Regulation of manganese superoxide dismutase and inducible nitric oxide synthase gene expression


Material Information

Regulation of manganese superoxide dismutase and inducible nitric oxide synthase gene expression
Alternate title:
Regulation of manganese superoxide dismutase and nitric oxide synthase gene expression
Physical Description:
x, 108 leaves : ill. ; 29 cm.
Chesrown, Sarah Elizabeth, 1945-
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Superoxide Dismutase -- genetics   ( mesh )
Superoxide Dismutase -- metabolism   ( mesh )
Nitric-Oxide Synthase -- genetics   ( mesh )
Nitric-Oxide Synthase -- metabolism   ( mesh )
Gene Expression   ( mesh )
Gene Expression Regulation   ( mesh )
Enhancer Elements (Genetics)   ( mesh )
Rats   ( mesh )
Molecular Sequence Data   ( mesh )
Base Sequence   ( mesh )
Department of Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Biochemistry and Molecular Biology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1994.
Bibliography: leaves 98-107.
Statement of Responsibility:
by Sarah Elizabeth Chesrown.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002321349
oclc - 48668920
notis - ALS4876
System ID:

This item is only available as the following downloads:

Full Text







To my Mother and Father--

by whose example I learned about love,

sacrifice and commitment.


With pleasure and heartfelt gratitude I thank the

members of my graduate committee for their assistance and

support. The input provided by Drs. Purich, Boyce, McGuire

and Ferl made my graduate work both better and more

enjoyable. I am especially indebted to my scientific

mentor, Dr. Harry Nick, for his friendship, unfailing

support and encouragement, his skillful personal counsel and

scientific guidance during the past five years.

For encouraging me and helping me clear many hurdles

before even beginning this graduate work, I sincerely thank

Drs. Dan Purich, Bob Van Mierop, Harry Nick, Ian Burr,

Harvey Colten and Lee Dockery. I also gratefully

acknowledge the unwavering support given me toward

fulfilling this professional goal by Dr. Doug Barrett,

Chairman of Pediatrics. To my friend and colleague, Dr.

Gary Visner, I offer special thanks. Gary has quietly and

consistently been there to give advice, cover the clinical

service and support my efforts on occasions too numerous to


I appreciate all the assistance and encouragement given

me by the members of Dr. Nick's laboratory. Without their

teaching and generosity this work could not have been


accomplished. I especially thank Jan Ling Hsu, Joan

Monnier, and Maureen Dolan-O'Keefe for their kind help and


I extend a special thanks to my parents, Ruth and

Richard Chesrown for their love, patience, understanding and

faith in me. My brothers, Michael, Thomas and Griff have

encouraged me throughout my studies, and I am delighted for

this opportunity to express my appreciation to them.

To my dear friend, Jane Day, who has unselfishly given

me encouragement, wise counsel, a shoulder to lean on and

patient understanding, I give heartfelt, special thanks. I

will forever be in her debt for her continual support during

this effort. Additionally, I thank all my other friends and

colleagues for believing in me and for their unwavering

encouragement. I share my accomplishment with all these

special people.


ACKNOWLEDGMENTS ... ............ iii


ABSTRACT . . . ix



Free Radicals--Biological Roles . 1
Superoxide Dismutases . 4
Inducible Nitric Oxide Synthase . 6


Materials . ..... 8
Methods . . 11


Introduction . . 27
Results . . 34
Discussion . . 43

GENE . . . 48

Introduction . . .. 48
Results . . 53
Discussion . . 59


Introduction . ... 67
Results . . 69
Discussion . . 85


Inducible Nitric Oxide Synthase . 91
Manganese Superoxide Dismutase . 94

REFERENCES ....... .... ... 98

BIOGRAPHICAL SKETCH .................. 108


2.1 Transient transfection experimental time line

3.1 Structure, restriction map and DNase I
hypersensitive sites of the rat MnSOD gene .

3.2 Human growth hormone (hGH) expression vectors
and restriction maps . .

3.3 Comparison of hGH expression from a vector
containing a minimally active thymidine
kinase constitutive promoter to a vector
containing the MnSOD 5' regulated promoter .

3.4 Rat MnSOD 5' flanking region and deletions .

3.5 Levels of hGH expressed from rat lung
epithelial cells after transfection
with vectors containing portions
of the 5' flanking sequence of the
MnSOD gene . . .

3.6 MnSOD genomic sequence from positions -545
to +38 . . .

4.1 Expression vectors assessed for MnSOD gene
enhancer activity . .. ..

4.2 Levels of hGH expression from vectors assessed
for enhancer activity compared to pHind-Eag
vector containing MnSOD 5' promoter alone

4.3 Comparison of hGH mRNA levels 72 hours after
transfection . .

4.4 Cloning of deletion constructs from pJM17 GH
to localize enhancer activity .

4.5 Localization of the MnSOD enhancer to the
3.8 kb Hind III-Hpa I genomic fragment .

5.1 Dose response analysis of induction of iNOS
mRNA by LPS . .

. 20

. 31

. 33

. 36

. 40

. 42

. 46

. 52

. 55

. 57

. 61

. 63

. .71


5.2 Time course of induction of iNOS mRNA by LPS 74

5.3 Northern analysis of RNA from unelicited
rat peritoneal macrophages exposed
to LPS . .. 76

5.4 Dose response of actinomycin D on LPS
induction of iNOS mRNA in RAW 264.7 cells 79

5.5 Dose response of cycloheximide on LPS induction
of iNOS mRNA in RAW 264.7 cells . 81

5.6 Densitometric analysis of IL-10 and TGF-f
modulation of LPS and INF-7 induction
of iNOS mRNA in RAW 264.7 cells . 84


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



Sarah Elizabeth Chesrown

December, 1994

Chairperson: Harry S. Nick
Major Department: Biochemistry and Molecular Biology

Both superoxide anion (02-) and nitric oxide (NO-) are

free radicals, possessing oxygen atoms with unpaired

electrons. They can mediate oxidative damage of cellular

proteins, lipids, and DNA, and assist phagocytes in killing

microorganisms and tumor cells. To minimize injury, cells

enzymatically regulate the synthesis of NO-, and the

catalysis of 02". These studies examined the regulation of

two key enzymes in this scheme: inducible nitric oxide

synthase (iNOS), and manganese superoxide dismutase (MnSOD).

By northern analysis, iNOS mRNA was undetectable in

unstimulated murine macrophages, but its induction by

bacterial lipopolysaccharide (LPS) was rapid and dramatic,

and was blocked by inhibition of new protein or RNA

synthesis. Expression of iNOS mRNA was induced by

interferon-y and interleukin-10, and combinations of these

cytokines, with or without LPS, synergistically elevated

iNOS mRNA. In contrast, iNOS mRNA remained undetectable in

human monocytes.

To identify cis-acting sequences regulating MnSOD

expression, transient transfection studies in rat lung

epithelial cells and deletion analysis of the MnSOD 5'

flanking sequence were used. MnSOD transcription starts 74

bp 5' to the first ATG signal (+1), and sequence from -227

to -42 was sufficient for both basal and inducible

expression of the reporter gene. After shortening the MnSOD

sequence to 77 bp (from -119 to -42) expression was

undetectable. LPS and tumor necrosis factor-a (TNF-a)

stimulated expression about 2.5 fold. The GC-rich MnSOD

promoter, from -227 to -42, contains 2 DNase I

hypersensitive sites, but no consensus sequences for known

transcription factors.

Adding 6.1 kb of MnSOD genomic sequence to an

expression vector containing 2.5 kb of 5' flanking sequence

had no effect on basal expression, but resulted in a 5-fold

induction by LPS, TNF-a, and interleukin-la (IL-la). This

MnSOD enhancer functioned well only in the 5' to 3'

orientation. Using deletion analysis, the enhancer activity

was localized to a 3.8 kb fragment from position +1105 to

position +4941. This enhancer may be novel with respect to

the combination of its orientation dependence, location

within a gene, and responsiveness to LPS, TNF-a and IL-la.


Free Radicals--Biological Roles

A free radical is any molecule that contains one or

more unpaired electrons. The two free radicals relevant to

this discussion, superoxide (02) and nitric oxide (NO), are

short lived, highly diffusable and reactive gases that are

widely distributed throughout our oxygen rich environment.

In living organisms both 02' (Maly et al., 1993; Smith and

Curnett, 1991) and NO (Karupiah et al., 1993; Flynn et al.

1993; Green et al., 1991; Lin et al., 1994) have been shown

to have important homeostatic roles in defense against

infection. However, if allowed to accumulate in excess, 02'

(Freeman and Crapo, 1982; Kehrer, 1993) and NO (Kilbourn et

al., 1990; Bredt and Snyder, 1994) become highly toxic.

Superoxide anion, or its reaction products following

association with a proton or metal cation, can cause

oxidation of sulfhydryl enzymes (Freeman and Crapo, 1982),

lipid peroxidation (Svingen et al., 1978) and base

modification or strand scission of DNA (Brown and Fridovich,

1981). Such O{ mediated tissue injury is thought to occur

during aging, inflammation, ischemia-reperfusion, and

breathing hyperoxic gas mixtures. NO excess may help

mediate the cardiovascular collapse that accompanies septic

shock (Kilbourn, et al., 1990a). In addition, these two

free radicals can react with each other to form another

toxic anion, peroxynitrite (ONOO') that may result in neural

and endothelial damage (Beckman et al., 1990; Beckman et al,

1993; Beckman et al. 1994).

One electron reduction of oxygen produces the

superoxide radical, 02, and it is formed as a by-product of

mitochondrial respiration primarily at the sites of

ubiquinone-cytochrome b (Boveris, 1977) and NADH-

dehydrogenase (Turrens and Boveris, 1980). In addition,

specialized eukaryotic cells have capitalized on the toxic

effects of this free radical. Phagocytic cells, including

neutrophils, eosinophils, monocytes and macrophages, employ

the enzyme NADPH oxidase to produce large quantities of

superoxide to help kill certain types of bacteria (Segal and

Abo, 1993). An inherited inability to generate superoxide

by these cells results in the disorder known as chronic

granulomatous disease (Dinauer, 1992), an immune deficiency

disorder characterized by recurrent lung infections and a

fatal outcome. Large numbers of these phagocytic cells are

recruited to every site of inflammation in the body to

remove necrotic debris. As a result, high local levels of

superoxide free radicals are generated that are potentially

toxic to neighboring healthy cells.

Certain pathogenic microorganisms escape the toxic

effects of superoxide and are well adapted to surviving and

multiplying inside phagocytic cells. These organisms

include some important human pathogens such as Mycobacterium

tuberculosis that causes tuberculosis, Plasmodium species

that cause malaria, and Listeria myocytogenes, an organism

that can cause sepsis and meningitis in newborns and

immunodeficient individuals. To help combat these

infections, macrophages and hepatocytes enzymatically

elaborate the free radical nitric oxide, NO (Green et al.,

1991; Wood et al., 1993; Liew et al., 1990; Evans et al.,

1993; Geller et al., 1993; Geller et al. 1993).

In addition to nonimmune host defense, many other

functions have been identified for superoxide and nitric

oxide free radical molecules. Both have demonstrated anti-

tumor activity (Wong et al., 1989; Lorsbach et al., 1993).

Nitric oxide also functions as a local intercellular second

messenger. In both the central and peripheral autonomic

nervous system NO is a neurotransmitter (Nathan, 1992;

Desai et al., 1991). Within the cardiovascular system

nitric oxide is a potent endogenous vasodilator produced by

endothelial cells that relaxes underlying vascular smooth

muscle (Kilbourn, 1990b).

Superoxide also forms as a result of electrons

"leaking" onto oxygen from various components of the

electron transport chains, such as those of the mitochondria

and the endoplasmic reticulum (Boveris, 1977; Turrens and

Boveris, 1980). The amount of leakage is increased as the

local concentration of oxygen is raised, resulting in one

source of oxidative stress for the cell. In the medical

setting, oxygen is commonly increased--either with the

administration of hyperoxic inspired gas mixtures or

following an interruption-then restoration of tissue

oxygenation. The resulting excess superoxide and its

reaction products are the basis of hyperoxia-induced

pulmonary toxicity and ischemia-reperfusion injury.

Superoxide Dismutases

The cellular components susceptibile to injury by

superoxide mediated oxidation include lipids, DNA, and

proteins. To protect against toxic effects of superoxide

and its metabolites, eukaryotic cells possess anti-oxidant

enzyme systems that include the superoxide dismutases

(Harris, 1992). These metalloenzymes catalyze the reaction

of superoxide with hydrogen ions, forming hydrogen peroxide

and oxygen, and thereby accelerating the destruction of

superoxide by about four orders of magnitude (Halliwell and

Gutteridge, 1990).

There are two intracellular forms of superoxide

dismutase and one extracellular form. Both copper and zinc

are located in the active site of the extracellular SOD and

the cytosolic Cu/ZnSOD, while manganese is in the active

site of the mitochondrial SOD. The regulation of

extracellular SOD has not been well studied, but our studies

as well as others have demonstrated that cytosolic Cu/ZnSOD

is abundantly and constitutively expressed in eukaryotic

aerobic cells (Visner et al, 1989; Dougall and Nick, 1991).

In contrast, expression of mitochondrial MnSOD is highly

regulated by mediators of inflammation such as bacterial

lipopolysaccharide (LPS), and cytokines including

interleukin-1 (IL-1) and tumor necrosis factor-alpha (TNF-a)

(Visner et al., 1989; Shull et al., 1991). The in vivo

importance of MnSOD in protecting pulmonary epithelium from

hyperoxic injury was directly demonstrated by Wisp& et al.

(1992) who produced transgenic mice in which the expression

of human MnSOD mRNA was directed by transcriptional elements

from the human pulmonary surfactant protein C gene. Human

MnSoD was expressed in a lung-specific manner in the

mitochondria of alveolar type II cells and nonciliated

bronchiolar cells. The activity of MnSOD, but not other

antioxidant enzymes, was increased in the transgenic mice

compared to control mice. When challenged by continuously

breathing 95% oxygen, the survival of the transgenic mice

was dramatically improved compared to nontransgenic


Understanding the molecular mechanisms that regulate

MnSOD gene expression has been a goal of this laboratory

since 1987. After cloning and characterizing the rat MnSOD

cDNA and genomic clones, Drs. Nick, Visner, Dougall and Hsu

characterized the induction of MnSOD mRNA steady state

levels in many cell types including rat pulmonary

endothelial and epithelial cell lines and rat hepatocytes.

Using nuclear run on studies Dr Hsu (1993) confirmed the

induction in MnSOD mRNA levels involves increased

transcription of the gene. Furthermore, results of her

DNase I hypersensitive studies on MnSOD chromatin in vivo

identified regions of the gene with increased nuclease

accessibility. These regions are sites of potential

regulatory protein-DNA interaction. My studies have

extended these observations by functionally defining regions

that regulate MnSOD gene expression by means of their

promoter and enhancer transcriptional activity.

Inducible Nitric Oxide Synthase

A family of isoenzymes, known as nitric oxide synthases

(NOS), produce the free radical nitric oxide in many

mammalian species including rodents and humans (Moncada and

Higgs, 1991). Two major classes of NOS are defined based on

whether their expression is 1)transcriptionally regulated by

mediators of inflammation and cytokines induciblee NOS,

iNOS) or 2)they are constitutively expressed, and post-

translationally activated by changes in intracellular

calcium levels (constitutive NOS, cNOS). The iNOS isoform

was originally cloned from a murine macrophage cell line

(Lyons, et al., 1992; Xie, et al, 1992) and shown to

significantly enhance macrophage function of killing

intracellular pathogens and tumor cells (Green et al., 1991;

Cunha et al., 1993).


Because excessive quantities of 02o and NO are toxic to

cells and because many of the same mediators of inflammation

and pro-inflammatory cytokines induce the expression of both

iNOS and MnSOD, I characterized the regulation of the iNOS

mRNA levels in murine macrophages.

Furthermore, it has been reported by Beckman (1990)

that the simultaneous production of superoxide and nitric

oxide, and subsequent formation of their combined product

peroxynitrite, at sites of infection and/or inflammation

might have deleterious effects on cells. If so, it might be

physiologically important that the expression of iNOS and

MnSOD be co-regulated and evolutionarily preserved in order

to prevent the simultaneous accumulation of both free

radicals. Winterbourn (1993) has suggested that electrons

from many different free radicals are transferred to oxygen,

yielding 02', and the subsequent removal of 0, by superoxide

dismutases is an important cellular mechanism to reduce

oxidative stress. Conversely, Oury et al. (1992)

demonstrated an increase in 02 toxicity in mice

overexpressing extracellular superoxide dismutase. Liochev

and Ftidovich (1991) also reported that the overproduction

of MnSOD superoxide dismutase in Escherichia coli blocks

superoxide induction of the soxR regulon resulting in an

increased sensitivity to paraquat, a compound that generates

superoxide anions. Thus, we were interested in examining

the regulation of iNOS mRNA by those inflammatory mediators

known to increase MnSOD expression.




Restriction endonucleases, T4 DNA ligase, calf

intestinal alkaline phosphatase, mung bean nuclease, and

Klenow, the large fragment of E.coli DNA polymerase, were

purchased from New England Biolabs. From Sigma Chemical Co.

we obtained cycloheximide, actinomycin D, an antibiotic/

antimycotic system (ABAM), ampicillin, DEAE dextran, trypsin

[2.5% (w/v)], glutamine, diethyl pyrocarbonate (DEPC),

chloroquine, Trizma base, proteinase K, lipopolysaccharide

from E. coli serotype 055:B5, bovine serum albumin (A-7511),

Dulbecco's Modified Eagle Medium (DMEM), RPMI-1640 medium,

DNase-free RNase, and Histopaquee-1077. Fetal bovine serum

(FBS) came from Intergen, and sodium pentobarbital, for

general anesthesia was purchased from Butler Labs.

Deoxymdenosine triphosphate alpha-32P, [a 2P]-dATP, was

purchased from ICN Biochemicals. For gel electrophoresis,

we obtained Seachem HGT agarose from FMC Bioproducts.

Ham's Modified (F12K) tissue culture medium came from

Gibco/BRL Life Technologies along with the random primers

DNA labeling system. We purchased cesium chloride

(Ultrapure Bioreagent) from VWR, and from Qiagen we

purchased columns containing anion exchange resin for

additional purification of all plasmid DNA used in

transfection experiments. Allegro human growth hormone

transient gene expression assay systems were purchased

Nichols Institute Diagnostics. Charged and noncharged

Zetabind nylon membrane was obtained from CUNO Laboratory

Products. Autoradiography film, Hyperfilm-MP, was

purchased from Amersham.

Cytokines were obtained as follows: recombinant murine

interferon-y (INF-y), Genzyme; recombinant human tumor

necrosis-a (TNF-a), Genentech; recombinant human

interleukin-1t (IL-1f), National Cancer Institute;

recombinant human interleukin-la (IL-la, Hoffman-La Roche;

recombinant human interleukin-10 (IL-10), R&D Systems; and

human recombinant transforming growth factor-fl (TGF-j3), was

a gift from Dr. Greg Schultz at the University of Florida.

Animals, Cell Lines and Primary Culture Cells

L2 cells, a rat lung epithelial cell line, were

purchased from American Type Tissue Collection (ATCC CCL

149). They were grown in F12K medium supplemented with 10%

fetal bovine serum, ABAM, and 4 mM glutamine at 37C, 5% CO2.

Only cells from passages 15 to 29 were used in these

experiments. The RAW264.7 and J774 murine macrophage cells

were also purchased from ATCC. They were grown in DMEM

tissue culture medium supplemented with 10% fetal bovine

serum, ABAM and 4 mM glutamine at 37C, in room air and 5%


Non-elicited, resident rat peritoneal macrophages were

isolated from male Harlan-Sprague-Dawley rats weighing

between 150-200 grams, purchased through the University of

Florida Division of Animal Resources. Macrophages were

isolated by repeated peritoneal lavage with 10 ml sterile

PBS immediately after the rats were anesthetized with 64.8

mg (1.0 ml) intraperitoneal sodium pentobarbital. The cell

suspension was placed on ice, spun at 600 g for 10 min and

the pellet resuspended in supplemented DMEM. Cells were

plated at 106 cells per 60mm plastic dish and macrophages

allowed to adhere for 3 hour at 37C, 5% CO2. Plates were

then washed to remove nonadherent cells and 4 ml fresh DMEM


Human peripheral blood monocytes were isolated from

heparinized samples of blood drawn from a peripheral vein.

Heparinized blood was diluted 1:1 (v/v) with Hanks buffered

saline solution (HBSS), layered over cold ficoll-hypaque and

spun at 800 g for 20 minutes at 40C. The mononuclear cell

layer-was collected, diluted with an equal volume of HBSS

and centrifuged at 400 g for 10 minutes. The cell pellet

was washed 3 times with HBSS and after each wash, the cells

were recollected by centrifugation at 400 g. Cells were

resuspended at a concentration of 5-8 x 106 cells per ml in

RPMI 1640 medium supplemented with 10% FBS. To enrich for

monocytes/macrophages and deplete of lymphocytes, 6 ml of

the cell suspension was added per 60 mm tissue culture

plate, and incubated for 1 hour at 370C, 5% CO2 to allow the

monocytes/macrophages to adhere to the plastic. Plates were

washed twice with HBSS in order to remove non-adherent cells

and replaced with 4 ml RPMI supplemented with 10% FBS, ABAM

and glutamine, 4 mM final concentration.

Molecular Probes

The 1.5 kb rat manganese superoxide dismutase (MnSOD)

cDNA clone used for synthesizing molecular probes was

isolated and cloned into the EcoR I site of pUC 19 by Dr.

Wm. Dougall working in Dr. Nick's laboratory (Dougall and

Nick, 1991; Hurt et al., 1992). The rat cathepsin B cDNA

(0.6 kb fragment cloned into the EcoR I site of a pUC

vector) was a gift from Dr. J. S. Chan, University of

Chicago; and the inducible nitric oxide synthase (iNOS) cDNA

probes were derived from a 1.89 kb Nco I fragment digested

from a murine cDNA clone that was a gift from Dr. James

Cunningham, Harvard Medical School. The vector containing

the human growth hormone cDNA was purchased from ATCC and

subcloned into pUC 19 for amplification.


Cloning the 5' Non-coding Sequence of the MnSOD Genomic
Clone into a Promoterless Human Growth Hormone Expression

In order to assist the reader, two numbering systems

will be used to designate positions in the MnSOD genomic

sequence. In the first system, the A in the first ATG

signal is +1, and all bases residing 5' to +1 have

negatively numbered positions, with -1 denoting the first

base 5' to +1. The second numbering system is based on

sequence data reported in the EMBL, GenBank, and DDBL

Nucleotide Sequence Databases under the accession number

X56600 RAT SOD-2 GENE (Ho et al.,1991). This second

numbering system designates the Hind III site located 2500

bp 5' to the start of transcription as #1. I included this

system in an attempt to facilitate the readers' use of

database sequence and database-generated restriction maps.

For example, the transcription start site may be denoted

either as -74 or 2499BL (Hurt et al., 1992).

Initially, the entire 17 kb MnSOD genomic clone (Figure

3.1) containing 1) 5 exons and 4 introns (6.5 kb), 2) 5'

non-coding sequence (4.5 kb) and 3) non-coding sequence 3'

to the last exon (6 kb), was removed from a pUC 19 plasmid

vector by digestion with EcoR I (Dougall, 1990). This

linear MnSOD genomic fragment was then digested with the

restriction enzyme Eag I, creating two fragments: 1)an

Eco/Eag fragment, from the 5' EcoR I site about 4.5 kb

upstream from the Eag I site at position -42 (2531mL), and

2)a 12.5 kb fragment from the Eag I site to the 3' EcoR I

site. The 4.5 kb Eco/Eag fragment was isolated by gel

electrophoreseis, electroeluted into Tris-buffered saline

(TBE), concentrated using secondary butanol, and cleaned by

separate extractions with equal volumes of 1) distilled,

Tris-saturated phenol, 2)a 50-50 mixture of phenol and

chloroform/isoamyl alcohol (24:1), and 3) the

chloroform/isoamyl alcohol mixture alone. Following

precipitation with three volumes of 100% ethanol and 2.5 M

ammonium acetate, the Eco/Eag fragment was resuspended in

Tris/EDTA buffer (TE), pH 8, 10 mM/1 mM respectively.

The 5' overhang ends of the Eco-Eag fragment were then

filled-in using DNA Polymerase I Large Fragment. Briefly,

50 pg of ethanol precipitated Eco-Eag fragment was dissolved

in one ml of lx E. coli PolymeraseI/Klenow Buffer [10

mMTris-HCl (pH 7.5), 5 mM MgCl2, 7.5 mM dithiothreitol]

supplemented with 33 pM each dNTP. Fifty units of Klenow

were added and the reaction incubated for 15 min at room

temperature. The reaction was stopped by adding EDTA to a

final concentration of 10 mM and heating to 750C for 10

minutes. The resulting blunt-end Eco-Eag fragment was

cloned into the Hinc II polylinker site in the promoterless

human growth hormone expression vector, pOGH, from Nichols

Institute (Selden et al., 1986). This cloning created a 9.3

kb plasmid hereafter referred to as pEco-Eag GH (Figure

3.2). It contained 4.5 kb of the 5' non-coding region of

the MnSOD gene with possible promoter activity in front of

the human growth hormone gene in a pUC 13 based vector.

Using transient transfection studies in L2 rat lung

epithelial cells, this construct was functionally assessed

for basal expression of the reporter gene product, hGH, and

for increased expression of hGH following treatment with

LPS, IL-la, or TNF-a.

Cloning Deletions of the 5' Flanking Sequence of MnSOD
in PD GH

To localize the cis-acting regions responsible for

promoter activity within the MnSOD Eco-Eag genomic fragment,

I used unique restriction enzyme sites to delete portions of

the MnSOD Eco-Eag 5' flanking sequence cloned into pO GH

(Figure 3.4). Each resulting vector, containing less of the

MnSOD 5' sequence, was then assessed in transient

transfection studies for its ability to drive basal and

stimulated hGH expression.

Digestion of pEco-Eag GH with the restriction enzyme,

Hind III, deleted a 2.1 kb fragment--from the Hind III site

in the ppGH polylinker (5' to the MnSOD sequences) to the

second Hind III site at position -2573 (laML)in the MnSOD

Eco-Eag fragment. Following religation of the Hind III ends

with T4 DNA ligase, the resulting 7.2 kb plasmid, called

pHind-Eag GH, contained 2.5 kb of the 5' flanking sequence

of the MnSOD gene. This plasmid was amplified in a large

scale plasmid preparation, purified both by

ultracentrifugation through a CsCl density gradient and by a

Qiagen anion exchange column. The pHind-Eag GH vector was

then assessed for promoter activity in transient

transfection studies.

All subsequent deletion constructs were created

directly form the pHind-Eag GH vector. The pHind-Eag GH

vector was shortened 922 bp by digestion with Hind III and

Sal I restriction enzymes. Both 5' overhang ends were

filled-in as described above, and the shortened construct


ligated. The resulting 7.1 kb plasmid, called pSal-Eag GH,

contained about 1.6 kb of the MnSOD gene 5' sequence from

the Sal I restriction enzyme site at -1651 (922EmL). This

vector was assessed in transient transfection experiments

for its ability to promote basal expression and to increase

expression in response to LPS, TNF-a and IL-la treatment.

Similarly, pHind-Eag GH was shortened by 2208 bp,

following digestion with Hind III and Sfi I. The Sfi I

restriction site is located at position -365 (2208ML).

Since Sfi I digestion leaves a 3' overhang, and since the

Klenow fragment can only fill-in 5' overhangs, the ends of

the vector could not be blunted by this DNA polymerase.

Therefore, the linear pSfi-Eag GH vector was treated with

mung bean nuclease, an enzyme that removes single-stranded

extensions (3' and 5') to create blunt ends. Briefly, 5 Ag

of linearized plasmid DNA was mixed with 1 unit of mung bean

nuclease in 11 Al of NEB Buffer #2. Following incubation at

300C for 30 min, the reaction was stopped by extractions

with phenol/chloroform/isoamyl alcohol, and the DNA

precipitated with glycogen and ethanol. To assure that any

possible remaining single-stranded 5' extensions were

eliminated, the plasmid was filled-in with a Klenow reaction

as described above before being ligated to create a 5.8 kb

pSfi-Eag GH vector containing 1024 bp of MnSOD 5' sequence.

Similarly, digestion of pHind-Eag GH with the enzymes

Hind III and Sac 2 created a linear Sac 2-Eag GH vector with

one 5'overhang from Hind III and one 3' overhang from Sac


II. The Sac II restriction site is located at position -227

(2246EmL). This linearized vector was also treated with

mung bean nuclease followed by Klenow in order to create

blunt ends for ligation. The resulting 4.857 kb pSac II-Eag

I vector contained 188 bp of MnSOD sequence. Finally,

pHind-Eag GH was digested with Hind III and the blunt-end

cutting Nae I restriction enzyme at position -119 (2454mmL).

The Hind III half-site was filled-in with a Klenow reaction,

and the shortened vector ligated to create pNae-Eag GH.

This 4.748 kb circular vector contained just 79 bp of the

MnSOD 5' flanking sequence upstream from the hGH gene.

Cloning of the MnSOD Hind III Fragment from Positions +1107
(3679mEO)to +7238 (9811m) into pHind-Eac GH

The 6.1 kb Hind III MnSOD genomic fragment, from

position +1107 to +7238, contains 5 of the 6 remaining DNase

I hypersensitive sites (#3 through 7) identified by Dr. Hsu

(1993). To test this fragment for possible interaction with

elements in the 5' sequence resulting in enhancer activity,

the Hind III fragment was digested with Hind III from a pUC

vector into which it had been subcloned. The 6.1 kb Hind

III fragment was isolated from a 1% agarose gel and cloned

in both orientations into the Hind III polylinker site of

pHind-Eag GH (Figure 4.1). This site is located 5' to the

Hind-Eag MnSOD insert. Expression of hGH from the resulting

13.45 kb vectors was compared to expression from pHind-Eag

GH to determine whether the Hind III insert contained

sequences with promoter, enhancer or repressor activity.

The 13.45 kb construct with the Hind III insert cloned in

the 5' to 3' orientation was named pJM17 GH, while the

construct with the Hind III insert in the opposite

orientation was named pJM13 GH.

Cloning Deletions of DJM17 GH and DJM13 GH

In order to begin to localize the regions) within the

Hind III insert responsible for the observed enhancer

activity, I utilized a unique Hpa I restriction endocuclease

site, located at position +4938 (7511EmL). Hpa I digestion

thus linearized pJM17 GH and pJM13 GH near the middle of the

6.1 kb Hind III insert (Figure 4.4). Subsequent partial

digest of these linear plasmids with Hind III resulted in 4

new, shortened vectors: 1) an 11.1 kb construct derived from

pJM17 GH containing sequence from the Hind III site at

position +1106 (3679,,mL) to the Hpa I site at +4938

(7511EML) (pSC17/11 GH); 2) a 9.6 kb construct derived from

pJM17 GH containing sequence from the Hpa I site to the Hind

III site at +7238 (9811EML) (pSC17/9 GH); 3) a 9.6 kb

construct derived from pJM 13 GH containing sequence from

the Hind III site at +7238 (9811,ML) to the Hpa I site at

+4938 (7511EmBL) (pSC13/9); and 4) an 11.1 kb construct

derived from pJM13 GH containing sequence from Hpa I site to

the Hind III site at position +1106 (36790.L) (pSC13/11 GH).

Two of these, pSC17/11 GH and pSC17/9 GH) were assessed for

promoter or enhancer activity compared to pHind-Eag GH and

pJM 17 GH in transient transfection experiments.

Transient Transfection Studies

The experimental protocol for the transient

transfection experiments is summarized in Figure 2.1. To

introduce hGH expression vectors into L2 cell monolayers, I

used the DEAE dextran mediated transfection method described

by Kriegler (1990). L2 cells at a passage between 15 and

29, were grown to about 75% confluence in 4 to 6 tissue

culture flasks (75cm2). The monolayers were removed with

0.25% (w/v) trypsin and combined in a suspension at 6 x 105

cells/ml. Cells were divided into 48 to 60 tissue culture

plates (60mm2) at 6 x 105 cells per plate. When the cells

had grown to 40-70% confluence, the medium was aspirated and

the cell monolayers washed with 4 ml phosphate buffered

saline (PBS) followed by 4 ml of tris-buffered saline (TBS).

To each 60 mm2 tissue culture plate was applied 180 pl of a

solution containing 5 yg (5 Ml) plasmid DNA, 31 Al TE pH 8.0

(100 mM Tris base and 1 mM EDTA), 72 pl TBS, and lastly, 72

Al of a 0.1% (w/v) solution of DEAE dextran in TBS (final

DEAE dextran concentration of 0.04%). Plates were incubated

at room temperature inside a laminar flow hood for one hour

and rocked every 5 minutes to assure that cells would not

dehydrate. The DNA-containing DEAE dextran solution was

then aspirated and the cells were washed with 4 ml TBS

followed by 4 ml PBS. The PBS was replaced with 4 ml of

F12K medium containing 10 MM chloroquine in addition to 10%

fetal bovine serum, ABAM and 4 mM glutamine and plates

incubated 4 hours at 370C in room air containing 5% CO2.

0 0 >i 5 >iU
0 (0 4 04 0
*1 14) 0V 4) V
4)) 0 4) 0 0

0010 8 M-
03 0 4 *I *?

--0 0 4 P
. r-4 0 I 0 0
r H1) 0 00 V 0
4) df 0 44 0 a )o E
*-A r- 0 (0 V I1 4
4- 04 4i -f ,-4 -H tO
V WA 0 V D Z O) )
r- 3 4)4 X E- A
CO0 0V 0 4) o) 0

e to 4 e 0 ) w 0 0

--I ) 00 W V 0 4)
ow %~ 0 40
d ^4 0 *'-A 0 -4 0
U0 C) U 004) V
C W 0 4-4 00 00I4-
0 k 0) 0 0 *4-444 00
00 3 000W 0 044
VV W 4) rC(O 4 ) r. W 4

) -H -H M 4 ) V M )
0 V (A 0 V V) V V 0 M
to r-44 V ) 0 > <44)
to10 4) 4) 0C 0 ) 0r 3

Sto H o 44M ) V 0
4J IfA 0 t 0 & M -H IV

V -4 : 400 0- 04)0
W*H 0 00 03 0-

HC *.4W4O
r.4 n 41 ON 4n
0 C B4 -'0 4C WO> o
k r-H .- 4 ) C4 0 M 1
E-1 X 4 0 C4 V 1 0 V H (1'+-
FA v 0 M H > 4) I4- 0 r-4 0

kaO4) *u 3 0 0kV


< -i


This medium was removed and replaced with identical medium

but without chloroquine (time zero). Twenty-four hours

later the inflammatory mediators LPS, TNF-a, or IL-la, were

added to triplicate or quadruplicate plates at final

concentrations of 2.5 gg/ml, 25 ng/ml and 5 ng/ml

respectively. Unless otherwise specified, 48 hours later

(72 hours after removing the chloroquin-containing medium)

one ml of medium was removed from each plate and frozen at -

700C until assayed for hGH.

Statistical Analysis of Transient Transfection Data

To determine whether treatments with inflammatory

mediators increased expression of the hGH reporter gene, the

mean concentration of hGH from plates transfected with each

vector and treated with LPS, TNF-a or IL-la was compared to

the mean concentration of hGH from untreated plates

transfected with the same vector using Student's unpaired,

two-tailed t test. A p value of <0.05 was considered

significant. In addition, a two-way analysis of variance

was performed to test for differences in basal and

stimulated hGH expression between vectors.

Radioimmunoassay Measurement of hGH Concentration

The hGH radioimmunoassay kit from Nichols was used for

measurement of hGH concentration. Samples of medium and all

reagents were warmed to room temperature. Into duplicate

tubes, 100 Al of either a hGH standard (0, 0.45, 1.4,and 5.0

ng/ml) or thawed medium from a plate of transfectd cells was


aliquoted. To this, 100 Al of an antibody mixture was added

containing 2 mouse monoclonal anti-hGH antibodies. The

antibodies, binding to different epitopes on hGH, contained

covalently bound 125 or biotin. After mixing, one avidin-

coated polystyrene bead was added to each tube and incubated

for 90 minutes at room temperature on a horizontal rotator

at 170 rpm. Beads were then washed twice with 2 ml of wash

solution and counted in a gamma counter for one minute.

Counts from duplicate samples were averaged, the average

counts from the zero standard subtracted, and the

concentration of hGH in the experimental samples was

calculated from the standard curve. The standard curve for

this assay is linear for hGH concentrations between zero and

15 ng/ml. According to the manufacturer, variations in

protein concentration of 25% have no detectable effect on

the value of hGH obtained. The sensitivity of this assay is

defined as the smallest single value that can be

distinguished from zero at the 95% confidence limit.

According to the manufacturer, the Nichols hGH Kit has a

calculated sensitivity of 0.06 ng/ml. The precision (intra-

assay-variance) of this assay was reported from replicate

determinations on each of three quality control sera in a

single assay (n=20). The coefficient of variation was 3.2--

4.3% for hGH concentrations from 3.6 to 19.5 ng/ml. The

reproducibility (inter-assay variance) was reported by

Nichols from data obtained for three quality control sera

assayed during a one month period (n=24). The coefficient


of variation ranged from 6.9 to 7.6% for hGH concentrations

from 3.8 to 18.0 ng/ml.

RNA Isolation and Northern Analysis

Total RNA was isolated from cell monolayers by the acid

guanidine thiocyanate-phenol-chloroform extraction method

described by Chomczynski and Sacchi (1987) and as modified

by Visner (1989). Briefly, to isolate total RNA from cells

in tissue culture, the medium was removed and cells lysed

with, 3 to 4 ml of GTC solution [4 M guanidinium

isothiocyanate, 25 mM sodium citrate, pH 7.0, 0.5% (w/v)

sarcosyl, and 0.1 M 2-mercaptoethanol]. The cell lysate was

scraped into a tube and extracted with 0.1 volume of 2 M

sodium acetate, pH 4, an equal volume of water-saturated

distilled phenol and 0.2 volume of chloroform:isoamyl

alcohol (49:1) (v/v). The final mixture was shaken

vigorously for 15 to 30 sec, incubated on ice for 15

minutes, then centrifuged at 10,000 g for 25 minutes at 40C.

The top aqueous phase was then transferred to a fresh tube

and an equal volume of isopropanol added to precipitate the

RNA. ,Following incubation at -20C for 2 one hour, the

tubes were centrifuged at 10,000 g for 30 minutes at 4C.

The RNA pellets were dissolved in 0.5 ml GTC solution and

transferred to individual eppendorf tubes that had been

treated with DEPC. The RNA was again precipitated with an

equal volume of isopropanol, cooled at 20C for at least one

hour and pelleted at 14,000 g for 20 minutes at 4C. After


removing the supernatant, the RNA was washed with cold 100%

ethanol, allowed to dry, and dissolved in 100 to 200 1l of

DEPC treated water. The RNA was precipitated once or twice

by the addition of 0.1 volume of DEPC-treated 3 M sodium

acetate, pH 5.2, and 2.2 volumes of cold 100% ethanol.

Depending upon its size, the pellet was dissolved in 100 1l

to 300 pl of DEPC treated water, and the concentration of

RNA estimated by absorbance at 260 nm (a reading of 1 equals

a RNA concentration of 40 pg/ml).

Depending upon the cell type in the experiment, ten to

twenty micrograms of total RNA was lyophilized and

resuspsended in 25 p1 of loading buffer containing 50% (v/v)

formamide, 6.5% (v/v) formaldehyde, 40 mM

morpholinopropanesulfonic acid (MOPS), pH 7.0, 10 mM sodium

acetate, pH 7.0, 1 mM EDTA, pH 8.0, and 3.0 Ag/ml of

ethidium bromide. The RNA was size-fractionated by

electrophoresis on a 1% (w/v) agarose, 6% (v/v)

formaldehyde, 40 mM MOPS buffer gel.

To confirm that equal amounts of RNA had in fact been

loaded into each lane, after electrophoresis, each gel was

photographed under UV light and the level of ribosomal RNA

assessed by ethidium bromide staining. If by ethidium

staining, ribosomal RNA was not even from lane to lane, the

gel was not processed further.

Equally loaded gels were washed twice in double

distilled water then soaked in 50 mM NaOH for 45 minutes.

Following neutralization in 1.0 M Tris-HC1, pH 7.5 for 45


minutes, and equilibration for 45 minutes in 0.1 M Tris-HCl,

pH 7.5, the RNA was electrotransferred to a charged nylon

membrane (Zetabind, CUNO), and covalently crosslinked to the

membrane with ultraviolet light (28).

Preparation of Radiolabelled cDNA Probes and Hybridization

The radiolabelled cDNA probes used in northern analyses

were prepared by random primer extension. Briefly, 500 ng

of the appropriate double-stranded cDNA fragment was

denatured by boiling for 5 minutes and then placed on ice.

From the Gibco/BRL Random Primers DNA Labeling System, I

added 15 Al of the random oligodeoxyribonucleotide

primers/buffer mixture containing 0.67 M HEPES, 0.17 M Tris-

HC1, 17 mM MgCl2, 33 mM 2-mercaptoethanol, 1.33 mg/ml BSA,

18 OD2e units/ml oligodeoxyribonucleotide primers

(hexamers), pH 6.8. I also added 2 Al of each of the

deoxynucleotide solutions containing: dCTP, dGTP, and dTTP

at 0.5 mM. After adding 100 ACi (10 Al) of [nP]-dATP (3000

Ci/mmol) and 6 units of DNA Polymerase Large Fragment, the

mixture was incubated at room temperature for 3 to 6 hours.

The resulting radiolabeled cDNAs were separated from

unincorporated nucleotides on a sephadex G-50 column, and

denatured by boiling for 5 minutes before being added to a

tube containing 10 ml of hybridization solution and the

nylon membrane(s) holding UV cross-linked RNA. The

hybridization solution bathed the membrane(s) at 600C for at

least 15 minutes before adding the denatured, radiolabeled


cDNA probe. Hybridizations were done under high stringency

conditions: 60C, and low salt concentration [0.76 M NaHP04,

ph 7.4, 7% (w/v) sodium dodecyl sulfate (SDS), 1 mM EDTA,

and 1% (w/v) BSA (crystalline grade, Sigma A-7511)]. The

stock NaHPO4 buffer was composed of 134 grams of Na2HP047H20

and 4 ml of 85% H3PO4 per liter. After hybridization,

membranes were washed 4 times (10 min per wash and one liter

per wash) under high stringency conditions with 1 mM EDTA,

40 mM NaHPO4, pH 7.4, and 1% (w/v) SDS at 650C. The

membranes with radiolabeled cDNA probe bound to

complementary RNA (northern) sequences were then exposed to

autoradiography film.

Quantitation of RNA by Densitometry

Where indicated, densitometry was performed on a Bio

Image Visage 60 video densitometry system. Any variation in

RNA loading was controlled for by comparison of the

autoradiographic signal intensity of a mRNA whose level did

not change as a result of experimental conditions. Thus,

each membrane was probed with at least two different

radiqLabeled cDNAs, one that bound to the mRNA under study

(iNOS, MnSOD or hGH), and an internal control that bound to

a mRNA whose level reflected loading differences and not

treatment effects (cathepsin B).



Expression of the rat MnSOD gene is regulated by pro-

inflammatory mediators including LPS, TNF-a, and IL-la

(Visner et al., 1989). Treatment of animals or cells in

tissue culture with these mediators results in increased

levels of MnSOD mRNA and protein (Visner et al., 1992). The

increased MnSOD levels subsequently protect the animals or

cells from oxygen radical-induced injury (Wispe et al.,


MnSOD mRNA increases 10- to 50-fold following 8 hours

stimulation with LPS, TNFa, or IL-la, and these increases

are blocked by co-treatment with actinomycin D, a drug that

inhibits transcription (Visner et al., 1989). In addition,

nuclear run-off studies, reflecting the number of mRNA

transcripts actively being synthesized at the time of

nuclear isolation, show a 3- to 10-fold increase in new

MnSOD transcripts following treatment with LPS, TNF-a, or

IL-la (Hsu, 1993). These results indicate that at least

part of the increase in MnSOD mRNA observed following

stimulation with these mediators is due to increased

transcription from the MnSOD gene.

Gene transcription involves, among other things,

interaction of nuclear proteins with specific cis-acting DNA

sequences (Dynan and Tijan, 1985). In eukaryotes, RNA

polymerase II and associated transcriptional activating

protein factors are required for the production of new

transcripts which are subsequently processed into mature

mRNA. RNA polymerase binds to genomic sequences located 5'

to a gene's coding sequence and synthesizes transcripts in a

5' to 3' direction along the gene template. Agents such as

LPS, TNF-a, and IL-la, that increase the production of new

transcripts from selected genes, do so by altering

interactions between nuclear proteins and either specific

cis-acting DNA sequences or DNA-bound proteins resulting in

increased mRNA production by RNA polymerase II (Ptashne,


One experimental approach to identifying these

regulatory DNA-protein or protein-protein stimulus-specific

interactions is DNAse I hypersensitivity analysis of

chromatin. The local alterations induced by proteins

binding to DNA result in increased local access of DNAse I

to that chromatin site (Gross and Garrard, 1988). This

increased access is measured as a stimulus-specific

increased sensitivity of the local chromatin to cleavage by

DNAse I treatment. These areas of increased chromatin

sensitivity are identified by autoradiography and are termed

DNAse hypersensitive sites.

Using L2 cells, Hsu (1993) identified a total of 7

DNAse I hypersensitive sites in the rat MnSOD gene located

from about 0.7 kb 5' to the transcription initiation site to

the end of the gene (Figure 3.1). The same 7 sites were

observed in both control cells and cells treated with LPS,

TNF-a or IL-la, but hypersensitive site 1 in the chromatin

from treated cells showed an altered intensity. High

resolution analysis of hypersensitive site 1 revealed 4

protein binding sites in control cells compared to 5 sites

in cells treated with LPS, TNF-a, or IL-la. No additional

hypersensitive sites were identified within 8 kb of the 5'

region nor within 17 kb of the 3' region of MnSOD.

Together, these data suggest that multiple areas of protein-

DNA interaction exist throughout the MnSOD gene under basal

and stimulated conditions and that treatment with

inflammatory mediators known to increase MnSOD

transcription, also creates a new DNase I hypersensitive

site in the 5' flanking sequence.

In order to determine whether hypersensitive site #1 in

the 5' flanking sequence of the MnSOD gene contained

promoter or enhancer activity, I cloned the 5' flanking

sequence from the MnSOD genomic clone into a promoterless

human growth hormone expression vector (Figure 3.2), and

compared the expression of hGH from control and treated

cells in transient transfection studies.

r- w
H w
9 C X .0 ,4 4
).c 0 0 41-

S44 > 0' H >

CC4J J04Bm B
oA *o) 0 O .. a
SV 4 J r. *
) 4t o H r* H
-,-A v 4., ro

,C! 0) 0 9 0 w C0
S*0 Ht-m H

W 9 CC ( UE- W (O
H. O) 0 4) r0 -H z
m -H O -. .-. 14 Q
4) ( a)4 U rP -4J 4)
0 0 .0 ( 10 to CO
Q U r- U, 4 0- 1 H ,-4

*4 )4 *-4 .4) H 0 V

(o > 41) r- 0':-
01 4 0 r) 4 ~
X,3 tU 0 0 a rC *
.a-HC e cC c-, c *
(dVa V E-4 9:
0 N4r4 ) 0 4 0 0 4

-HV 4)M tOH 0.
44- a) V A -) H *u x*

0 (U ) M14
* :- 4- 4 H 4) 0i

W 4) 0 -4 4 4) 0 *-4
o) c () r. U 0 V 4-)
t ) c0 *A 04 4 c)
1 r-4 4)aE-1 k 1-

0 Z aV) -X 3H V V.
M 0 4 > 04
W M 4 U *. 4) 014 )

*P 0 O0) > > *
0 ( VC r V V M 0V

C-HC 4 'H-4 0 -1
t444 Z CO4)HCQC
0 0 0 (U
0) V ) 10 4 M

V#P N 0 o1 k
,4 cxO A r. A t
I^U # UIBHJ ,C r.

I m



m m-


-mu ()

o) l-



E e
In Z
Cn Z

II 0
w cI

C\i X


U 0

0) i ON 0
4-4 04)

9: q> 0 .9 -r 4O
M -1- 4 O
4c 0

$4 w1 O
u 0 wi 0 0

0 0 00 > 0 4
4)4J40) 4C0 4)

00 W 0 S4
O) r. U4 WO
4 M 0 0) C0

0 0 ^ 0 +Jr. V-
Sv 4J 4) 0 W )

go > 0 a) 0)
*r-HHN 0 4 4-.H

0 0 0 04 M 04
O4) in ON u
-WWWSi *Md
0 00 0 Vi C -
0 aC E u) r 0 W
) 0 >00 00

4HN 4 O4)H
O V -H00 0O 0

O (0 a0 04) 0

.0 0 FA 4)
004 WO 00
to w H 0 O
M *r4 O V

) 04) H 0 U 4 0
C k r0 0 0 04

0 *-4 VM 0 .44) r-
ObOHO 004 OH

* 0- H 0

S4 t- r-4 0 > In 0























The 5' flanking sequence of the MnSOD genomic-clone

from the EcoR I to the Eag I restriction sites (4.5 kb) was

cloned into the Hinc II site of the promoterless hGH

expression vector, pOGH (Nichols), creating the vector pEco-

Eag GH (Figure 3.2). Using DEAE dextran, 5 ig of this

vector was transfected into 64 plates of L2 cells in 4

independent experiments (Figure 2.1). The concentration of

hGH in the 4 ml of medium covering each plate was measured

in samples obtained 72 hours later. Figure 3.3 illustrates

the basal expression of hGH from the sixteen untreated

plates transfected with this vector was 0.83 0.08 ng/ml

(mean S.E.). Continuous treatment with LPS or TNF-a, from

24 through 72 hours after transfection, significantly

increased this expression to 1.87 0.18 ng/ml and 1.65

0.14 ng/ml respectively (p = 0.0001), indicating that the

MnSOD Eco-Eag fragment contains cis-acting sequence elements

sufficient for LPS and TNF-a responsiveness. On the other

hand, expression of hGH from the 16 plates transfected with

pEco-Eag GH and similarly treated with IL-la only reached

1.06 0.09 ng/ml, and was not significantly different from

untreated controls (p = 0.08).

I next evaluated whether the observed inductions in hGH

expression were specific to the MnSOD 5' flanking sequence,

and not due to non-specific effects of LPS, TNF-a, or IL-la

on transfected L2 cells secreting hGH. I compared the

effects of these treatments on cells transfected with pEco-

O 4 U >A
-4 0 04 0
.rq 4.) 0 r-4 .0 4 4- 9 4
la 0 (a -Iq 441 0 U 0 U r-4 0 41
V V $ MU 0 > 10 00 1 -4 I0V
C 4.) 0 -I -4 q > O U O0
0 k C 0 ) ia .3 to (d -q-4
*4)00 4 U E-4AA> V
V00 *4 0 QU 9

U k O )CV1C0 W IV MVP--
) 04 rC (a V r. -4 C 0 -4

> $4 -r4 FA M.-4 r-4ooeh 044 4 4WP
0004 r0 W 0W 0 40 .4

014) V 0 43-I IW 4.4 4) 0 C P4 V
0 3 tP 0 p4 U u 0 W 0

N VW 0 W '09 in X) tv -r4 4
44 -40 r-4-4 4) 0 r. 4) 4 ) > U 4
S000*r- >'0.r'0 M 0 .--I t4O

c m'4'O W-HVV9 0 V 0 -4 44 4)
00 *c u a U c0 to-0 t7 >

-HI 4)J u 0) IA4J V w 0 c go
W U 0 N 0 0 44 -1 0 I) N O' to IV
-'.04) U 404)IAW*4%

IA r-44k M A C e4- M- >i-i A
1) 4) 590 4)) r t V 4) 0 04 04 W U
N*o4 4 t-4I Ov4 0 fa 40
090 0) 0 4) 0 4M ) 4 -I-.-4V
X 9 M 4) V d 4J V V 0 Id 4) C 3
r-4 O ) r-4 N ) <-1 WU >i) U 4
0 0in n: 4J ) On 4) -i to -H 0
* CQ *IVO H A -)O C r4 ) -IV
.0 5 O O 4 '0 C a U
0HC 0 0 IA k OIV 0 4 0Ure-. r0 0
0-4 01 4-0Z -.) 4 0 0 0 P >

0> X W 4) 4)4-" M : M O rV
-'44 4 10 w--IV4)to *m
04)d5.0Cl) W O10t V9V 0-1-1.M:
OUCM < V .M 0 )4-I) V 4)
0-H 4-40 ZC N tto E 0)
W0 >k4- IE 0 V0 0 v
0 >+j t 0 V N 4 U4PWO O
*V-I 0 M040 *M V 0 .4 0 043
*' 00 U 9 4)V0 134M 0U 3 0 44 W O
0 0H & 0 4 ) 0 W 0 -4 V
C > 44) I< 0 0 m 00 >4 4
>i0 V 4)r-4 W) 5q o t 0 0V
r-4 C W -I0V V V V M 0) V :.0

*)4 W -4 r- 0( 4) >4 W4 M # 00
r4 4 .0 -'UI 0 4) 0 U 4 0
004)4 4r.M04 4J 0 r )90 V 44 wr
P4 > 0 WO 4) UE- > -0V4)V-

mH 1u

%,%% M

(iW/Bu) HOLI

Eag GH to cells transfected with pTK GH, a vector that

contains a minimal thymidine kinase promoter for low level,

constitutive expression. None of the treatments had any

significant effect on hGH expression from pTK GH, while LPS

and TNF-a significantly increased expression from pEco-Eag


In order to assure that the hGH in the medium reflected

the majority of the total hGH produced by the transfected

cells, we measured hGH in the cellular monolayer. These

data showed that only 2-3% of total hGH expressed was

retained in the cells and >97% was secreted into the medium.

To more closely define those regions) within the Eco-

Eag fragment of the MnSOD gene that are necessary for basal

and stimulated hGH expression, I used unique restriction

endonuclease sites to successively delete portions of the

MnSOD sequence and tested each resulting shortened pGH

construct in transient transfection studies. The 5 vectors

are illustrated in Figure 3.4, and the results of the

transfection experiments are summarized in Figure 3.5.

Basal expression between vectors was not significantly

different until the MnSOD promoter fragment was shortened

from 188 bp (Sac II-Eag) to 79 bp by digestion with Nae I.

Mean basal expression of hGH from the 18 untreated plates

transfected with pNae-Eag GH was only 0.05 0.01 ng/ml -

below the level of detection by this assay. The Nae I

restriction site at position -120 (2454xmL) is located only

46 bp 5' to the MnSOD transcriptional start site at position

-74. These results indicate that at least some portion of

the MnSOD 5' flanking sequence between positions -157 and -

48 is required for basal expression of the rat MnSOD gene.

Similarly, LPS and TNF-a stimulated hGH expression was

not different between vectors until the responsiveness to

these mediators of inflammation was eliminated by shortening

the MnSOD sequence using Nae I digestion. Treatment of

transfected cells with either LPS or TNF-a significantly

increased expression of hGH between 2 and 3 fold until the

MnSOD 5' sequence was shortened by Nae I digestion.

Expression of hGH from the 54 plates of cells transfected

with pNae-Eag GH and treated with LPS, TNF-a or IL-a

remained near undetectable levels at 0.07 0.01 ng/ml (mean

S.E.) and was not significantly different from untreated

plates (p > 0.1). Therefore, LPS and TNF-a stimulated

expression requires at least some portion of the 85 bp of

MnSOD sequence between the Sac II restriction site and the

Eag I site. This sequence contains the DNase I

hypersensitive subsites #6 and 7 located within the strong

hypersensitive site #1 identified by Dr. Hsu (1993).

IL-la treatment of cells transfected with vectors pEco-

Eag GH, pHind-Eag GH, pSfi-Eag GH, pSac-Eag GH or pNae-Eag

GH caused no significant increase in hGH expression.

Unexpectedly, IL-la treatment of cells transfected with

pSal-Eag GH (n = 21 plates, 9 experiments), significantly

increased hGH expression from 1.8 .17 ng/ml to 2.8 0.39

J ) 0 A4
1 4 1 *

S 'O CG 0 M C(

0 P 4*P* *J -,0 -
) U tP M W N tn
(0 4- 0 r 0 *4 -*
0O ,4) V E4 0 Q0
H 0 0 04...0

0 9 0 V 0 ,MI
-,- 0) C V *-H
4) N O W (04-4 >1(a

go oM* U) V) .0) U *-
r.4 J 4J4 V C V 4)
C -I (0 *H -* 4V A
0 1 M43 0 10 0

0) r-4J V ) 1) t0o F
kr *4 0 0A > C V
La X C 0 *1 -* -1 *-
O- E- H 4
*i- $ rH CH
M V *r -r4 E-4

H 1 ) C to
4-4 44 >i o ) ) *
- 0 o4 *4- -r4 Ud
iu) -H4 C k um U E
0 4V to 0 U H
i -4 U X: C OU o
M O -i 0 4) 134

4) 4) 4) ( ) 0
V M P r-4 CU
( 4) 4) c ol) r. 4)0

4 0 0 *40 n 4
*Ur. U >4A8
M -f o) W V C
V tP U x *-r W H a)
) t CrO 0) .0
14 910 ).QM9 .4 4)
0 W 0 EU t EU *
tP A Cn tl) 4) O 9 C:
-*H4 U c ) .r-iH C *
FM74 MW4.P cMr-H:


U) C%~41

() 0) 0)
0 0 u

C u
co ot
2 \ c o -

-0 z cy \C^u-^ & 1-

E ce
0) ) 'a


mi CI

i~o Ic


E =e -

o C- o c -
a) I'






WI i-

0. V 4
C 0 ) 0) W rC 0 CO
HH rl A %D \ -H 43 D to r. 4
Hr4 V V.P 4J V 0P U 4) U
IV -1 0) 4 (d (M 44 PC 4 -rqi
OH IV A 000 0 4) t Q 44 0 0
r-4 Vp E-f 14 4) C *-r4 4 14 -r4 -r 44i

V U '0 0 4) V Or- ir

3%4 OM44r)WPA00 C e 0
zt ON 4.4 to 9 A 0 w 40 C B
*I a0 rMC4o B <) 3a 0 4 -

a-Ine .a'>.o 9 Ln 0 0V >Vr4 W-
a4CO 4 -i c 3n V r. V 04 V rr i -4 n AO

tp o r A-I
k 0 4 d) b $ 4) U 0 $4I 0r.N4)rV
to o g kOi ) m to 0q *0 -4 r AU
ka 45 4) to M=p~ 1 V'- 01 ( >P IVUBA
4) V 0 4) H fd ) 044 V'-^ O
0 0 W V > (a) u 9: 4 u 09-0n1 $-04 '5
$4 44O 0VW0M1 w00 )1V 4J4
444 0 P VH 9' P wP a 0) 1=a 0V 4B rl.
V- 441 4) 04 10 :3 Ln M *iq 91 0 4) n $4 3 ON
10 *Hi V 0 O.4 N a) go )
Q) 3r U' V 0 0 x tS X: (u) 41J >j ; 39 nM FA
tO Moir a 0 1 FA o AOM 1 a)
tO 9 ) k. 4 k W -4 4 r-I X 4 (0 V 1* 0 00)
(D 0 ;J 0 WA4^ZNM U u-i 0) $4 93 P V 10
xu to p ) 0 > > 4)--l
44J 00 4V 0- 0 V r-0 0 0 0

> nr-, 0 4) o U k a4r4 4)
k r-I :3 V N+ V 1 tO 4) 4 U fd <1) 0
.c Id 9 ) -i u (0 r. a) 4i 4) -ri > 39+ gV
POT C l. 4 0 H 4Cr 4) V0

04 V 0 0 0 9 A 0-H 0 C4O 0 W 0443 4) V
T0 r-40> 0 )0 $r- W 10 4 V 400
k 44 0 -I4 *4 r-4 WX: 04 *4 MV*V4 r.03

r4)- *rH-, U 0 H44 V r.4J V A 0VV
) 4-4 n 4 '0 ( -) V Ce O 10
> to 0 0 r44-4VE-4.P 4- *V' 0. -
4) Q) 0 W) > U ) 0 o 4) t ) 0

0.0 W H r4 -4 $4 ) r0 4) $4 4) 4-) 4) CA 04
4 VOO( CO'0i 4C C 44 r. 4J) k 4) P --f
*0 044-1 4 0 0 V 4 4 > U0) 0 ( 0 44 C 0 t) )
M 00%44 r > > V o 0 0 *0
m Nk0 0 -V 9 9H rlr 4g
TA..VXC 00-04J0 m-4 ---r-llCM
r-m4J00' 04. *-40WT44000

-> 94-) 4) 4) --I U 4J *H ) tQ M V M U M
4) 0 u co 0 V W r- t MA AC
m to n-l0 4) '-l '-I 0 3 0-i 01i-1. i ) u 0 ) 0 ) -H 0 d 4) 0-4
3<-4J B00 0r-4-H Mv-4-4 M 0 M'0 t
M4 04 r M r-i* 0 r. M4 P &to
*04 0 0 0O 0 0 0 "-r 000 044 XJ V 0C) 0W
rZ4 0 U 0 9 0 4 -4 4 ki 4 4V 0i 0 A IV r4 O .0



xM x

Lo 't C\ r-

(Iw,6u) HL












% % % % %


ng/ml (p = 0.01) compared to untreated controls. Since no

significant IL-la responses were observed in any of the

other 5 constructs, it seems likely that this apparent

increase is due primarily to a large increase in hGH

expression measured in only one of the nine independent

pSal-Eag GH experiments. Based on these data, I conclude

that the MnSOD 5' flanking sequence is not sufficient to

provide the IL-la induction observed when the endogenous

MnSOD gene is stimulated with IL-la.


Transcription of the MnSOD gene is induced by the

inflammatory mediators, LPS, TNF-a, and IL-la. Chromatin

studies have demonstrated a strong DNase I hypersensitive

site in the 5' flanking sequence of MnSOD that changes in

intensity in cells treated with LPS, TNF-a, or IL-la. Using

transient transfection studies, I have tested a 4.5 kb

fragment of the 5' flanking sequence of the MnSOD gene for

promoter activity. Results indicated that not only is this

fragment capable of promoting basal expression, but it also

permits LPS and TNF-a inducible expression. It did not,

however, effect a significant increase in reporter gene

expression in response to IL-la treatment. Furthermore,

the levels of induction in response to LPS and TNF-a

treatment were modest, only 2 to 3 fold, compared to

inductions of endogenous steady state MnSOD mRNA levels of

10 fold or greater.

The MnSOD promoter is G-C rich and does not contain

TATA or CAAT sequences (Figure 3.6). The sequences that are

important in transcriptional regulation in GC rich promoters

are not as well understood as those containing a TATA motif.

The MnSOD 5' promoter contains one AP-1 protein-binding

consensus sequence (GTGAGTCA) beginning at -502 (20710L)

and two hexanucleotide Spl protein-binding consensus

sequences (GGGCGG) at -305 (2268maL) and -81 (249200L).

These sequences are important in regulating basal

transcription of many genes including the mouse

dihydrofolate reductase promoter (Dynan et al. 1986).

Deletion analysis of the 4.5 kb 5' flanking sequence

indicated that the level of basal promoter activity did not

change significantly until the MnSOD sequence was shortened

from position -228 (2346FL) to -120 (2454EL). At that

point, reporter gene expression was completely eliminated.

However, although these data demonstrate that the Sac II-Nae

I fragment is essential for basal transcription, no

consensus sequences are located within that portion of MnSOD

5' flanking sequence. In addition, induction of genes by

LPS or TNF-a is often (MUller et al., 1993; Lowenthal et

al., 1989), but not always (Zhang and Rom, 1993; Patestos,

et al., 1993) mediated through the binding of nuclear

transcription factor NF-xB to its consensus binding sequence

(GGAAAAGTCCCC). Two NF-KB decanucleotide consensus binding

sequences exist in the MnSOD 5' promoter: at -1515 (1058EL)

and -426 (2147EBL). However, LPS and TNF-a responsive

Figure 3.6. MnSOD genomic sequence from positions -545 to
+38. Selected restriction enzyme sites are illustrated
along with potential binding sites for known transcription
factors. The adenine nucleotide of the first ATG signal is
+1. The start of transcription occurs at position -74.







I- -









Isfi I





INae I

IEag I


expression persists in the MnSOD promoter even when it is

shortened to -228 by Sac II digestion. To eliminate LPS and

TNF-a responsiveness, it was necessary to shorten the MnSOD

5' flanking sequence to position -120. Neither of the NF-KB

consensus sequences are located in the Sac II Eag I


The 2 to 3 fold induction in hGH expression observed

following LPS and TNF-a treatment is similar to the

increases in nascent MnSOD transcripts measured in nuclear

run off studies employing isolated nuclei (Hsu, 1993).

However, it is much lower than the induction of endogenous

steady state MnSOD mRNA levels following LPS or TNF-a

treatment (Visner et al., 1989). This fact, together with

the absence of IL-la inducibility suggested that other

regions might exist within the MnSOD gene with important

regulatory function. This hypothesis was tested in the

experiments described in Chapter 4.



Chromatin studies on the MnSOD gene have revealed six

DNase I hypersensitive sites within the gene in addition to

the hypersensitive site in the 5' flanking sequence (Hsu,

1993). Each hypersensitive site indicates an area of

chromatin with increased DNase accessibility relative to

surrounding chromatin, and these areas are potential sites

of protein-DNA interactions with important regulatory

function. Although I showed the 5' flanking sequence of

MnSOD contains promoter activity sufficient to drive basal

expression and to permit induction by LPS and TNF-a, the

magnitude of the induction was relatively low compared to

their induction of endogenous MnSOD mRNA levels (Visner et

al., 1989). In addition, unlike the endogenous MnSOD gene,

the 5- flanking sequence was not sufficient for significant

induction in expression following IL-la treatment.

In other genes, such as the human and mouse

immunoglobulin kappa gene (Queen and Baltimore, 1983; Judde

and Max, 1992), platelet-derived growth factor gene

(Takimoto and Kuramoto, 1993), alcohol dehydrogenase-l-S

(Callis et al., 1987) or collagen (Simkevich et al., 1992)

genes, regulatory sequences with promoter or enhancer

activity have not been limited to the 5' flanking sequence,

but have also been identified within introns or in 3'

flanking sequence distant from the gene in question. We

hypothesized that MnSOD sequence contained within the other

DNase I hypersensitive sites might contain promoter or

enhancer activity. Therefore, we created constructs

containing hypersensitive sites 1 and 3 through 7 to test

this hypothesis.

Typically, promoters are short, cis-acting sequences

located 5' and close to the start of transcription of the

gene they help regulate through sequence-specific

interactions with DNA binding proteins (Dynan and Tjian,

1985). Enhancers, on the other hand, although also short

and cis-acting sequences, typically function whether located

5' or 3', or near or far away from the gene whose expression

they help regulate (Ptashne, 1986; Serfling eet al., 1985;

Tjian and Maniatis, 1994). Promoters are only active in

their native 5' to 3' orientation, but many enhancers are

equafly active when cloned in either orientation. In order

for enhancer elements to modify transcription of a gene, it

may be necessary for the DNA containing the enhancer to be

distorted or bent such that the enhancer element can

physically interact with promoter sequences in the 5'

flanking sequence (Lilley, 1992; Wolffe, 1994; Giese and

Grosschedl, 1993; Cullen et al., 1993; Su et al., 1990).

There is also evidence that the function of some enhancer

sequences is dependent upon their being on the same face of

the supercoiled DNA helix as other regulatory cis sequences

(Hochschild and Ptashne, 1986; Dunn et al., 1984). If the

position of these enhancers is moved by 5, 11, 15 or 24 base

pairs, thereby placing them on the opposite side of the

helix, their function is destroyed.

In order to test whether the addition of the MnSOD

genomic sequence containing DNase I hypersensitive sites 3

through 7 could modify expression compared to the 5'

flanking sequence alone, we created several new constructs

(Figure 4.1). The 6.1 kb Hind III fragment from position

+1179 to +7311 contains introns 2, 3 and 4, in addition to,

exons 3, 4 and 5. This fragment also contains all but one

of the 6 DNase I hypersensitive sites in the MnSOD gene

other than those in the 5' flanking sequence. The Hind III

fragment was cloned in both orientations into the Hind III

site of the pHind/EagGH expression vector (Figure 4.1). The

resulting vector with the Hind III fragment in the 5' to 3'

orientation is denoted pJM 17GH. The vector containing the

Hind III fragment in the 3' to 5' orientation has been

denoted pJM 13GH. The ability of each vector to drive basal

GH expression and LPS, TNF-a or IL-la stimulated expression

was then assessed in transient transfection studies. The

observed enhancer activity was then further localized.

Figure 4.1. Expression vectors assessed for MnSOD gene
enhancer activity. Schematic representations and
restriction maps of the MnSOD genomic clone and the
expression vectors constructed to assess potential enhancer
activity are shown. The restriction enzyme sites are
indicated above each sequence, Hind III (H3). The
expression vectors contain the 6.1 kb Hind III fragment from
positions +1197 to +7238 of the MnSOD gene, with DNase I
hypersensitive sites 3 through 7 (*), in either the 5' to
3'orientation (pJM17 GH) or the 3' to 5' orientation (pJM 13
GH). In addition, each expression vector contains the MnSOD
5' cytokine-responsive promoter sequence from the Hind III
site to the Eag I site.

Cloning of Vectors Assessed for
MnSOD Enhancer Activity

Sac NaeE
Sfi Eag



Hpal +7238

* ** *

H3 Eag

pHind-Eag GH


pJM17 GH



07 0 a 7238 Eag Eco H3
I. Il i. I.--------

* ***



* hGH

1107 Eag

pJM 13 GH

*** *










Within the same experiment, when compared to GH

expression from the Hind/Eag GH vector, neither pJM 17GH nor

pJM 13GH significantly altered basal, unstimulated

expression (Figure 4.2). However, pJM 17GH dramatically

increased LPS, TNF-a and IL-la stimulated expression

compared to Hind/Eag GH, indicating an enhancer activity

existed within the 6.1 kb Hind III fragment of the MnSOD

genomic sequence. In contrast, both basal and stimulated

expression from pJM 13GH was not significantly different

from Hind/Eag GH. Thus, the enhancer function of the Hind

III MnSOD fragment is dependent upon its orientation.

To confirm that the changes in hGH gene expression

observed at the secreted protein level reflected changes in

transcription, and not changes in translational control, we

compared levels of hGH mRNA isolated from cells 72 hours

after transfection (Figure 4.3). As expected, non-

transfected L2 cells expressed no detectable hGH mRNA. L2

cells transfected with pHind-Eag GH or pJM17 GH expressed

basal and LPS-stimulated hGH mRNA. LPS increased hGH mRNA

expression 3 fold from pHind-Eag GH transfected cells

compared to a 9 fold induction from pJM17 GH transfected

cells. These data support the enhancer function identified

in the protein data with pJM 17 GH. Expression of

endogenous MnSOD mRNA was detectable in non-transfected and

untreated cells. LPS treatment increased MnSOD mRNA levels

by 11 and 16 fold in cells transfected with pHind-Eag GH and

o %0
mW 0
Ot I100
>0 O 0 0 '4 0
M> O 1 n1 )O 0 41 0
4 4)4P 0 .) C 1 0) 0
*4)400 A0 M04) 0
I atv 0 W 0 4n 0 M C
> & 4) r I 4* 4) 4) r- ) >10

. 04 C C0 4- A 0 0 C U
M 04-) ri 0 4) 9 U 4 -4 S<0 0
0 0 w 0 k V 4)0044
O 0>U 0 0 40 4 1 -4 $4
0) 0- -) -4 ) M 3 10 0) ,.-.O0
U >.UX040< (0c
SE-0 I 0 4) M 0 140 -4)
-4 -V V 0. V M 04 > -.I

-0 0o 0 0 ~ 4 04 )
014 444 0 >*i
S.-.. A C4 *),- -,-A*
S0 4 d .-4 > 0 U-)V 4)-
So 0 0-444. >4 > 0 0
o04) 1 t o 14 4) r-4 *0
t9 0 H4 8 a 4) > 00 4) 4) () 0 k
m0 o-p 0 A! w m B o -4 4)
0 M 0E *4 r. 0 E ) v V r-H 0 0 0) 0 IV >04)0Vr. m 9
U 4 10d *ri A 144 ) A (i) 0 4) (d
O i 0U Oa C E-4 M M U 00 M
>0 10 HOr- 4). 0- 14C
U aWr40 r4 V.CO H
r.C Z 04) 4 N 0 0(I
* -4 E-4 H tQ > HU >4-1 m
*- fo r O r-i E- tom00v-

mE t (1 0 (HE 4) 4 Ut0 IIA
c 0e-4 W > 4 4- 0 -H ) 0 -)
9C-f- ONCfa 4v-r U go MH VId
Or C44 M.40 H M tQ4) EU 4r-I0to

4 0 H 0 0 E0 4-) 4r-4 0 0 U n 0 H
o u t to -H 4 1)F4-r0 4 0i
O 0 r 4) I ( 04- 4) U i *.-I
124U U *-14' MX .-HBdH+U 0
X 4-) t r (1) V 4- U 00 .4
0 0 U r- 0 H 3 C r-4 (0
M V0 ) 3 E0P --E)4- -4

40 C H 9 944 9 000 V
> 4 d C- M O 0U -H M W-'0
0 t- M04) Hd 4) > 4) U -IVE
0 P 04r-40 (d CP V 4) X l) <4) 4) 4 tl

M: 0 O 0 -4 *-. k Ht m 8 '-H
44 0 ) 0C 0 N 4VP1 14 0 r M
0 4 044 r4. 0) 04-H) Cn

0 M V0 0 W 0 *O Qk
04 go -- M r 4) 0 zo tc
*O- 00 44 m 9r c ( 00 m 0 4
Pk 0 eH 0*- :t Q) AV t = )M

WOO6 *ai'*l-M'aVZnW## a


X! !1



cD C C'J

% % % %H


4) (a
> -I
*H 1 ) dO0)
k 4O a)+J f-4 c 4) >
4~1 0 0 4J M 0 .-r4 V0
*p o u *P B o-'i N
>AU C g 0 C. m 00)
410) 14 -) HO -4 W-1 to 44 <
4)O U 14 0-1 M 0 4)
*o >#<)t7'iW*iM nC D'- ( ( S *
V@ 9 .Q: OU 9 4H M@ 0
U i 0 M 0 00410 $4 -Hj
4-> C BC Q3 M(# C ^QAB
4)H 9 0 o A 29 rZ 4) vt 9 V *> W
14*-HrNCC 043C-0 OM
4W4- -9r.0 U 0 m
(l) t O IV -4 () g <'V CV4-) -r 4) W CO 0
MCri NI mmovaCw
4Mc4 E 2 ) u a3 H a
Law09 m Ctor 0
40 U 044 J 0 V '1-4041 4 U44-H 41

to0 4 ONL00 4I
$4 (1 0 ) V4 440< (1)p V O 90 -4.f~ V i
3^oS oxoU mm o a c
w01.4r.L 0. 0-J9000
0 0) 4- 0VM E- M In 4) IV (
) k-i ) m n v 00
N to > 0 M N ie 9 V- Mo V- tos
HD gr-0 H 0 r. 0C
H-i -4) + r.r-I M 4 k h'o -<- ) Im
to ) 1 3 IV 0 -44 V40
H-I u O V0 -1 .Ca = >4'0 r.
4) ON (4)> n M OU H 4) S-1
H> '0 I-. 3 a 0 4 0 u
0 4CMO 40 C-.-0
0cV2 k oa)t41a 41 0 ot

0C V r+-) Id 01 kt.L C 0-H-
(a wM $4 V :3
Zr H-H)0 0+Ho -1'-4H.)- V U
r-4H00Lk OV-H 0 CV4)
il CMC *-400V 0C
S0W0 04 0CV4-i N #) 0-H
.) X: f r-I *-rl 4) (a *-1'0-I Q0 H Oi 0
. *HOV -- M *HH O) 0.1 'o
.&441-- 1. -La004-JG
*0 In -4 C 40 0M -H 4'--
(W- 4 O) O 0 M > C 4 o
0 t)UQ 0 Q)0-H40 V ) 4)0 Cr.V4-4
r-I 0Cr0 4k UiC4. Va
C ON 4) CO V r. >4 V n 00
OCH C r. 0) EDr Wa 3 4
10 cyg w 0 3- O 4 q0 p-i
*-ri -i toi ) p 04 A to tn rc *
k H- 0L0 9UM.H 00 >40 4J 0
" 90 ON V V 0 M0 AC *
MMM a c ** 4J^'c'oioMj
0 O *H ,C <- l l 0 <<) C (Q O
0 04 wD 9 9 --143kar r
*U r. 0 & U ) V gu)0 IV
O) go A C41 4U) R-HC > -I 04
*0>-H 1w01-4H 1.OHX
0 >H -4 k 044 0-I 4 H 04 00
V 0 9 V41) U r >U
) i Hl ? 0 w 0 0-H 0H I
.)-o11 -HcH 44 0 ro H>
1.4Lac-( 0L~ria0L-Ua 0oto
40) IM C 4) M 0fl 0 U W0
09 4 0^ V M94-H rl +400 r-l 4
0 in 9) Q (0 0n VZ o vmH e n
*H4 1-H 0-H H 1.4 N IV Q ) r 0 0) %
&V 4 0 U M -4 r" U O -H U r-H

\ | I Iu I


I m
0 0 c

2 0
C 0.

pJM17 GH respectively, indicating that expression from the

endogenous intact regulatory gene sequences is still more

efficient than from either transfected plasmids.

To begin to localize the enhancer-active sequence

within the Hind III fragment, I linearized pJM17 GH using

the Hpa I restriction site at position +4942 (7515mmL)

followed by a partial digest with Hind III. I then isolated

both shortened linearized constructs, and filled in the 5'

Hind III overhang using a Klenow fill-in reaction (as

previously described) before lighting the shortened

plasmids. The resulting two deleted hGH expression vectors,

pSC 17/11 GH and pSC 17/9 GH (Figure 4.4), were assessed for

enhancer activity compared to pHind/Eag GH and pJM 17 GH.

The vector containing the 3.8 kb Hind III to Hpa I fragment

from +1106 (3679EmL) to +4942 (7515EmL), denoted pSC 17/11

GH, includes DNase I hypersensitive sites 3 and 4. The

vector containing the 2.3 kb fragment from Hpa I site to the

downstream Hind III site at +7238 (9811mll) includes DNase I

hypersensitive sites 5, 6 and 7, and is denoted pSC 17/9 GH.

Results of transient transfection experiments in the

rat lung L2 cells are summarized in Figure 4.5. The data

again demonstrate the enhancer activity of pJM 17GH

significantly increased stimulated, but not basal, mean hGH

expression compared to pHind/Eag GH. The enhancer increased

LPS-stimulated expression from 2.26 0.14 ng/ml in

pHind/Eag GH transfected cells to 8.7 2.46 ng/ml in JM 17

GH transfected cells (p<0.05), a 4 fold increase. For TNF-a

treated cells, the enhancer in JM 17 GH increased hGH

expression pHind/Eag driven expression from 2.33 0.28 to

7.49 1.37 ng/ml (p<0.05). In contrast to the 5' flanking

sequence in pHind/Eag alone, with the addition of the

enhancer-containing sequence in JM17 GH, IL-la treatment

significantly increased hGH expression. IL-la treatment of

pHind/Eag GH transfected cells only increased hGH expression

from 1.0 0.47 to 1.54 0.25 (p>0.05). With the enhancer

in pJM 17 GH present, IL-la stimulated expression increased

from 1.83 0.35 to 5.2 0.98 (p<0.05).

The enhancer activity of JM17 GH was retained in the

pSC 17/11 GH vector containing the upstream, 3.8 kb, Hind

III-Hpa I fragment with DNase I hypersensitive sites 3 and 4

(Figure 4.5). In contrast, when this 3.8 kb fragment was

deleted from pJM 17 GH, creating pSC 17/9 GH (containing HS

sites 5, 6, and 7), expression was indistinguishable from

pHind/Eag GH.


"These experiments demonstrate the presence of a

functional enhancer in the Hind III fragment of the MnSOD

genomic sequence between +1179 and +7311 from the start of

transcription. The enhancer more than doubles expression

compared to the 5' flanking sequence alone. It has been

further localized to the upstream portion of this sequence

Figure 4.4. Cloning of deletion constucts from pJM17 GH to
localize enhancer activity. The Hpa I-cut pJM17 GH vector
is represented along with a restriction map, DNase I
hypersensitive sites (*), and exons. The Hind III (H3) to
Eag I 5' cytokine-responsive promoter region of MnSOD is
denoted by the horizontal white rectangle near the 5' end.
The H3 fragment, from positions +1107 to +2738 of the MnSOD
gene,is denoted by a gray shaded rectangle. It has been cut
at position +4938 by Hpa I digestion. Following
linearization of pJM17 GH with Hpa I, partial digestion with
Hind III created the 11 kb and 9 kb deletion constructs
called pSC 17/11 GH and pSC17/9 GH. The H3 (+1107) to Hpa I
(+4938) fragment contains hypersensitive sites (*) 3 and 4
and forms part of pSC17/11 GH. Following ligation, the
orientation of this fragment has been reversed relative to
the parent vector, pJM17 GH. The Hpa I (+4938) to H3
(+7238) fragment contains hypersensitive sites 5, 6 and 7,
and is found in pSC17/9 GH. Following ligation, this
fragment is in the same orientation relative to the parent
vector, pJM 17 GH.


Cloning Deletions to Localize
MnSOD Enhancer

Hpa I



Hpa I cut
pJM17 GH

in *




pUC **

pSC 17/11 GH

7238 Eag


* hGH

Hpa I

pSC 17/9 GH


** *





Hpa I







IIll llliii ;ii llllli


- fl V0 '
in (n jWcV '
o 4) rHi 41E io0
O a -) >
0 .C 00
a> ) U o >i :9-4 0 0
0 -r4 w a t 4 tP A 04.0 V 44 )
r.V0*V4 i 4ON= r. r. 44 90
cpU r.4 0 0 0CC-4 ra.
ou cq wo 0 -r o to 9o:-i
^ e w-43 i (4 44 J 4 $4 $4M-4
Q0 vM-10 0 0 -to0O
t-l fl) 4W & X U.crnO M 0 if0U O *
N C S4J w-4 4 r- 0 kME -HO4 Ui4 r-4
Q U 4- 4 V V P HI
toc 4 Mn H > r-4i
I am 4) ) 0 B >,#i) 4) T
en >' P M* r -i >^u

W.0 00041 A #. O*M g M41
Q0E40049-4 d 0- 0t
a Vo crE ad -O. ra Ea 41 n

41 0 W v-4 r4 t >H W v-W WQ4k
to 9 r. o -i j t0 n m a) X V ri4

4 ) 41 V r-I 0v- VV00 H 0

00 c4 c-WV 0 Ni>cV0M U

co M 0a H9*- o CO 44 044 4 U -t-ED 0 4)
V 4)OC *H *OV4 W 4) W V-
(>0) a) ) -4 = v0 4) r. -H ) a o' u E-)r

4-) V tmr 04 9W k10i d V 4 4 X
4i ) O N >". c0 < -'r. 0V
4- Q) 9 k V > CIN r-4 44 44 r-0 EQ <0 4) VV

U A j4)$ 40 ; 04 0 r-4 CO i41M:
9: c s: l 0 k *r3 0 M v 0 in HV 0 M
40 0 a 0a0 to*u N4 W 0
40M W 040 :1W tQV-4P9 U -
Oj -ptp *MD aucm^ icMC 4 i 10

o 4 A4A 0 *) M aO W0# 0 W
0000 OC rA -44 *H 4)1-41

S4 OD 0 0 V 4-4VMdV4V V-

*0 W V M 0 t 0 00 9 Ora w U
BCO Ur M 401 ( 4 0 i ) 4

Z 0 00 4) Vr O* r.0
u H M Q i+ -aao *-i o<-i -i u

v 0 a ) 4-) 0 CW 0 r4 0

U in r. 0 0$4 O il0 4r. r 01 r.
0' 4 4) 1 4) 0'0 U 0 O *4OC 4 0Z
V V 1 00 4 4 to arO* *
0o > J .34 0V 4 0 rc A V 0
S -4 W --H 40 0E 4)W Ir M *HOr
r- o ) o FA .9 (a 4 : > w c to a 1)

0 W n 00O4 -4 PV -H m 9 U '- W 4)4
.r40> --4 ) -r4 o0 *4 ) 1)* 0-4 0 C 4

VA 9 0 W 4) 0 0 $4 0 -44V a

10 0 W 5*V1 SI W00 FA -" 0
N V 0 r. M 44 to c 4 -Oi *) 0 u

0 *r-I H)- r I k B
t eo d3 *r k m ai 44

u 0 4 ZL 0 -i 4 0 >x W $
0z1 B--1 V 0 ) H. V to 4)' V 043 4
0J V U k LO u 4 O>9 r
*BQ C 4 3 9 lO a # 4f- 1-n

rt 4J > c 4C 044a i 4 B M CP ON 0% 0
* 9-) v to r4 0 a Vc OIV
#4 H -4 0 9 )Vr. M -H r
. 0 I P 7 1 0 0go > r- (1) 01 f-0
Vs i- H -4 4t 1-4>
9 0 FA > $4 Z V ) U tQ V t r COA
C) FA 4) 4 E i-4 14 .0 WH -rj 4 F
gn -U )r-4 --I k 4) o0 M o M c M 4)

M 4 *r- 0 U 0 H I m 4 k 0
rq & k > 4 ^0 014 44 a) 1*10 #rj
Pk 44 V 0 IU r. 00 u > c; H0 Vii- 2t

LL. -

-0 "0
(c j=




0 co c "

C~j 0

(|UW/Bu) HO9


* I--


between the Hind III restriction site at +1179 and the Hpa I

siteat +5013. This sequence

contains 6 NF-IL6 consensus sequences compared to 1 such sequence

in the downstream fragment from Hpa I to the 3'Hind III site.

The NF-IL6 enhancer-binding protein is also now termed

CCAAT/enhancer-binding protein # (C/EBP3) (Stein et al., 1993).

Binding of transcription factors NF-IL6 and NF-KB, to sites

within the same promoter, has been shown to synergistically

activate transcription of genes coding for IL-6 and IL-8

(Matsusaka et al., 1993; Mukaida et al., 1990), and serum amyloid

A (Li and Liao, 1992; Betts et al., 1993). In addition, NF-IL6

and NF-KB have been shown to functionally and physically

associate and synergistically stimulate expression of reporter

genes from synthesized promoters containing a NF-IL6 binding

site, a TATA sequence and a NF-KB binding sequence (Stein et al.,

1993). The p50 subunit of NF-KB has been shown to specifically

bind proteins encoded by cDNAs for the NF-IL6 transcription

factor (LeClair et al., 1992). This association has been shown

to take place between the N-terminal Rel homology domain of NF-xB

and the ldUcine zipper motif of NF-IL6. The interaction mediates

functional synergy in promoters with NF-IL6 binding sites (Stein

et al., 1993).

My studies are the first to demonstrate enhanced expression from

a MnSOD genomic sequence combining 5' promoter sequences with a

distant downstream sequence containing a portion of intron 2, all

of exon 3, intron 3, exon 4, and part of intron 4. Based on

sequence analysis, it is possible that this enhanced expression

is mediated by an interaction between proteins within DNase I

hypersensitive site 1 (subsites 6 and 7) with NF-xB or NF-IL6

bound transcription factors in the downstream enhancer sequence.

Interestingly, however, the MnSOD 5' flanking sequence that I

have shown induces LPS and TNF expression does not contain either

NF-KB or NF-IL6 consensus binding sequences.

Although most enhancers function in an orientation independent

manner, the enhancer from the MnSOD gene was much more effective

at boosting expression when present in the same, 5' to 3',

orientation as the promoter sequence. The mechanism by which

distant enhancers interact with promoter elements to increase

transcription is unknown, but most of the possibilities mentioned

in the literature propose direct contact between proteins bound

to the enhancer and promoter sequences alterring DNA architecture

and increasing the frequency of transcript initiation by an as

yet unknown mechanism. The intervening DNA may be bent (Giese

and Grosschedl, 1993), looped out (Schleif, 1987), or wrapped up

(Wolffe, 1994). The DNA may be twisted or unwound from histones

(Felsenfeld, 1992), or the regulatory proteins may recognize a

specific site on DNA and then slide along the DNA (or thread it

through the protein) until the second critical regulatory site is

encountered (Ptashne, 1986). Alternatively, after binding to its

specific sequence, a regulatory protein may then assist the

binding of another protein to adjacent sequences, which in turn

helps another to bind next to it, and so on, until a procession


of proteins has oozed out from the control sequence to the gene

where transcription is initiated (Ptashne, 1986). These latter

two models seem unlikely when enhancer and promoter elements are

separated by thousands of base pairs--as in the MnSOD gene. The

deletion analysis data presented, does not permit us to determine

the exact bases that are critical for the MnSOD enhancer and

promoter activity. Alternative experimental approaches will be

required as described in Chapter 6.



Nitric oxide (NO), a free radical gas with crucial

immune, cardiovascular and neurological second messenger

functions, is synthesized by several isoforms of the enzyme

nitric oxide synthase (NOS)(Moncada, et al., 1991; Nathan,

1992; Lowenstein and Snyder, 1992; Stuehr and Griffith,

1992). Expression of the various isoforms is regulated

differently. Inducible NOS (iNOS), cloned from a murine

macrophage cell line (Lyons, et al., 1992; Xie, et al.,

1992; Lowenstein, et al., 1992) and human hepatocytes

(Geller et al., 1993), is regulated at the transcriptional

level, whereas expression of the endothelial and neural

isoforms is constitutive (cNOS), and is regulated

predominantly at the post-transcriptional level.

Most studies on the expression of NOS have inferred

changes in NOS activity from measurements of nitrites (NO2")

and nitrates (NO3-), soluble, stable metabolites of NO gas.

Conclusions about which NOS isoform produced the observed

increases in NO or N03-, have depended largely upon indirect

evidence including cell type, time course, stimuli, and

calcium/calmodulin dependence or independence (Stuehr

and Griffith, 1992). However, recent cloning of the cDNA

encoding the transcriptionally regulated macrophage iNOS

has made possible direct and specific measurements of the

regulation of its transcription in various cell types.

In murine macrophages, treatment with bacterial LPS or

cytokines such as INF-y increases iNOS mRNA and levels of

NO and NO3" (Lorsbach et al., 1993; Chesrown et al., 1994).

Enhanced killing of microbes and tumor cells also results

(Green et al., 1991; Higuchi et al., 1990). Conversely,

treatment of murine macrophages with other cytokines such as

IL-10, IL-4 and TGF-f have been reported to inhibit NO

production and killing activity (Gazzinelli, et al., 1992;

Oswald, et al., 1992; Nelson, et al., 1991). There have

been no similar reports on human macrophages or peripheral

blood monocytes (PBMC).

My studies have characterized the induction of iNOS

steady state mRNA levels following LPS stimulation and the

mechanisms involved. I chose to evaluate mRNA levels rather

than NOS activity measurements in order to directly link

stimulus-dependent responses to a specific NOS isoform. I

examined this induction in both murine cell lines and rat

primary cell cultures. Cytokine treatment failed to induce

iNOS mRNA in human PBMC.

In addition to decreasing NO2 production, the

cytokines IL-10 and TGF-3 have been shown to down-regulate

many other macrophage functions (De Waal Malefyt et al.,

1991a and 1991b; Fiorentino et al., 1991). However,

Corradin et. al. (Corradin et al., 1993) recently

demonstrated enhanced NO0 release by IL-10 in bone marrow-

derived mouse macrophages stimulated by INF-y plus TNF-a.

IL-10 had no effect in cells stimulated with INF-y alone or

in combination with LPS in this report. Therefore, I

studied the effect of IL-10 and TGF-0f treatment on the

induction of iNOS mRNA by LPS and INF-y. Results

demonstrated opposite effects of IL-10 and TGF-f on INF-y

induced iNOS mRNA steady state levels providing further

documentation of the complex regulation of immunological and

inflammatory macrophage functions.


Characterization of iNOS Response to LPS

To define the potency of LPS on iNOS mRNA induction in

a murine macrophage cell line, I exposed RAW264.7 cells to

increasing concentrations of LPS for 8 hours before RNA

isolation. Figure 5.1 shows the results of the Northern

analysis. In untreated RAW264.7 cells, no iNOS mRNA was

detectable even after prolonged exposure of autoradiographs.

The iNOS mRNA levels increase in response to LPS in a dose-

dependent fashion. Induction of the message was apparent at

20 ng/ml, and was near maximal at 200 ng/ml.

To compare the time-dependent LPS induction of iNOS

mRNA in two different murine macrophage cell lines, and to

demonstrate that this response is not unique to a single

cell line, I exposed RAW264.7 and J774 cells to LPS (200

C co

4C 4 V
0 rr- 0

** t 0 0
Cr-I U

t 0

-4 0

0 0

V 0
S0 V *
o '-4l 0
t5 *4 I8-C
0 k C(

-4 0 m 0

S4 4.*)
to 0) -10
0 0r-4 C
S4JA & 4J


oH 0t 0)

#0 0 4 0




ng/ml) and isolated RNA at varying times. Figure 5.2 from

RAW264.7 cells illustrates obvious induction of iNOS mRNA at

4 h, with a peak at 12 h, followed by a decrease from 24 to

40 h. Overexposure of this autoradiograph showed that

induction of the iNOS mRNA was detectable by 2 h. In J774

cells, the kinetics of LPS induction of iNOS mRNA were very

similar (not shown). Basal levels of iNOS mRNA were

undetectable. Induction of the message was obvious at 4 h,

peaked at 8 h, and was decreasing at 24 h. Again in these

cells, overexposure of the autoradiograph revealed

detectable iNOS mRNA at 2 h.

To show whether the effect of LPS on iNOS mRNA levels

was limited to murine macrophage cell lines, I isolated

unstimulated and unelicited peritoneal macrophages from

pathogen free Sprague-Dawley male rats. Following their

adherence to tissue culture plastic, I treated them with LPS

at 200 ng/ml for 6 hours before isolating RNA. Unlike

murine cell lines, Figure 5.3 shows that primary culture rat

macrophages unstimulated in vitro have a detectable basal

expression of iNOS mRNA. Like the murine cell lines,

expression of iNOS mRNA is strongly enhanced by LPS.

Effect of Actinomycin D and Cycloheximide on LPS Induction
of iNOS mRNA in RAW264.7 Cells

LPS induction of iNOS mRNA steady state levels could

result from an increase in transcription, a prolongation of

iNOS mRNA half-life, or both. To address these

alternatives, RAW264.7 cells were grown to near confluence

4 -0

CM to*-V
S0 0

:) V m V 4) *4
r-1 0 010

n 0.0

# M r ,.,o4J
09 0 )44 t

0 040
En -AC -rq

0 $"4 V A *,.
C > 0)
w O
Z 0 0) 4 4
(M t -,'

4 OI J E
OQ 0)

0 Hv C t0

U) *H a )

*-4 W 4- 10
O0- 4.C) 0

0 to 0 0

0 q rX
' 0 4) l

- M)4 04 H

4(D .Vo 04
to V v to -i-4
0 0C 1 U 0
14 -H U 0 4)
0 ) 1-H >
0 E to
U0 CV 0% *
.4) V -H
-4 0) r. I 0) >4

IN 4 to C9 to
* 4 00
ILn ) 'o 4) C
a) 4J N )
3-l r-4 (A 04-)
-4 0 Hr- k
H 4) to >4 >4 0
Ow 0 -4 Q C 2









U) V
0 4.'


Figure 5.3. Northern analysis of RNA from unelicited
resident rat peritoneal macrophages exposed to LPS.
Peritoneal macrophages were plated at 106 per dish and
after 3 h, non-adherent cells were removed by PBS
washing. Adherent macrophages were treated with medium
alone (C) or LPS (200 ng/ml) for 6 h. Levels of iNOS
and cathepsin B mRNA were determined by Northern
analysis. RNA loading in each lane was documented by
cathepsin B control.





under standard conditions and then treated for 4 hours with

LPS (50 ng/ml) alone or co-treated with LPS (50 ng/ml) and

increasing concentrations of actinomycin D or cycloheximide.

I employed a dose-dependent analysis of these inhibitors due

to the observation that both agents were toxic to these

cells at concentrations typically used (Visner et al.,

1989). Figure 5.4A shows the response of iNOS mRNA to LPS

and actinomycin D alone and the effects of increasing

concentrations of actinomycin D on LPS-induced levels of

iNOS mRNA. At concentrations ranging from 4nM to 4 pM,

actinomycin D alone caused no detectable iNOS mRNA

expression, and LPS-induced iNOS mRNA levels were decreased

in a dose-dependent fashion by low doses of the

transcriptional inhibitor. The LPS induction was completely

inhibited by actinomycin D at 400 nM.

The effect of LPS on iNOS expression in RAW264.7 cells

was first detectable after 2 h of LPS exposure (Figure 5.2).

This provides time for new protein synthesis, which may be a

prerequisite for the LPS induction. To determine if new

protein synthesis played a role in iNOS induction, RAW264.7

cells'were exposed for 4 h to LPS alone (50 ng/ml),

increasing concentrations of cycloheximide (20 nM to 20 pM)

or co-treatment with LPS and increasing concentrations of

cycloheximide. Figure 5.4B illustrates the dose-dependent

cycloheximide inhibition of the LPS induced increase in iNOS

mRNA in RAW264.7 cells. Overexposure of the autoradiograph

revealed underloaded but intact RNA in the 2uM cycloheximide

H 4J-4i
p a5 0 0

2 LO *

44 04% 0
0-I0 I .0

P-4 C
9 *- 0

4J 0 4-)P

.O *-E--4
0 4 0 04

0. C 0 -4

OCi 04 *

0 0 0: r
r4 OdH ti
to H

0 4 W4
0 r- H

0 9 0


ua 4 -H a)
4) 0 0 to to
S4) 0u 4)
E0 0 o 0
44 N -H -0 0 >

i+ +

+ +

I +


c n
a .





*'4l 0

,. : Z 0

-U 0) 14
U4 C ) )
.,.0 C ..-.0 :- U
0 o 4-i t

0 r4 r.%HO
S -4 > 0,

Qa ) m 0

. .4 4 ,-4 r
ol 0 H OH
E to M zC
S0 0) 10
v0 *0 0 4-) 0)
C'O tn 4 0
*H M 4C
0 04 H 4 AA
m 0 0 -
4 0 04 *r- 4
0 *4 4 a
4) 0

020 0
0 0 0

)x x 04a)
X0 ) 0 0

P41 *- V- 4
0 -4 -.4
0$4 u M Cm0

o 0 0) 0) 0

0 0 0V

4 0 0 (10
0-n 0 0 V r
S0 w0 -4 0
0)O & (8

r- *4 r. 4 a)
00) Z 0
Q ) 0 0 9: t

P.-,4,4 U <

~I + +




alone lane. The inhibition is evident at 2 pM cycloheximide

and was almost complete at 20 pM of this inhibitor.

IL-10 and TGF-B Effects on LPS and INF-'y induction of iNOS
mRNA Levels in RAW264.7 Cells

To further characterize the induction of iNOS mRNA, I

treated confluent RAW264.7 cells for 8 hours with LPS (50

ng/ml) and INF-y (50 U/ml) alone or in combination. Then

using co-treatment, I assessed the ability of TGF-f (10

ng/ml) and/or IL-10 (100 U/ml) to modulate the LPS and INF-y

induced iNOS mRNA expression. Figure 5.5 shows the absence

of iNOS mRNA expression in unstimulated RAW264.7 cells (Bar

1). Exposure to either IL-10 or TGF-j alone showed no

induction after 8 hours (Bars 2 and 3). LPS and INF-y alone

(Bars 4 and 8) each induced iNOS mRNA to similar levels

after 8 hours. However, IL-10 co-treatment synergistically

augmented INF-y induction of iNOS mRNA expression (Bar 9)

while slightly decreasing the induction of iNOS mRNA by LPS

(Bar 5). In contrast, TGF-j had only minor inhibitory

effects on LPS (Bar 6), INF-y (Bar 10) or their combination

(Bar 12).

As reported by others (Lorsbach et al., 1993), I found

a synergistic increase in iNOS mRNA expression following

exposure of cells to both LPS and INF-y (Bar 7). IL-10

treatment combined with LPS and INF-y further increased this

high iNOS mRNA expression (Bar 11). Finally, cells treated

simultaneously with LPS, INF-7, IL-10 and TGF-6,

ro w p 1 < o P 44 44
C. r_4 1 0 r. C .-I to 0 0
<-)I fH 0 0 m
(0U + I H 0) E-i0) 0 0
a ,- > ) 4 ,-I V
r- -4 En?-+2+ I 4J 0) 4.
h I r1 o Po r to % P ( to r4-
0 ID Z I E-Z ) 0 ) W 4 M
,O > PE 00H)w 0013
c -. + >*-rq c oH -4
0*0 vH AM V 4 0 00 0 0

H4)q 4 0r40 W 0Q
4 'P + -oiH V 00 4) *-I4
to ) 0 ) ) ) t (0 V V 0
O r-- +4 u0 M 0UC
4E1 0 H *. EH ~ *C 4 4-
Su 1C '.HO z 0 00 01)t4
0 3 PH (L) aI C o- 0 4 N. 0 P i
i Mo +00 0 EIH c
C -r-4 U V 4)0 -H40)P 0
M r-4 t-4 W Ca < 9C 4 ) V W O) 0 4
O 0) r. 1i 0 4 -) 0A.I-V
-,.40 0 104 0 0 .:- 4 41
-o u .- o0 .0 4 p 0 00
I M 0 Z P 0 w m 0 H O0r-4
P .4 H MrMIV-4Ir. $4u '-4
Ha 0 + l A 40) IB
waON+ -Cm napHN V
4O N r i ? -N-ri-I rE ) iv. w C
%H .-N 4 4r* 4J 4) r- Hid 4-4 P
H g -OOZk O O OM C44 0
to) 0 :3- m41 2 M4 0 P
_r4 $w- Blz 0U E H 0< 0 $4 M to 0
10 M0( 0) 3Z. X X r14 0rU 10 r.
C04 W) i a (D o >i V 0-1
I r: 1 P N c 4 0 r-I 9 1 $M
c IV EU o EU ) 04 a0) to EU-o-I mi4 N.c1
o U) r. ri4 ) 9C

S4) 1r-0 E O N ?r v 00 0r O
-*(0 *H H aF4) +M 4 rj 0

.A 0 + H 4 VP .9 rtf 441 Id
w w + MUCOU 0O0004
0 0* M + r-I W >< *V Z 0 A
Q-H 0 10 0>0 C-- 0>> o 0
VJ PO rj 4134 4) 4) 0 C 4) > > f
u 0 E- Z r-4 V o-4-i >I* -i4 ) $4
*'dmeUQafl *-a XU-r-I EU 3

%U0 V> U HO9 OrQ V CV*VA *
It 9 0) 1 -. W N 0 014)H rd 0)
*d $4VN4 $4 E-1E-4 -A-4 r. 0) > CO
) V U 00 00 -1 Q 0 P-4 A -1 t-d)r 4- V
I 0) P 4O-4 4 M C 4) rj W -l
&z.$4.- -0 (-04MV M 10$ V -' I M- Or-
*-HZ )-.OCQMM Cq r.-r 0 EU U r W 0) )
Pri H 3 N r-I %- -r-4 EU C r. > U) :3 E0 P $4 U



w 1

VNEIW SON! dO S13A31 3AIiV-13I








expressed iNOS mRNA at a level similar to the combination of

LPS, INF-7 and IL-10 (Bar 13).

Effect of LPS and Cytokine Mixtures on iNOS mRNA Levels in
Other Rat and Human Cell Types

To determine whether iNOS mRNA could be induced in

other rat cells or in human cells, I did Northern analyses

looking for iNOS mRNA expression in rat pulmonary artery

endothelial cells, a rat lung epithelial cell line (L2

cells), and human PBMC from both a healthy volunteer and a

patient with chronic granulomatous disease. In none of

these cells was iNOS mRNA basal expression detectable nor

was induction observed after 8 h exposure to any of the

single agents (LPS, TNF-a, IL-1f0, and/or INF-7) or a mixture

of these four cytokines. The cytokine mixture tested was

the same as that reported to maximally induce the expression

of iNOS mRNA in human hepatocytes (Nussler et al., 1992).

Using a rat cathepsin B probe as a control, Northern

analysis of the membranes from these experiments confirmed

equal RNA loading. Inclusion of a lane of RNA from LPS

treated RAW264.7 cells confirmed the function of the

radiolabeled iNOS cDNA probe. (Data not shown.)


Activation of any of the isoforms of NOS results in

production of NO that rapidly reacts with oxygen in water -

ultimately forming stable nitrates and nitrites (Stuehr and

Griffith, 1992). These compounds are easily measured in

tissue culture supernatants as indirect measures of total

NOS expression. Recent cloning of cDNAs for specific NOS

isoforms from different tissues and species now allows more

direct measurement of expression of specific NOS isoforms

(Lyons, et al., 1992; Xie, et al., 1992; Lowenstein, et al.,

1992; Marsden, et al., 1992; Kishimoto, et al., 1992;

Nunokawa, et al., 1993). Characterization of the cytokine-

inducible NOS (iNOS) confirmed that it is regulated at the

transcriptional level. However, except in more recent

reports, most studies on the regulation of iNOS expression

have shown measurements of nitrates and nitrites in tissue

culture supernatants rather than changes in iNOS mRNA

levels. My studies describe the effects of various

inflammatory mediators on the steady state levels of iNOS

mRNA in several cell types from mouse, rat and humans.

To extend previous observations on iNOS regulation, I

used a mouse macrophage-derived iNOS cDNA clone to analyze

regulated expression in several cell types including two

murine macrophage cell lines, as well as in primary cultures

of rat peritoneal macrophages. Northern analyses showed

that low concentrations of bacterial LPS induced iNOS mRNA

levels dramatically in each case. The cell lines had

undetectable basal levels of iNOS mRNA while the unelicited

rat peritoneal macrophages demonstrated low basal iNOS mRNA

expression. In two mouse macrophage cell lines the time

course of LPS-mediated iNOS induction was similar. It

occurred rapidly, within 2 hours, peaked at about 12 hours,

and was declining by 24 hours. Inhibition of iNOS mRNA

induction by low concentrations of actinomycin D is

consistent with the requirement for de novo transcription.

Cycloheximide, at a very low concentration, inhibited LPS-

mediated increases in iNOS mRNA. This suggests that newly

synthesized polypeptide(s) are either necessary for LPS-

signaled iNOS transcription, and/or that newly synthesized

polypeptide(s) attenuate a pre-existing ribonuclease


Both TGF-# and IL-10 have been shown to inhibit

multiple macrophage functions. Specifically, IL-10 has been

reported to decrease INF-y induced NO biosynthesis (Cunha

et al., 1992; Oswald, et al., 1992; Bogdan et al., 1991)

and INF-y induced parasite killing (Gazzinelli et al., 1992)

by mouse peritoneal macrophages elicited by intraperitoneal

thioglycolate. In contrast, I observed that IL-10 co-

treatment with INF-7 dramatically augmented the iNOS mRNA

induction in a murine macrophage cell line. IL-10 also

further increased the iNOS mRNA level induced by the

synergistic effect of LPS plus INF-7. These data contradict

the previous reports where NOS activity was measured The

differences in these observations may result from the

different cells studied or the existence of multiple NOS

isoforms, and be due to the fact that NOS expression was

evaluated by different means. We intentionally chose to

study mRNA levels using a gene-specific probe in order to

avoid any shortcomings of the less direct activity assays

which cannot differentiate between isoforms. It is,

therefore, possible that the discrepancy with previous data

may arise from differential regulation of each isoform at

either the transcriptional or post transcriptional level.

In addition, this may also represent a difference between

the cells studied. Namely, the previous studies (Gazzinelli

et al., 1992; Cunha et al., 1992; Oswald, et al., 1992)

evaluated macrophages from an inflammatory exudate that had

been elicited by intraperitoneal periodate or thioglycolate

and stimulated by INF-7, whereas we studied the effects of

IL-10 alone or in combination with LPS and INF-7 on a murine

cell line. Clarification of this issue will require the

availability of inhibitors specific to each of the NOS

isoforms, so that activity measurements can independently

reflect the different isoforms.

More recently, IL-10 has been reported to enhance NO0

release from murine bone marrow-derived macrophages

stimulated in vitro with INF-7 plus TNF-a, but not INF-7

alone (Corradin et al., 1993). These authors also report

that IL-10 inhibited NO release from these cells treated

with LPS in combination with INF-7. Thus, in this study as

well as the present study IL-10 has been shown to increase

iNOS expression. In the macrophage cell line employed in

the present study, IL-10 increased iNOS mRNA induced with

INF-7 alone or INF-7 in combination with LPS. It is

consistent with the notion that the regulation of iNOS is

very specific to cell type, cell activity and species.


Again based on NO2 assays, TGF-3 has been reported to

inhibit NO biosynthesis in activated macrophages (Ding et

al., 1990; Nelson et al., 1991), as well as in cytokine-

induced human smooth muscle cells (Junquero et al., 1992)

and in rat renal mesangial cells (Pfeilschifter and Vosbeck,

1991). Therefore, I also examined the effects of TGF-f

alone on iNOS mRNA expression and on LPS and INF-7 induced

expression of iNOS mRNA levels. In contrast to IL-10

effects, and in agreement with published activity data co-

treatment of cells with TGF-3 and 1)LPS, 2)INF-7 or 3)LPS

plus INF-7, slightly decreased iNOS mRNA induced levels. As

IL-10 and TGF-f have been shown to suppress cytokine release

from macrophages by different mechanisms (Bogdan et al.,

1992) they may also employ different mechanisms in

regulating iNOS mRNA levels.

There have been 2 reports of NO production and

enhanced microbial killing in infected human macrophages

derived from peripheral blood monocytes (Denis, 1991; Mufloz-

Fernandez et al., 1992). Another recent report infers basal

NO biosynthesis and release from human neutrophils and

peripheral blood monocytes obtained from both healthy

volunteers and patients with chronic granulomatous disease

(CGD)(Condino-Neto et al., 1993). However, other efforts to

induce NO- biosynthesis in human mononuclear phagocytes have

been largely unsuccessful (Schneemann et al., 1993). This

has prompted one group of investigators to conclude that NOS

is not a part of the antimicrobial defenses of human

macrophages (Schneemann et al., 1993). In agreement with

these authors, I observed no induction of iNOS mRNA in human

macrophages isolated from peripheral blood monocytes of a

healthy volunteer or a patient with CGD stimulated with the

mixture of inflammatory mediators used successfully by

Geller et al. (1993) to induce iNOS mRNA in human

hepatocytes. In addition, I did not detect any cytokine-

dependent effects on iNOS expression in human or rat

pulmonary artery endothelial cells, although other

endothelial cells have been reported to have inducible NO

production. These data, therefore, extend to the mRNA level

previous reports that have found no inducible NOS activity

in human macrophages.

In conclusion, these results underscore the importance

of independently evaluating the specific enzymatic isoforms

of NOS following stimulation of nitric oxide production.

They also further demonstrate important species and cell

type specificity of iNOS regulation.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E4D0P7G96_9AS0U8 INGEST_TIME 2012-02-20T21:00:47Z PACKAGE AA00009044_00001