-mediated polymerase chain reaction (LMPCR) to examine this region for stimulus-specific contacts.

Materials and Methods

Cell Culture

The L2 rat pulmonary epithelial-like cell line (ATCC CCL 149) was grown as a monolayer in 150 mm cell culture dishes containing Ham's modified F12K medium (GIBCO) supplemented with 10% fetal bovine serum, 10 ptg/ml penicillin G, 0.1 mg/ml streptomycin, and 0.25 gg/ml amphotericin B at 370C in humidified air with 5% CO2. At approximately 90% confluence, cells were treated with 0.5 gtg/ml Escherischia coli (E. coli) LPS (E. coli serotype 055:B5, Sigma), 10 ng/ml TNF-a (kindly provided by the Genentech Corp.), or 2 ng/ml IL-1 3 (kindly provided by the National Cancer Institute) for

0.5 to 8 hr to induce MnSOD gene expression. Untreated cells were used as controls.









73

In Vivo DMS Treatment

L2 cells were cultured as described above in 150 mm plates. The medium was removed and cells washed with room temperature phosphate buffered saline (PBS, 10 mM sodium phosphate, pH 7.4 and 150 mM NaCl). The PBS buffer was removed and replaced with room temperature PBS containing 0.5%-0.25% dimethyl sulfate (DMS, Aldrich) for 1-2 min at room temperature. The PBS containing DMS was rapidly removed, and the cell monolayer washed with 40C PBS to quench the DMS reaction. The cells were lysed in 67 mM EDTA pH 8.0, 1% SDS, and 0.6 mg/ml proteinase K, followed by incubation overnight at room temperature. Genomic DNA was then purified by phenol/chloroform extractions (Sample was extracted once with an equal volume of Tris-equilibrated phenol followed by two extractions with a 24:24:1 [v/v/v] mixture of Trisphenol-chloroform-isoaml alcohol, and finally by one extraction with a 24:1 [v/v] mixture of chloroform-isoamyl alcohol.) and the aqueous phase collected each time by centrifugation at 14,000 g for 10 min at room temperature was ethanol precipitated. Samples were then treated with 100 ptg/ml RNase A, organic extracted, precipitated and suspended in TE (10 mM Tris pH 8.0, and 1 mM EDTA). The DNA samples were digested with BamH I, and strand cleavage at modified guanine residues was achieved by treatment with IM piperidine (Fisher) at 900C for 30 min. Naked genomic DNA was harvested and purified from cells without any DMS treatment and restricted with BamH I.









74

In Vitro Guanine-Specific Chemical Reaction for Protein-Free DNA

Twenty-five microgram of DNA sample was resuspended in 10 jil H20 followed by the addition of 190 [t1 dimethyl sulfate (DMS) buffer (50 mM sodium cacodylate and 0.1 mM EDTA) and DMS (final concentration is 0.25%). Each sample was incubated at room temperature for 30 sec. The reaction was quenched by adding 68.1 p.l cold DMS stop solution (7.35 M NH4OAc and o.2 p g/tl E. coli tRNA) and cold 100% ethanol, and sample was immediately incubated at dry ice-ethanol bath for at least 5 min followed by centrifugation at 40C for 15 min. Each sample was immediately incubated in dry iceethanol bath for at least 5 min followed by centrifugation at 40C for 15 min. The chemical waste was discarded. Two hundred and fifty microliter of common reagent (1.875 M NH4OAc and 0.1 mM EDTA) and 750 jd cold 100% ethanol was added into the DNA pellet followed by the incubation in dry ice-ethanol bath for at least 5 min. Each sample was centrifuged at 40C for 15 min and then lypholized and resuspended in 90 PI H20. Piperidine cleavage (final concentration = 1 M) was performed at 900C for 30 min. Ethanol precipitation of the sample was done after the sample was cooled down to room temperature. The final lypholized sample will be ready for ligation-mediated PCR as described below.

Ligation-Mediated Polymerase Chain Reaction (LMPCR)

The LMPCR procedures was performed as in Materials and Methods in Chapter

2. Except the following six primer sets were used. Top strand primer set: D. primer 1 5'GTTAATTGCGAGGCTGGCAA-3', primer 2, 5'-CCCTAACCTCAGGGGCAAC-








75

AAAG-3'; E. primer one 5'-GTCGTTTTACATTTATGGTGG-3', primer two 5'GGGTTTAGTCAGGAAAGATGAACCTGGC-3'; F. primer one 5'-GGAAAAACCACCCGGAAC-3', prime two 5'-CAGTGGCAGAGGAAAGCTGCC-3'; bottom strand primer set: K. primer 1, 5'-CGGTGTGGCTATGCT-3', primer 2, 5'-GCTCCACCCTCAGACTAGGCCCCGCCT-3'; L. primer one 5'-CTTTTCCATTCCTGGTTCTGG-3', primer two 5'-CAGAGCCATGGCGTAATCAGGGGCCT-3'; M. primer one 5'-CATCTCAGGTTTTAGTGTGTTC-3'; primer two 5'-CTTTGTTGCCCCTGAGGTTAGGG-3'. Their relative positions are shown in Figure 2-1. Preparation of M 13 Single-Stranded DNA Probe

The same procedures were performed as in Materials and Methods section in Chapter 2.

LIP-cDNA Transient Transfection into L2 Cells

L2 cells were cultured in 150 mm plates to about 80% confluent as previously

described in Chapter 2. After the removal of medium, cells were washed with 25 ml prewarmed to 370C PBS followed by another wash with 25 ml freshly prepared TBS (100 mM Tris, pH 7.5, 137 mM NaC1, 5.1 mM KC1, 0.75 mM Na2PO4, 1.3 mM CaC12, and 0.49 mM MgC12). TBS solution was aspirated off followed by the addition of 1780 gl DNA/DEAE-Detran/TBS mixture (8 pg LIP cDNA expression vector/712 pl 0.1% DEAE-Detran in PBS/712 p l TBS, and 348 pl Tris-EDTA [TE, 100 mM Tris, pH 8.0 and

1 mM EDTA]) to the cells. Cells were incubated at room temperature inside a laminar flow hood. The plates were rocked every 5 min for 1 hr. At the end of incubation, the









76

DNA/DEAE-Detran/TBS mixture was aspirated off followed by a wash with 25 ml TBS and another wash with 25 ml PBS. Fresh medium was then added into plates after the aspiration of PBS and incubated at 370C in 5% CO2 humidified air for 24 hr. L2 cells transfected with LIP cDNA were experimental group, cells without transfecting with LIP cDNA were control group. After a 24 hr incubation period, the media was aspirated off followed by a PBS wash, and 1 ml of 0.25% (w/v) Trypsin and 20 mM EDTA was added into each plate for 2.5 min to detach the cells from the plate. Seven mls of fresh medium was then added into each plate. Plates (150 mm) from experimental and control groups were split into two 100 mm plates, and incubated for another 2 hr followed by the addition of LPS (final concentration = 0.5 ptg/ml) to one plate from each group for 4 hr. RNA was then isolated and examined for the mRNA levels of MnSOD. RNA Isolation and Northern Analysis

RNA Isolation. Acid guanidinium thiocyanate-phenol-chloroform extraction

method (Chomczynski and Sacchi 1987) with modifications was employed to isolate total RNA. Medium was removed from 100 mm plates followed by the addition of 3 ml of GTC solution (4 M guanidinium isothiocyanate, 25 mM sodium citrate, pH 7, 0.5% [w/v] sarcosyl, and 0.1 M P-mercaptoethanol). Cells were scraped off the plates and mixed with 0.1 volume of 2 M sodium acetate, pH 4.0, and then mixed with an equal volume of water saturated phenol, with 0.2 volume of chloroform-isoamyl alcohol (49:1) (v/v). After vigorously shaking for 10 sec, the samples were centrifuged at 10,000g for 15 min









77

at 40C. The aqueous phase was transferred to a polypropylene centrifuge tube and mixed with an equal volume of isopropanol followed by incubation at -200C for at least 1 hr. The sample was centrifuged at 10,000 g for 25 min. RNA pellet was dissolved in 500 ptl GTC solution and transferred to a diethylpyrocarbonate (DEPC) treated 1.5 ml microcentrifuge tube. RNA was precipitated by the addition of 500 [1 isopropanol and incubated at -200C for at least 1 hr. Sample was centrifuged in an Eppendorf centrifuge for 15 min at 40C. The RNA pellet was rinsed with cold 100% ethanol, air-dried, and resuspended in 300 -l of DEPC-water (0.1% DEPC, v/v) followed by ethanol precipitation, two times, with 0.1 volume of 3 M sodium acetate, pH 5.2, and 2.2 volume of 100% ethanol. The pellet was centrifuged for 15 min at 40C, rinsed with 100% ethanol, air-dried, and resuspended in 200 gt DEPC-water. The concentration of RNA was estimated by absorbance at 260 nm.

Northern Analysis. Fifteen microgram of total RNA was lyophilized and

resuspended in 25 [1 of loading buffer containing 20 mM morpholinopropanesulfonic acid (MOPS), pH 7.0, 6 mM sodium acetate, pH 7.4, 0.5 mM EDTA, 17.5 % (v/v) formaldehyde, and 50% (v/v) deionized formamide. The sample was incubated at 500C for 5 min followed by two separate incubations for 5 min at 650C. Five microliter of loading dye (0.3 jtg/gl ethidium bromide, 0.4% xylene cyanol, 0.4% bromphenol blue, 1 mM EDTA, and 50% [v/v] glycerol) was added to each sample. RNA was size fractionated on a 1% (w/v) agarose/2.2 M formaldehyde gel at 45 volts for 16 hr. After electrophoresis, the gel was soaked in 50 mM NaOH for 45 min followed by









78

neutralization in 100 mM Tris-HC1, pH 7.5 for another 45 min. The gel was then twice equilibrated in 50 mM TBE buffer (50 mM Tris-Borate, pH 8.3, and 0.05 mM EDTA) for each of 30 min. After equilibration, the gel was electrotransferred to a nylon membrane (Cuno). RNA was covalently crosslinked to the membrane by UV irradiation. Preparation of Random Primer Extension Probes

This method was used for making probes for Northern analyses of MnSOD, LIP, as well as cathepsin. One hundred nanograms of the appropriate DNA template was denatured by boiling for 5 min, and immediately incubated in a ice bath for at least 5 min. A buffer containing random primers, dCTP, dGTG, and dTTP (GIBCO) was added into the template solution followed by the additions of 100 ptCi [a-32P]-dATP and 10 units of Klenow DNA polymerase. The mixture was incubated at room temperature for 3-4 hr. The probe was separated by a Sephadex G-50 (in a buffer containing 10 mM Tris-HC1, pH 8.0, 1 mM EDTA, pH 8.0, and 750 mM NaC1) column. The 32P-labeled probe was boiled for 5 min after elution from the column. DNA templates were derived from appropriate restriction enzyme digestions of the rat MnSOD cDNA, rat LAP cDNA (kindly provided by Dr. Ueli Schibler at University of Geneva), or cathepsin cDNA.

Results

Identification of One Stimulus-Specific Binding Site

High resolution DNase I HS site studies by Dr. Jan-ling Hsu suggested that there existed important regulatory cis-acting elements in the promoter region for induced expression of rat MnSOD gene. I employed DMS in vivo footprinting and LMPCR





























Figure 3-1. Detection of guanine contacts specific to LPS, TNF-a, or flL- 13 exposure. In Vivo DMS footprinting of the top (-410 to -393) and bottom (-415 to -395) strands. In
(A), (B), and (C) I illustrate the LPS, TNF-a, and IL- 13-specific footprinting sites, respectively. Primer set K was employed for LMPCR for the top strand, and primer set D for the bottom strand. Lanes G are genomic protein-free DNA, and lanes C are in vivo DMS treated control cells. Lanes L, T, and I are in vivo DMS treated cells, previously exposed to LPS, TNF-L or IL-13, respectively. Filled circles represent enhanced guanine residues. The nucleotide positions relative to the transcriptional initiation site are illustrated on the left of the figure.









80

(A)


Top Strand Bottom Strand
in vivo in vivo
G G C C L L G G -C C L F
-410 .395


4W j4

-393 -415



(B)

-."'To p Strand Bottom Strand
in vivo in vivo
G G C C T T G G C C T T
-410 .395




-393



(C)

Top Strand Bottom Strand
in vivo in vivo
G G C C I I G G C C I I
-410 -395




-393 -415









81

to resolved these cis-acting elements at single nucleotide resolution. The positions of primer sets (D, E, F, K, L, and M) used in LMPCR are shown in Figure 2-2.

L2 cells were treated with LPS for 30 rin, 1 hr, 4 hr, or 8 hr, TNF-ct for 1 hr or 4 hr, or IL-10f for 4 hr. The same results were obtained for different stimulants and various time treatments. Figure 3-1 illustrates representative autoradiograms from samples treated for 4 hr with LPS, TNF-a, or IL-1P. Using primer sets #D and #K, I observed enhanced guanines at positions -404, and -403 on the top, and bottom strands, respectively. I also examined the promoter as far 5' as -720 bp and was unable to detect any further contact. Computer analysis of this region revealed a complete identity with the NF-IL6 consensus DNA binding sequence, 5'(A/C)TTNCNN(A/C)A, (Akira et al. 1990).

NF-cB Does Not Bind To The Rat MnSOD Promoter

NF-B was proposed as an oxidative stress-responsive transcription factor of

higher eukaryotic cells (Schreck et al. 1992). It is also one of the common transcription factors being activated by LPS, TNF-a, and IL-13 to induce the targeted genes. Das et al. (1995 a, b) reported that there is an associated relationship between the activation of NFKB and the elevated steady-state levels of MnSOD mRNA by TNF-cc or IL-I in lung adenocarcinoma (A549) cells. However, recently, Borello and Demple (1997) suggested that the induction of human MnSOD gene is NF-cB independent but parallel to the activity of AP-1. To address whether NF-idB plays a role in the induction of the rat MnSOD gene, I use computer analysis and found a putative NF-KcB binding site from








82











Top Strand Bottom Strand
in vivo in vivo
G GC C L L G G C C L L
-359 -338 __putative
'Wputative ...... NF-KzB site
-350 Le NFB site-355- i















Figure 3-2. Lack of NF-cB Binding on The Promoter Region of The MnSOD Gene. Primer set K was employed for LMPCR for the top strand (-359 to -350), and primer set D for the bottom strand (-355 to -338). Lanes are designated as in Figure 3-1. The same results were observed for TNF-a, or IL- 1 3 treated cells.









83

-353 to -344 in the promoter region of the MnSOD gene. The sequence of the putative NF-B binding site perfectly matches its consensus DNA binding sequence, GGG(G/A)(C/A/T)T(T/C)(T/C)CC (Lenardo and Baltimore 1989). However, I did not find any protein bound to this putative NF- KB binding site in vivo on either DNA strand of the MnSOD promoter as shown in Figure 3-2. Expression of Liver-Enriched Inhibitory Protein (LIP) in L2 Cells Does Not Affect The Induced Expression of The Rat MnSOD Gene

I observed enhanced guanines at positions -404, and -403 on the top and bottom strands, respectively, by employing genomic in vivo DMS footprinting. The flanking sequence of these two enhanced guanine residues matches the identity of NF-IL6, which I designated NF-IL6-like. I then attempted to evaluate the importance of NF-IL6 in the stimulus-dependent expression of MnSOD gene.

LAP (liver-enriched transcriptional activator protein), a rat NF-IL6 homologue,

has been cloned and characterized (Descombes et al. 1990). Although it is expressed in a variety of tissues, interestingly, the highest level of LAP mRNA was observed in lung. In addition, LAP/NF-IL6 was reported to be expressed at a low level in normal tissues with some studies indicating that, like the MnSOD gene, this regulatory factor was dramatically induced by LPS, TNF-a, or IL- I P (Akira et al. 1990; Akira et al. 1992). Furthermore, post-translational modification of LAP/NF-IL6, such as phosphorylation of a Ser residue(s) within its activation domain, has been shown to increase its affinity for its binding sequence, implying that de novo protein synthesis is not required for the stimulation of genes bearing the LAP/NF-IL6 recognition sequence (Trautwein et al.









84

1993; Trautwein et al. 1994). The activity of LAP/NF-IL6, therefore, may be regulated either at the transcriptional or post-translational levels. Previously, our laboratory demonstrated that de novo protein synthesis is not required for the regulation of the rat MnSOD gene by inflammatory mediators based on studies in which L2 cells were cotreated with cycloheximide and LPS, TNF-ct, or IL-103 (Visner et al. 1990). This further implicates LAP/NF-IL6 as a potential candidate transcription factor in the induction of the MnSOD gene.

I utilized a naturally existing dominant negative derivative of LAP/NF-IL6 known as LIP (Descombes and Schibler 1991; Buck et al. 1994). This natural protein is translated from an internal AUG thus generating a protein which lacks the putative transcriptional activation domain, but has the same DNA binding domain as LAP. LIP can then compete with LAP to bind to the same cis-acting element, but does not function as an activator since it lacks a transcriptional activation domain. LIP is thus thought to function within the cell as a dominant negative regulator (Descombes and Schibler 1991; Buck et al. 1994). Studies in L2 cells transiently transfected with a expression vector overexpressing LIP driven by a CMV promoter, however, resulted in no changes in basal or stimulated induction of the MnSOD gene as shown in Figure 3-3. Among the samples without transfected LIP plasmid, I also observed that there is significant basal LAP expression with only a minor induction, if any, in response to LPS (Figure 3-3).




























Figure 3-3. Overexpression of LIP in L2 cells did not affect the induction of MnSOD gene. Lanes 1-4 represent samples without transfecting LIP plasmid, and lanes 5-8 are samples transfected with 8 ptg of LIP plasmid. C represents control cells, and L represents cells exposed to LPS for 4 hr. After transient transfection of LIP plasmid and exposure of LPS, RNA was extracted and purified. The same samples were separated into two groups and loaded onto two separated gels and subjected to Northern analysis as described in Materials and Methods. Membranes were hybridized with MnSOD/cathepsin, or LAP/cathepsin cDNA probes. The LAP cDNA was kindly provided by Dr. Ueli Schibler at University of Geneva.






86



1 234 5 67 8
LIP
CL C LC LC L


MnSOD UW



Cathepsin owwo 0foto*

LAP U04ininA
LIP
Cathepsin ao ft i m*ao o









87

Discussion

MnSOD mRNA levels show an 18 23 fold induction after stimulation of L2 cells with LPS (Visner et al. 1990), similar results were observed on cells treated with TNF-a or IL-i. Studies with cycloheximide, an inhibitor of protein synthesis, showed no effect on LPS, TNF-a or IL-l-dependent induction of MnSOD mRNA level. On the other hand, L2 cells co-treated with stimulant and actinomycin, an inhibitor of mRNA transcription, inhibited the stimulus-dependent induction of MnSOD mRNA levels (Visner et al. 1990). Nuclear run-on studies demonstrated a 3-9 fold increase in nascent RNA transcription in response to these pro-inflammatory mediators. The above data suggest that the regulation of MnSOD gene expression is, at least, partially, transcriptionally dependent. Furthermore, Dr. Jan-Ling Hsu in our laboratory has identified a single LPS, TNF-L, or IL-i-specific hypersensitive subsite by using high resolution DNase I hypersensitive (HS) site analysis (Hsu, 1993). One of the transcription factors, NF--KB, has been shown to be utilized by LPS, TNF-u, and IL-I P to induce a variety of genes. I did not detect an in vivo NF-AB footprint on its putative binding site (-353 to -344) in the promoter region of the rat MnSOD gene. However, I observed two stimulus-specific enhanced guanine residues at positions -404, and -403 on the top, and bottom strands, respectively.









88

A Model of The In Vivo Promoter Architecture of The Rat MnSOD Gene

Based on chromatin structure studies of H~su (1993) and my genomic in vivo

footprinting data, I propose models invoking the chromatin structure change following the treatment of LPS, TNF-ax, or LL-l j. This model is shown in Figure 3-4. Figure 3-4

(A) shows a strong 5' boundary for hypersensitive (HS) site I as observed in control cells, whereas following stimulation the boundary is replaced with an additional HS subsite (Hsu 1993) as well as the detection of two stimulus-dependent enhanced guanine residues (Figure 3-1). It is possible that before cells are stimulated with LPS, TNF-L, or IL-ip, a phased nucleosome is positioned at or near the binding sequence for a transcription factor. This nucleosome is displaced following treatment with inflammatory mediators, presumably allowing the binding of the transcription factor and leading to both the enhanced guanine residues and the observed alterations in chromatin structure. Interestingly, Dr. Rich Rogers in our laboratory showed that the region containing the enhanced guanine residues I observed by using genomic footprinting is not functionally required for the induction of the rat MnSOD gene by employing a transient promoter/reporter system. It is therefore also possible that these enhanced guanine residues were caused by chromatin structure changes only, as shown in Figure 3-4 (B). A strong 5' boundary for HS site exists in control cells, whereas following stimulation, the boundary is replaced with an additional HS site (Hsu, 1993) as well as two stimulusdependent enhanced guanine residues (Figure 3-1). In summary, the current hypothesis,




























Figure 3-4. Models of the In Vivo Architecture of the MnSOD Gene Promoter. The spacing between each binding site is approximately scaled. The thin arrow represents the basal expression and the thick arrow represents the induced expression of the MnSOD gene.
(A). The top portion of this figure illustrates the presence of ten basal binding sites and also illustrates the potential presence of a phased nucleosomal boundary 5' to binding site 1. The bottom portion depicts the induced state of the MnSOD promoter. As illustrated, I have identified a potential NF-EL6-like stimulus specific binding site. A more open and accessible chromatin structure is evident following stimulation, thus allowing for stimulus-specific protein-DNA interaction.
(B). Alternatively, the stimulus-specific enhanced guanine residues are caused by the chromatin structure changes only without the involvement of protein-DNA interaction following the induction of LPS, TNE-a, or IL-I P.








90



BASAL
ii III Ivv vivilviii Ix x





LPS, TNF-cc, or IL- I INDUCED

I HHI Fvv vrviivila ix x









BASAL
I II in Ivv vivilvili Ix x





LPS, TNF-a, or IL- 1 INDUCED

II III Ivv vrviiviii Ix x CG









91

regarding the molecular mechanism that leads to stimulus-specific enhancement of DMS reactivity at these guanine residues, relates to their possible involvement in the observed alterations in chromatin structure (Hsu, 1993). It is possible therefore that the enhanced guanine residues detected in vivo reflect a chromatin structure which allows for proper access of the promoter by the transcription factors involved in enhancer activity. This could result from either the binding of a transcription factor or through changes in DNA structure which result in an enhancement of DMS reactivity, a situation that is important in vivo but may not be necessary in a transient promoter/reporter system. Is LAPJNF-1L6 The Stimulus-Specific Activator for The Induction of The Rat MnSOD Gene?

Methylation interference data for NF-1L6 has demonstrated that guanine residues in the central portion of the binding site for NF-I1L6 on the top and bottom DNA strands are important for its binding activity (Akira et al. 1990). This is consistent with my guanine enhancements seen in vivo, except that the in vitro data also predicts that two other guanine residues are also important for binding (Akira et al. 1990). I summarize the comparison of consensus DNA binding sequence for NF-1IL6 and its methylation interference with my in vivo data as followings:

My in vivo footprinting data identifies two stimulus-specific enhanced guanine residues, which are bolded and underlined:

-409ATTACGCCA
TAATGCGGT

The consensus DNA-binding sequence for NF-1IL6 based on Akira et al. (1990): 5' (A/C)TTNCNN(AIC)A