The roles of metallothionein expression and dietary zinc in zinc metabolism and cytoprotection in metallothionein transg...

MISSING IMAGE

Material Information

Title:
The roles of metallothionein expression and dietary zinc in zinc metabolism and cytoprotection in metallothionein transgenic and metallothionein knockout mice
Physical Description:
viii, 109 leaves : ill. ; 29 cm.
Language:
English
Creator:
Davis, Steven Roger, 1968-
Publication Date:

Subjects

Subjects / Keywords:
Metallothionein -- Physiological effect   ( lcsh )
Gene expression   ( lcsh )
Zinc -- Metabolism   ( lcsh )
Mice -- Physiology   ( lcsh )
Food Science and Human Nutrition thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Food Science and Human Nutrition -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2000.
Bibliography:
Includes bibliographical references (leaves 94-108).
Statement of Responsibility:
by Steven Roger Davis.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 025800190
oclc - 46687039
System ID:
AA00020439:00001

Full Text










THE ROLES OF METALLOTHIONEIN EXPRESSION AND DIETARY ZINC IN
ZINC METABOLISM AND CYTOPROTECTION IN METALLOTHIONEIN
TRANSGENIC AND METALLOTHIONEIN KNOCKOUT MICE














By

STEVEN ROGER DAVIS


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

UNIVERSITY OF FLORIDA


2000






























This work is dedicated to the memory of my aunt, Leona Goguen. She has been, and will
always be my role model. I dream that one day, I too will possess the wisdom, character,
and courage that she displayed each day of her life.














ACKNOWLEDGMENTS

I would like to thank my committee chairman, Dr. Robert J. Cousins for his

guidance throughout this work.

I would like to thank my committee members Dr. Jesse Gregory, Dr. Rachel

Shireman, Dr. Susan Percival, and Dr. Stephen Roberts for their advice and input into

this work.

I would like to thank Walter Jones, Virginia Mauldin and Warren Clark for their

technical assistance and sense of humor throughout my stay in Dr. Cousins' laboratory.

I would like to thank Dr. Nora Holquist, Dr. Christina Khoo, Dr. Barbara Davis,

Dr. Vicki Sullivan, Dr. Lorraine Lanningham-Foster, Dr. Jay Cao, Monique Coy,

Jennifer Moore and Juan Liuzzi for their friendship and support.

I would like to thank Dr. Raymond Blanchard and Dr. Robert McMahon for their

friendship and tutoring in the laboratory.

I would like to thank my family for their patience during this process.

Lastly, I would like to thank Amy Mackey for her technical and professional

assistance and especially for her personal support during the final stages of this work.















TABLE OF CONTENTS

page


ACKNOWLEDGMENTS ................................................................................................ 111i

ABBREVIATIONS ....................................................................................................... vi

ABSTRACT ...................................................................................................... ...vii

CHAPTERS

1 INTRODUCTION .......................................................................................................... 1
Literature Review ..........................................................................................................2
Hypotheses and Research Objectives ...........................................................................9

2 THE EFFECT OF METALLOTHIONEIN EXPRESSION ON ZINC
ABSORPTION IN METALLOTHIONEIN TRANSGENIC AND
METALLOTHIONEIN KNOCKOUT MICE............................................................. 11
Introduction .......................................................................... ....................................... 11
Materials and Methods ....................................................................... ..................... 12
R results ........................................................................................................... .......... 15
D discussion ................................................................................................................... 2 1

3 REGULATION OF METALLOTHIONEIN EXPRESSION
AND ZINC METABOLISM BY DIETARY ZINC IN METALLOTHIONEIN
TRANSGENIC AND METALLOTHIONEIN KNOCKOUT MICE ........................ 28
Introduction ................................................................................................................. 28
Materials and Methods ................................................................................................ 33
R results ......................................................................................................................... 32
D discussion ................................................................. .................................................. 36

4 THE EFFECTS OF METALLOTHIONEIN GENE EXPRESSION AND
SUPPLEMENTAL DIETARY ZINC IN PROTECTION AGAINST
HEPATOTOXICITY IN METALLOTHIONEIN TRANSGENIC AND
METALLOTHIONEIN KNOCKOUT MICE ............................................................ 41
Introduction ............................................................................................ ..................... 4 1
Materials and Methods ................................................................................................ 43
R results ...................................................................................... ................................... 47
D discussion ................................................................... ............. ..................................... 56










5 EFFECTS OF METALLOTHIONEIN GENE EXPRESSION AND
SUPPLEMENTAL ZINC IN PROTECTION AGAINST OXIDATIVE
STRESS IN PRIMARY HEPATOCYTE CULTURES FROM
METALLOTHIONEIN TRANSGENIC AND METALLOTHIONEIN
KN OCKOUT M ICE .................................................................................................... 66
Introduction ........................................................................................... ...................... 66
M materials and M ethods .............................................................................. ............ 67
Results .......................................................................... ............... ......... 72
D discussion ................................................................. .................................................. 76

6 SUMMARY AND CONCLUSIONS ....................................................................... 85
Zinc and Metallothionein in Zinc Absorption and Metabolism................................... 85
Zinc and Metallothionein in Defense Against Oxidative Stress ..................... ........89

LITERATURE CITED ..................................................................................................... 94

BIOGRAPHICAL SKETCH ................... .............................................................. ........... 109














ABBREVIATIONS


AAS atomic absorption spectrophotometry
ALT alanine aminotransferase enzyme
BBM brush border membrane
CK control mouse strain for KO mice
CO corn oil
CT control mouse strain for TG mice
Dex dexamethasone
DTNB 5,5'-dithio-bis(2-nitrobenzoic acid)
GSH reduced glutathione
GSSG oxidized glutathione
i.p. intraperitoneal
I1-1 interleukin- 1
11-6 interleukin-6
KO metallothionein-null (knockout) mouse
LDH lactate dehydrogenase enzyme
LPS lipopolysaccharide
MRE metal response element
MT metallothionein
MTT thiazoyl blue
NPT nonprotein thiol
pv perivenous
SPAEC sheep pulmonary artery endothelial cell
SSA sulfosalicylic acid
TBH tertiary-butyl hydroperoxide
TCA trichloroacetic acid
TG metallothionein overexpressing transgenic mouse
TT total thiol
WME William's medium E














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

THE ROLES OF METALLOTHIONEIN EXPRESSION AND DIETARY ZINC IN
ZINC METABOLISM AND CYTOPROTECTION IN METALLOTHIONEIN
TRANSGENIC AND METALLOTHIONEIN KNOCKOUT MICE

By

Steven Roger Davis

December 2000

Chairman: Robert J. Cousins
Major Department: Food Science and Human Nutrition Department

The biochemistries of zinc and metallothionein are intricately linked. Zinc

induces metallothionein gene expression and allows the protein to resist proteolysis. In

turn, binding of zinc to metallothionein allows cellular zinc accumulation, and

metallothionein may be involved in intracellular zinc trafficking. Further, metallothionein

induction and zinc accumulation often are coupled. This relationship makes it difficult to

determine if effects of zinc supplementation are due to zinc, metallothionein induction, or

both. Metallothionein overexpressing (transgenic) mice and metallothionein null

(knockout) mice provide unique models to study the effects of zinc supplementation and

metallothionein gene expression on physiological processes. We used these mouse

models to determine (1) the effects of metallothionein expression and dietary zinc intake

on zinc absorption and tissue zinc accumulation, and (2) the effects of zinc








supplementation and/or metallothionein gene expression on susceptibility to oxidative

stress.

Metallothionein expression was inversely proportional to serum zinc

concentrations 2 h after an oral zinc dose, which strengthens the theory that

metallothionein impedes zinc absorption. Intestinal zinc accumulation was inversely

related to metallothionein expression, however, which argues that metallothioneinein

does not act by simply sequestering zinc in the intestine. Metallothionein protein

expression was directly proportional to intestinal and liver zinc concentrations after 3-7 d

of zinc supplementation, but only at dietary zinc levels 20- to 50-fold of the requirement.

Intestine and liver zinc concentrations did not change over a wide range of zinc intakes,

and knockout mice maintained serum zinc levels as well as control mice. These points

suggest that at typical zinc intakes, and in the absence of significant stresses,

maintenance of zinc homeostasis is not metallothionein-dependent.

Metallothionein expression protected against carbon tetrachloride-induced

hepatotoxicity. Neither zinc supplementation nor metallothionein overexpression

provided further protection, however. The heightened toxicity in KO mice after carbon

tetrachloride treatment was associated with their lack of control over zinc homeostasis. In

primary hepatocyte cultures, metallothionein induction was associated with increased

susceptibility to oxidative stress. This was likely due to the observed depression of

cellular glutathione. These results argue against direct antioxidant roles for

metallothionein expression and supplemental zinc in mouse liver.














CHAPTER 1
INTRODUCTION


The objectives of the research reported in this dissertation were to determine (1)

the roles that dietary zinc and metallothionein have in regulation of zinc absorption and

tissue distribution, and (2) the roles that dietary zinc and metallothionein have in defense

against oxidative stress. The production of mouse strains with perturbed metallothionein

expression provided the opportunity to investigate the above objectives under conditions

of metallothionein absence and metallothionein excess in an intact animal model.

Previously, investigators studying the effects of metallothionein expression on zinc

metabolism or oxidative stress used a myriad of treatments to alter metallothionein levels

before their experiments. Although they successfully altered metallothionein levels, other

physiological pathways may have been perturbed as well. For example, zinc pretreatment

induces metallothionein, but undoubtedly affects other components of the machinery that

regulates zinc metabolism (e.g., zinc transporter abundance). Similarly, treatment with

other metals, hormones, and cytokines is known to affect more than metallothionein. By

using mice with altered metallothionein expression the complications associated with

such treatments were avoided. Because of the artificial nature of these models, we must

use caution when interpreting the results of studies in which they were used.

Nevertheless, the results of such studies provided strong support for existing theories, as

well as new insights for putative roles for metallothionein.








Literature Review

Zinc

Zinc is an essential nutrient whose recommended dietary allowance (RDA; 15

mg/d for men, 12 mg/d for women) ranks with iron as highest among the trace elements

(National Research Council 1989). Zinc functions in more than 50 enzymes, serving in

catalytic, structural, and regulatory roles (reviewed by Vallee and Falchuck 1993,

Cousins 1996). These enzymes are involved in the synthetic and catabolic pathways of

many biomolecules, including proteins, nucleic acids, carbohydrates and lipids. Zinc also

is a component of zinc finger transcription factors. Zinc is distributed throughout the

body, including 57% residing in skeletal muscle, followed by 29% in bone, 6% in skin,

and 5% in liver (reviewed by Jackson 1989). Within cells zinc is distributed ubiquitously.

Thiers and Vallee (1957) reported that 43% of liver zinc is cytosolic, 37% nuclear, 13%

microsomal, 5% mitochondrial, and 2% in connective tissue. Zinc deficiency is

associated with a 20% reduction in whole body zinc, but it is unknown which

intracellular pool(s) is most affected (Hambidge 1989). Although severe zinc deficiency

is uncommon, mild zinc deficiency may be prevalent in many parts of the world (Prasad

1991).


Metallothionein

Metallothioneins are a family of small (6-7 kDa), cysteine-rich metal binding

proteins found in vertebrates and invertebrates (Dunn et al. 1987). These proteins are

further characterized by a lack of histidine residues, aromatic amino acids, and disulfide

bonds. The protein is capable of binding up to (10) copper atoms, or up to (7) cadmium

and/or zinc atoms in two distinct clusters of the protein (Neilson et al. 1985).








Metallothioneins are thought to function in metal homeostasis, including zinc absorption

and tissue distribution, and protection against heavy metal toxicity (Cousins 1985, Liu et

al. 1995, Masters et al. 1994A). Metallothioneins are transcriptionally regulated by

metals through metal response elements (MREs) in their gene promoters (Carter et al.

1984, Dumrnam and Palmiter 1981, Stuart et al. 1984). Metallothioneins also are

transcriptionally regulated by glucocorticoid hormones and cytokines, which implies a

role for metallothioneins in inflammatory and stress-related responses, such as the acute

phase response (Cousins and Leinart 1988, Etzel et al. 1979).

Although four forms of metallothionein have been discovered, the two most

widely expressed are metallothionein-1 and metallothionein-2 (Hunziker et al. 1995,

Kagi et al. 1974). These forms are found in most tissues, and are especially prevalent in

the liver, kidney, pancreas, and intestine (Hunziker and Kagi 1985). Metallothionein-3

and metallothionein-4 forms were discovered recently, and are expressed mostly in the

brain and skin, respectively (Masters et al. 1994B, Quaife et al. 1994). Subforms of

metallothionein (e.g., MT-la) exist in primates, but not in mice (Stennard et al. 1994,

Twunoo et al. 1978). Because metallothionein-1 and metallothionein-2 are the

predominant forms in the liver and the intestine, they are the focus of discussion from

this point on.


Metallothionein in Zinc Absorption and Tissue Distribution

Chapters 2 and 3 of this dissertation focus on the role of metallothionein in

dietary zinc absorption and tissue zinc accumulation. The quantity of zinc absorbed by

the body depends on several processes. These include the ingestion and digestion of zinc-

containing foodstuffs, uptake of zinc by the intestinal mucosa, and transport of zinc from








the intestine to the vascular supply. Zinc uptake by the intestine, followed by transfer to

the portal blood supply is referred to collectively as zinc absorption. Zinc is absorbed

throughout the entire small intestine (Lee et al. 1989). The predominant site in humans is

the jejunum, while both the duodenum and jejunum appear to be major sites in rodents

(Davies 1980, Lee et al. 1989). Although some information is known about zinc

absorption, a clear understanding of this process and its regulation has not been reached.

The process can be divided into three parts: (1) uptake from the lumen at the brush

border membrane (BBM), (2) transport across the epithelial cell, and (3) transfer of zinc

across the basolateral membrane to the vascular bed. Zinc absorption is dependent on the

concentration ofbioavailable zinc in the lumen. Although 30% of zinc is absorbed

from typical diets, greater efficiency is achieved in animals fed zinc-deficient diets, and

lesser efficiency in animals fed diets containing supplemental zinc (Hempe and Cousins

1992, Sandstrom 1989, Smith et al. 1978). In vitro transport studies using brush border

membrane vesicles from rats showed that BBM transport was greater from zinc-deficient

rats compared to zinc-adequate rats (Menard and Cousins 1983). Dietary zinc intake did

not affect transport into basolateral membrane vesicles, however, suggesting that control

of zinc absorption occurs at the apical, but not the basolateral membrane (Oestreicher and

Cousins 1989).

The last area for control of zinc absorption is the enterocyte cytosol itself. The

cytosolic protein metallothionein has been the focus of research in this area because it is

induced by high dietary zinc and parenteral zinc administration, but is depressed at low

zinc intakes (Menard and Cousins 1983, Smith et al. 1978). One model of zinc absorption

suggests that intestinal metallothionein is an integral part of the regulatory machinery,








acting as a damping agent during periods of excessively high dietary zinc intakes

(Cousins 1989, Hoadley et al. 1988, Richards and Cousins 1975). In this model a high

zinc influx into the mucosal cell induces metallothionein production. Metallothionein

then chelates the cytosolic zinc, limiting its passage from the enterocyte to the portal

circulation. Enterocyte to lumen efflux of zinc, combined with regular sloughing of

enterocytes from the villus tip result in reduced zinc absorption. This model is consistent

with results from animal studies wherein high zinc diets or parenteral zinc administration

elevated metallothionein levels in the intestine and resulted in decreased zinc absorption

from subsequent meals (Coppen and Davies 1987, Hoadley et al. 1988, Richards and

Cousins 1975). Similarly dietary zinc restriction depresses tissue metallothionein content

and results in enhanced zinc absorption from subsequent meals (Hoadley et al. 1987,

Smith and Cousins 1980).

Contrary to results with diet-related changes in metallothionein, several inducers

of intestinal metallothionein do not inhibit zinc absorption. For instance, bacterial

infection, bacterial lipopolysaccharide injection, and interleukin-1 administration each

increased zinc absorption, even though each induces intestinal metallothionein (Pekarek

and Evans 1975, Pekarek and Evans 1976). Further, mouse studies showed no clear

correlation between metallothionein level of the intestine and zinc absorption using either

oral dosing, stomach tube feeding, in situ duodenal loop feeding, or injection of a zinc

solution into the duodenum in situ (Flanagan et al. 1983, Olafson 1983, Starcher et al.

1980). Other studies have shown no increase in intestinal zinc retention even when

metallothionein is induced (Hempe et al. 1991). Thus the role of metallothionein in

regulation or zinc absorption is still in question.








Metallothionein also is believed to affect zinc accumulation in tissues.

Metallothionein gene expression is regulated in a tissue-specific manner in rats (Blalock

et al. 1988). Hepatic metallothionein induction occurs after excessive dietary zinc intakes

in rats, and is associated with hepatic zinc accumulation (McCormick et al. 1981). Other

treatments that induce hepatic metallothionein, such as administration of interleukin-1,

interleukin-6, and lipopolysaccharide, result in hepatic zinc accumulation and serum zinc

depression (De et al. 1990, Huber and Cousins 1993). Taken together, these data suggest

that metallothionein expression exerts a strong influence on tissue zinc distribution.

Within cells, metallothionein also may function as a zinc reservoir from which

this metal is made available for incorporation into apometalloenzymes or other

metalloproteins. This idea is supported by the highly rapid exchange rates of zinc in

metallothionein, which are far faster than the exchange rates from other proteins (Li et al.

1980, Udom and Brady 1980). Further, incubation with Zn-metallothionein reconstitutes

a number of enzymes and transcription factors (e.g., apocarbonic anhydrase and Spl) and

rescues their activities (Li et al. 1980, Udom and Brady 1980, Zeng et al. 1991).

Exchange rates are more rapid and more zinc is exchanged in the presence of oxidized

glutathione, possibly linking zinc release to cellular redox status (Jiang et al. 1998).

Interestingly, zinc is also liberated from metallothionein by a number of oxidants

(Berendji et al. 1997, Fliss and Menard 1992).

We undertook research concerning the role of metallothionein in regulation of

zinc absorption due to the importance of zinc for health. Proper zinc nutrition is

particularly important to support growth and immune functions (reviewed in Keen and

Gershwin 1990, Rivera et al. 1995). In particular, zinc deficiency is known to retard








growth and inhibit sexual maturity, both of which can be reversed to some degree with

zinc supplementation (as reviewed in Prasad 1991). Further, supplemental zinc can help

to alleviate secretary diarrhea and morbidity in third world countries (Sazawal et al.

1995). Consequently, determining the mechanisms) and regulation of zinc absorption

provides information that can be used to efficiently and effectively maintain proper zinc

homeostasis, and to support health.


Metallothionein and Zinc in Defense Against Oxidative Stress

The pathogenesis of aging, cancer, atherosclerosis, cataracts, neurodegenerative

disorders, and ischemia-reperfusion injury are associated with oxidative stress (Blot et al.

1993, Weidau-Pazos et al. 1996, Fraga et al. 1990, Rengstrom et al. 1992, Rosen et al.

1993, Taylor et al. 1992). Oxidative stress occurs when the balance between oxidative

attacks and oxidative defense systems favors oxidation. The mediators of oxidative

damage include reactive oxygen species (i.e., H202, 02 and "OH), and radical species that

are not oxygen-centered (e.g., -CC13). These species can cause damage to lipids, proteins,

and nucleic acids (Farber 1994, Loft et al. 1994, Oliver et al. 1990). Living organisms

combat oxidative stress through the use of antioxidant nutrients such as tocopherols and

ascorbate, as well as endogenous antioxidant scavengers like superoxide dismutase and

glutathione (Yu 1994). When damage does occur, organisms have damage repair systems

to fall back on. For example, DNA excision repair systems remove oxidized DNA bases,

and glutathione peroxidase can convert lipid hydroperoxides formed by membrane

oxidation to less reactive lipid hydroxides. Further, proteolytic and lipolytic enzymes

degrade damaged macromolecules when damage is irreversible. But when these systems

are overcome, the cell or organism may not survive. Since the generation of reactive








oxygen species is unavoidable, maximizing antioxidant defense systems has become a

research priority (Cohen 1994).

Adequate zinc nutrition may help protect against oxidative stress. Zinc-deficient

rodents display markers of oxidative damage, and are more susceptible to subsequent

oxidative stresses (DiSilvestro and Carlson 1993, Miceli et al. 1999, Oteiza and Keen

1995). These results may reflect depressed activity of Cu/Zn superoxide dismutase,

increased oxidation of sulfydryl groups that are normally protected by zinc binding, or

increased exposure of zinc binding sites within proteins to copper- and iron-induced

oxidation (as reviewed in Powell 2000). Increased oxidative stress also might be due to

perturbation of a number of other zinc-dependent processes, including maintenance of

cellular metallothionein (Blalock et al. 1988, Schroeder and Cousins 1990). This is due to

reduced activation of metallothionein gene expression, as well as enhanced susceptibility

of metallothionein protein to proteolysis when zinc is not available for binding (Smith et

al. 1978, Feldman and Cousins 1976). Metallothionein expression may help defend

against oxidative stress since metallothionein is capable of scavenging free radicals, and

the hydroxyl radical in particular (Thomalley and Vasak 1985).

Beyond preventing zinc deficiency, supplemental zinc provides additional

protection against certain oxidative stresses (Blain et al. 1998, Dhawan and Goel 1995).

Although the mechanism of protection is uncertain, it may include the induction of

metallothionein protein. Metallothionein is induced by a number of chemicals that

generate oxidative stress (Bauman et al. 1991, Satoh et al. 1996, Shiraishi et al. 1989;

Tate et al. 1995). Preinduction of metallothionein by a number of agents (including zinc,

other metals, hormones, and cytokines) is associated with protection against the toxicity








of subsequent metal, chemical, and other stresses in cell culture and in vivo (e.g., Coppen

et al. 1988, Moffat et al. 1996, Satoh et al. 1992, Schroeder and Cousins 1990). Similar

results were seen when cells were transfected with metallothionein genes (Kaina et al.

1990, Schwarz et al 1995). Other reports showed no protection by metallothionein

expression against free radicals, however, casting doubt on metallothionein's role in

oxidative defense (Kaina et al. 1990, Kelley et al. 1988).

Based on these observations, supplemental zinc and metallothionein expression

may or may not affect the outcome of conditions characterized by oxidative stress

(Oteiza et al. 1995, Prasad 1991). Characterization of zinc and metallothionein as

participants in oxidative defense is important because oxidative damage is associated

with so many disease processes. For example, oxidative stress is a component of

inflammatory bowel disease, which is associated with a reduced metallothionein content

of the bowel (Mulder et al. 1991). Also, free radical production and oxidative stress

characterize many diseases of the liver, and the liver is an organ that contains high zinc

and metallothionein concentrations after zinc supplementation (Cohen 1994).


Hypotheses and Research Objectives

Based on the information summarized above, we developed two hypotheses

regarding the biochemical actions and interactions of zinc and metallothionein:

1. Zinc absorption is inversely related to intestinal metallothionein production,

while tissue distribution of absorbed zinc is directly related to the

metallothionein content of the tissue.

2. Zinc and/or metallothionein protect mouse livers and hepatocyte cultures from

oxidative damage.








We examined these hypotheses using relatively new models for biological

research transgenic mice. We used metallothionein transgenic overexpressing mice

(designated TG), metallothionein knockout mice (i.e., null; designated KO), and their

respective control strains in all of the studies that follow. Metallothionein transgenic mice

carry 56 copies of the metallothionein-1 gene in their genome, and those genes are

responsive to the same stimuli that induce endogenous metallothionein-1 genes (Palmiter

et al. 1993). In contrast, metallothionein knockout mice produce no metallothionein-1 or

-2 protein under any conditions in the tissues we studied (Masters et al. 1994A). This

allows us to assess the effects of a broad spectrum of metallothionein expression levels

on zinc metabolism and oxidative stress. Further, studies with metallothionein null mice

allow determination of biological effects of zinc that are independent of metallothionein.














CHAPTER 2
THE EFFECT OF METALLOTHIONEIN EXPRESSION ON ZINC ABSORPTION IN
METALLOTHIONEIN TRANSGENIC AND METALLOTHIONEIN KNOCKOUT
MICE

Introduction

The mechanisms that regulate zinc metabolism are not understood. When dietary

zinc intake is restricted in experimental animals and humans, the efficiency of zinc

absorption increases and endogenous zinc excretion decreases. Furthermore, zinc

absorption is depressed after ingestion of zinc-rich diets. The biomolecules that mediate

the regulation of zinc metabolism by the dietary zinc supply have not been fully

described. The cytosolic protein metallothionein (MT) may be a principal participant in

this regulation. This metalloprotein is inducible by many factors (stimuli) and may act as

a zinc pool or buffer that is influenced by body zinc levels. In addition, the redistribution

of endogenous zinc associated with stresses such as acute infection and physical trauma

may require metallothionein. The induction is believed to involve interleukin-1,

interleukin-6, and glucocorticoid hormone-mediated changes, all of which can be linked

to elevated expression of metallothionein in the liver and other tissues (reviewed in

Cousins 1989, 1996).

Metallothionein transgenic mice (TG) and metallothionein knockout mice (KO)

provide a unique model to study the effects of metallothionein expression on zinc

absorption. Metallothionein transgenic mice have elevated metallothionein protein in

many tissues, including the liver and intestine (Iszard et al. 1995, Liu and Klaassen









1996). The larger cytosolic metallothionein pools might convey protection against zinc

deficiency (Dalton et al. 1996). Conversely, metallothionein knockout mice allow

examination of how zinc metabolism differs when no metallothionein is produced

(Masters et al. 1994, Michalska and Choo 1993). Others have shown that KO mice have

altered zinc metabolism, including inability to sequester zinc in the liver after injections

of zinc or lipopolysaccharide (Coyle et al. 1995, Philcox et al. 1995). Further,

hepatocytes from KO mice were incapable of accumulating zinc in response to

interleukin-6 or dexamethasone treatment. Hence, significant perturbations of zinc

metabolism occur in mice with altered metallothionein expression. This study was

directed at determining the effects of altered metallothionein expression on zinc

absorption.

The intestine is a major control site for zinc homeostasis and is also a major

metallothionein-expressing organ (Cousins 1989). While some data support an inverse

relationship between intestinal metallothionein expression and zinc absorption (Hoadley

et al. 1987, Hoadley et al. 1988, Menard et al. 1981, Smith and Cousins 1980), other data

do not (Flanagan et al. 1983, Starcher et al. 1980). To examine how metallothionein

influences the intestinal processing of zinc, a zinc dose was delivered by gavage (zinc

tolerance test), and the increase in the serum zinc concentration was used as a measure of

absorption.

Materials and Methods

Animals

The founder mice used in this study were obtained from The Jackson Laboratory,

Bar Harbor, ME. The transgenic mice (designated TG mice) were derived from the








C57BL/6 strain crossed with the SJL strain (Palmiter et al. 1993). The metallothionein

knockout mice (designated KO mice) were derived from the 129/SvCPJ strain crossed

with C57BL/6 (Masters et al. 1994). Mice of the appropriate background strains served

as controls, designated CT and CK, respectively. Only adult male mice were used for

experiments. They were housed in plastic cages with wood shavings as bedding and with

a 12 h light/dark cycle. The mice had free access to a commercial diet (Laboratory

Rodent Diet No. 5001, PMI Feeds, New Albany, IN). All experiments were started

between 8:00 and 10:00 AM. Care and treatment of the mice received approval of the

University of Florida Institutional Animal Care and Use Committee.


Radioisotopes

The 2a"P-dCTP was from Du Pont NEN (Boston, MA), and the 109Cd (1.35 x 106

kBq/nmol Cd) was from Isotope Product Laboratories (Burbank, CA).


RNA Isolation and Northern Analysis

Total RNA was isolated from intestine and liver using TRIzol reagent (Life

Technologies, Gaithersburg, MD). Briefly, 50 to 100 mg of the proximal duodenum and

the liver were homogenized in 2 mL of TRIzol reagent. After addition of chloroform, the

RNA was processed and analyzed as described previously (Blanchard and Cousins 1996).

Equal quantities of RNA from mice of each group were pooled and subjected to Northern

analysis (20 pjg total RNA per lane). Equal loading was confirmed by ethidium bromide

staining. Northern blot analyses were carried out using a rat metallothionein-1 cDNA

probe (Blanchard and Cousins 1996). It was radiolabeled with cc3P-dCTP using the RTS

RadPrime DNA Labeling System (Life Technologies) as described previously








(Blanchard and Cousins 1996). The metallothionein cDNA probe hybridizes to

metallothionein mRNA from the control mice and the KO mice. The metallothionein

mRNA of the KO mice contains a premature stop codon, however, and can not be

translated into MT protein.


High Performance Liquid Chromatography of Intestinal Cytosol

In some experiments, the mucosa was homogenized with a Potter Elvejhem tissue

grinder using 2 volumes of ice-cold buffer (S-12 buffer, 154 mmol/L NaCI, 10 mmol/L

TrisCi, 3 mmol/L NaN3, and 10 mmol/L MgSO4) plus protease inhibitors (0.1 mmol/L

phenylmethylsulfonyl fluoride, 1.2 tmol/L leupeptin and 1.5 pmol/L pepstatin A). After

centrifugation at 100,000 x g (30 min), the cytosol fraction was filtered (0.22 ptm) and

200 ptL was applied to two Superdex 75 chromatography columns (1 x 30 cm; Pharmacia

Biotech, Piscataway, NJ) in series using an isocratic elution of S-12 buffer (Hempe and

Cousins 1989). Metallothionein was measured in all fractions using the cadmium ('9Cd)-

binding assay (Eaton and Toal 1982).


Oral Zinc Dosing

Mice that had been fasted overnight were administered 0.5 mmol Zn/kg body

weight as ZnSO4 in saline, or saline alone via a stomach tube and were killed 2 h later.

Blood was obtained by cardiac puncture and serum was prepared for serum zinc analysis.

Liver and intestinal zinc concentrations were measured to determine if the zinc dose

produced a change in tissue uptake/retention.








Analytical Methods and Statistical Analysis

'9Cd was measured using a Packard Cobra II gamma spectrometer equipped with

a 3 inch crystal (Packard, Downers Grove, IL). Metallothionein protein was measured by

the cadmium (1"9Cd) binding assay (Eaton and Toal 1982). Briefly, tissue extracts are

boiled, and the resulting supernatant is incubated with "9Cd. After removal of unbound

'"9Cd using hemoglobin as a chelator, '9Cd bound to MT is measured by y-counting, and

converted to moles of MT using the Cd-MT binding stoichiometry of 7:1. Total protein

was measured by the method of Lowry et al. (1951). Serum zinc concentrations were

measured by flame atomic absorption spectrophotometry (AAS) (Hempe and Cousins

1989). Tissue zinc was also measured by AAS, after sections of liver and intestine were

digested with acid (HNO3/H2SO4; 3/1) as described previously (Dunn and Cousins 1989).

Data were analyzed by ANOVA followed by the Student-Newman-Keuls multiple

comparison test where appropriate (InStat, GraphPad, San Diego, CA). Logarithmic

transformation of some data was used to obtain homogenous variances before analysis.


Results

Metallothionein Expression in Intestine and Liver after Oral Dosing

Northern analysis of metallothionein mRNA in liver and intestine demonstrates

both basal and zinc-induced levels of expression in these genotypes (Fig. 2-1). All groups

expressed metallothionein mRNA, but TG mice had several-fold greater metallothionein

expression than either of the control strains in both the intestine and liver.Metallothionein

mRNA was also expressed in KO mice, but this message contains a premature stop

codon and is not translatable. Zinc treatment resulted in induction of metallothionein

mRNA in the intestine and liver of all mice, but expression was greatest by far in the TG








mice. These data confirm that KO mice, control mice (CK and CT), and TG mice

represent groups with distinguishably different basal and zinc-induced metallothionein

mRNA levels in both intestine and liver. Consequently, if metallothionein is important in

regulating zinc absorption, these groups should display different absorption

characteristics.

The knockout mutation was confirmed by identifying metallothionein protein in

size exclusion chromatography fractions of intestinal cytosol (Fig. 2-2). Metallothionein

protein was measured in each 0.5 mL fraction using the '9Cd binding assay (Eaton and

Toal 1982). A large metallothionein peak is seen between 32 and 35 mL in the

chromatography profile from TG mice, but none is identifiable in the profile from the KO

mice. No peak was seen in profiles from zinc treated KO mice (data not shown).

Therefore, although metallothionein mRNA is produced by KO mice, no metallothionein


+ + + + Zinc

Intestine





Liver


CK KO CT TG

Figure 2-1. Intestinal metallothionein mRNA in metallothionein knockout (KO),
knockout control (CK), transgenic (TG), and transgenic control (CT) mice that
consumed diets containing either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d. Equal
amounts of total RNA were pooled from intestine of 3-5 mice per group and analyzed
by northern analysis using a metallothionein-1 cDNA probe, with P-Actin used for
normalization. (A) KO and CK mouse intestine. Although metallothionein-1 mRNA is
present in KO mice, it is not translatable. (B) CT and TG mouse intestine. (C)
Graphical representation of metallothionein-1 mRNA expression as a function of








protein results from that message.Although metallothionein is the predominant '09Cd

binding compound in the cytosol, a small but finite amount of 9Cd binding is seen in

nearly all other fractions in the profiles of TG and KO mice (Fig. 2-2). Others have

observed the same phenomenon (Liu et al. 1996A). Since no "Cd binding activity is


VI

-4->






e0


1.2



1



0.8



0.6



0.4



0.2


18 20 22 24 26 28 30 32 34 36 38 40
Elution Volume (mL)

Figure 2-2. Superdex 75 size exclusion chromatography of metallothionein (MT) in
cytosol from intestine mucosa of metallothionein-transgenic (TG) and metallothionein-
knockout (KO) mice 2 h after an oral dose of saline. Mucosal cytosol was separated by
two superdex 75 columns run in series and 0.5 mL fractions were collected. The MT
content of fractions 18 through 40 is expressed as tg MT equivalents. Metallothionein
elutes between 32 and 35 mL.








seen in any chromatographic fractions in the KO mice, including the fractions that

contain metallothionein in normal mice, any '9Cd binding activity associated with KO

mice cytosol should be considered background. These levels do not increase when zinc

treatment is given (data not shown). To better define the metallothionein content of

tissues we deducted the average '9Cd binding value associated with the cytosol of KO

mice from all groups when measuring metallothionein in the intestinal mucosa (0.1 mg/g

protein; 15 nmol/g protein) and liver (10 p.g/g liver; 1.5 nmol/g liver). It is clear that the

saline treated controls have little metallothionein protein present in intestinal mucosa

(Fig. 2-3). TG mice, however, have significantly elevated metallothionein levels

compared to CT mice. Zinc treatment significantly increased metallothionein in all but



1.2
E] Saline L Saline C
c 1-0 Zinc I Zinc
0.8
C-
tm 0.6
I--
2 0.4
0) b
E0 b b
a a a a
0.0-
CK KO CT TG

Figure 2-3. Metallothionein content of intestine mucosa of metallothionein-transgenic
and metallothionein-knockout mice after an oral zinc dose. Metallothionein (MT) was
measured by the '9Cd binding assay (mg MT/g mucosa protein) in the intestine mucosa
of metallothionein-transgenic (TG), TG control (CT), metallothionein-knockout (CK),
and KO control (CK) mice after an oral dose of either saline (D) or saline and 0.5 mmol
Zn/kg bw (0). Data are reported as mean +/- SEM of 4-5 mice/group. A constant value
of 0.1 mg MT/g mucosa protein was deducted from all measurements to account for the
nonspecific '"Cd binding observed in KO mouse mucosa. Data was analyzed by
ANOVA followed by the Student-Newman-Keuls multiple comparison post test. Bars
labeled with different letters within a graph are statistically different from each other (p
< 0.05)








the KO mice group. Again, the induction was greatest in TG mice, being eight fold

higher than the zinc treated CT mice. Overall metallothionein protein was not present in

KO mice, was only present in CT and CK mice after zinc treatment, and was always

greatest in TG mice.

Metallothionein was also measured in liver (Fig. 2-4). Unlike in the intestine,

liver metallothionein in saline treated controls was greater than in KO mice, and similar

to TG mice. Zinc treatment elevated the mean value of metallothionein 5-fold in CT (p >

0.05) and 10-fold in TG mice livers, but the increase was only significant in TG mice. No

increase in liver metallothionein was seen in CK mice after zinc treatment. This was not

completely unexpected, however, since differences in zinc induction of liver

180
150 Saline L Saline b
150 U Zinc Zinc
>. 120
0)
E go-
Pa,b
S60-

30 b b a a
a a a a1

CKO KO CTG TG

Figure 2-4. Metallothionein content of liver in metallothionein-transgenic and
metallothionein-knockout mice after an oral zinc dose. Metallothionein (MT) was
measured by the '"Cd binding assay (jig MT/g liver) in the liver of metallothionein-
transgenic (TG), TG control (CT), metallothionein-knockout (KO), and KO control
(CK) mice 2 h after an oral dose of either saline (E) or saline and 0.5 mmol Zn/kg bw
(M). Data are reported as mean +/- SEM of 4-5 mice/group. A value of 10 gg MT/g
liver was deducted from all measurements to account for the nonspecific 'Cd binding
observed in KO mouse liver. TG and CT data was log10 transformed prior to statistical
analysis to achieve homogeneous variances. Data was analyzed by ANOVA followed
by the Student-Newman-Keuls multiple comparison post test. Bars labeled with
different letters within a graph are statistically different from each other (p < 0.05).








metallothionein have been seen among mouse strains (Farr and Hunt 1989). Hence, each

group had a different amount of metallothionein protein present in intestine and liver

after zinc treatment. Since metallothionein may provide a zinc storage pool in these

organs, mice with the greatest metallothionein production (TG mice) have a greater

capacity to deal with a zinc load than those with the least metallothionein (KO mice).


Intestine, Liver and Serum Zinc Responses Two Hours after the Oral Zinc Dose

Although metallothionein in intestine and liver differed among groups, these

differences were not correlated to detectable differences in intestine and liver zinc

concentrations (data not shown). When saline was given, all mice had roughly equivalent

zinc concentrations in both intestine and liver (approximately 0.48-0.62 pmol Zn/g and

0.37-0.59 pmol Zn/g, respectively). No significant elevation of liver zinc was seen after

zinc treatment. All groups had significantly elevated zinc concentrations in intestine after

zinc treatment, but there were no significant differences among the zinc-treated groups

except in the KO group. In the KO mice, zinc treatment increased the intestinal zinc

concentration significantly compared to the zinc-treated CK mice (1.37 0.22 vs. 0.86

0.11 pmol Zn/g, respectively; p < 0.05). In contrast, the intestinal zinc concentrations in

zinc-treated CT and TG mice were similar (0.97 + 0.18 and 1.03 0.18 umol Zn/g,

respectively). Hence, absence of metallothionein resulted in a detectable increase in zinc

accumulation in intestine. However, overexpression in the TG mice did not influence

intestinal zinc retention. This suggests that metallothionein does not alter zinc

metabolism simply by sequestering zinc in the intestine. The change in serum zinc

concentration two hours after the oral zinc dose was used as an indicator of the quantitiy

of zinc absorbed (Fig. 2-5). In contrast to tissue zinc, serum zinc was markedly affected








by metallothionein expression. Although all groups had similar serum zinc

concentrations when given saline (15-30 gmol/L), mice with greater metallothionein

expression had lower concentrations after zinc treatment than mice with less

metallothionein expression. Zinc treated control strains had serum zinc concentrations 4

to 5 times higher than saline treated controls. KO mice, however, had 10-fold greater

serum zinc values after zinc treatment. Conversely, TG mice had only 2.3-fold greater

serum zinc concentrations after zinc treatment. Thus, serum zinc concentrations were

inversely proportional to the amount of intestinal metallothionein expressed. This

relationship is illustrated in Fig. 2-6.


Discussion

Our hypothesis was that intestinal metallothionein acts as a negative regulator of

zinc absorption. This relationship has been examined in the past, but with conflicting

300
U Saline b U Saline
250 U Zinc U Zinc

E 200

150 b

E 100 -_
*- a
50 a a a a


CK KO CT TG

Figure 2-5. Serum zinc in metallothionein-transgenic and metallothionein-knockout
mice after an oral zinc dose. Zinc was measured by atomic absorption
spectrophotometry (utmol zinc/L) in serum of metallothionein-transgenic (TG), TG
control (CT), metallothionein-knockout (KO), and KO control (CK) mice 2 h after and
oral dose of either saline (0) or saline and 0.5 mmol Zn/kg bw (M). Data are reported
as mean +/- SEM of 4 mice/group. Data were analyzed by ANOVA followed by the
Student-Newman-Keuls multiple comparison post test. Bars labeled with different
letters within a graph are statistically different from each other (p < 0.05).









results. For instance, Dr. Cousins' laboratory (Hoadley et al. 1987, Menard et al. 1981,

Smith and Cousins 1980, Smith et al. 1978) found that the quantity of zinc absorbed by

the isolated perfused rat intestine was inversely related to the zinc content of the diet

consumed prior to the experiments. In addition, giving rats a large zinc dose (i.p.) 18 h

prior to experiments depressed zinc absorption. Since zinc absorption was inversely

proportional to intestinal metallothionein throughout those experiments, it was proposed

that metallothionein serves as a damper of zinc absorption. Similarly, studies where rats

consumed diets ranging from 5 to 80 mg Zn/kg also showed that zinc absorption was

inversely related to metallothionein-bound zinc (Coppen and Davies 1987). Furthermore,

Hoadley et al. (1988) found that elevated metallothionein levels in intestines of fasted

rats were associated with greater mucosa to lumen transfer of absorbed zinc by the

isolated perfused rat intestine. They proposed that metallothionein depresses zinc

absorption by providing a sink that holds zinc in the intestine, allowing more opportunity

for transfer of zinc back into the lumen.

Contrary to the results cited above, other studies found no relationship or positive

correlation between intestinal metallothionein and zinc absorption. For example, bacterial

infection, endotoxemia, and interleukin-1 administration to rats all elevated liver

metallothionein, and resulted in 50-100% greater zinc absorption and liver zinc

accumulation from 65Zn doses (Kincaid et al. 1976, Pekarek and Evans 1975, 1976).

Intestinal metallothionein expression was not evaluated in those experiments, but

endotoxin has been shown to induce both metallothionein-1 and metallothionein-2 in

mouse intestine (De et al. 1990). Also, interleukin-1 is thought to be a mediator of

metallothionein induction by LPS in some tissues, and thus may induce the protein in the









intestine as well (reviewed in Cousins 1996). Furthermore, small zinc doses (0.2 pmol/kg

body weight; i.p.) caused metallothionein induction in the mouse intestine and

corresponded to enhanced zinc absorption 18 h later (Starcher et al. 1980). Flanagan and

coworkers (1983) observed no difference in zinc uptake or absorption in relation to

intestinal metallothionein in intestinal perfusion experiments with mice. They did,

however, see greater zinc absorption in zinc-deficient vs. control mice when doses of


35O
X KO
300 CK

A CT
3 250
_. TG
0
E
2 200
0
E
0
.C 150
N
E
| 100 A
0 A
A\

50-


0 0.2 0.4 0.6 0.8 1 1.2

Intestine Metallothionein (mg/g protein)

Figure 2-6. Serum zinc concentration as a function of intestine metallothionein content
in zinc-treated mice. Response of serum zinc (Y) vs. intestinal metallothionein content
(X) in metallothionein-transgenic (TG), TG control (CT), metallothionein-knockout
(KO), and KO control (CK) mice 2 h after gavage with 0.5 mmol Zn/kg bw.
Metallothionein (MT) is expressed as mg MT/g mucosa protein and serum zinc as
limol Zn/L. The inverse relationship between intestinal MT and serum zinc can be
expressed mathematically with good fit (Y = 1690/X + 0.75; r2 = 0.94).








zinc were delivered by gavage. They also demonstrated that differences exist in zinc

absorption characteristics between rats and mice, particularly the response of increased

absorption during zinc deficiency.

Although the studies cited above focused on the effect of metallothionein on zinc

absorption, the methods used to alter intestinal metallothionein levels varied. Treatments

used to induce the protein included intraperitoneal, intragastric, and dietary doses of zinc,

fasting, bacterial infection, lipopolysaccharide and interleukin-1 administration, and

various forms of physical stress. Although these treatments manipulate metallothionein

expression, each has effects not related to this protein that may cause physiological

changes and complicate interpretation of the results. Using knockout and transgenic

mouse models, it is possible to focus on zinc absorption as directly related to

metallothionein expression.

Giving animals a large oral dose of zinc by gavage, we were able to determine the

effects of metallothionein induction on zinc absorption by measuring serum and tissue

zinc concentrations. This avoids the potential for isotope dilution, which can cloud

interpretation ofradioisotopic tracer studies using 65Zn. We have used the oral dosing

approach previously (Menard et al. 1981). It is equivalent to the zinc tolerance test used

with humans (Sullivan et al. 1979). We used fasting and dosing in saline to prevent

nonspecific binding of zinc to food in the gut, and to allow for a maximal gastric

emptying rate. The 2 h time point used was determined to be the time point of maximal

serum zinc response in these mouse strains (data not shown), and agrees with data from

rats (Menard et al. 1981) and humans (Sullivan et al. 1979, Valberg et al. 1985). Further,

all dosing was done between 8 AM and 10 AM. The 0.5 mmol/kg dose given is 2.5 to 3.1








times greater than the typical dietary zinc intake of these mice (0.17 to 0.22 mmol/kg

body weight). Although greater than the typical intake, this dose is attainable through the

diet, and is therefore nutritionally relevant. Further, since the dose produced fivefold to

tenfold increases in serum zinc, we anticipate that smaller doses will also result in

significant differences, albeit smaller. Menard et al. (1981) showed that intestinal

metallothionein synthesis in rats was increased by 3 h after an oral zinc dose is given.

Furthermore, a second dose of zinc caused induction of metallothionein synthesis and

resulted in better regulation of serum zinc concentrations. In the present experiments,

serum zinc doubled in TG mice and increased 10-fold in KO mice when zinc was

delivered by gavage. We interpret the inability of the KO mice to handle the zinc load

compared to the TG mice to the difference in metallothionein expression. Specifically,

the TG mice controlled serum zinc concentrations more tightly than did the KO mice.

Serum zinc concentrations remain elevated for a considerable time after the oral dose.

We cannot rule out that the observed differences wee related to different kinetics of

absorption in these genotypes, however.

A drawback of this approach is that the role of other tissues in clearance of zinc

from the circulation cannot be explained adequately. Since the gene addition in TG mice

and gene deletion in KO mice are not tissue specific, we cannot rule out the possibility

that MT expression in some other tissue affected zinc clearance from the serum.

However, we did measure the zinc content of the liver, the main zinc storage organ and

the key organ in the regulation of zinc metabolism (Cousins 1996, Coyle et al. 1995). In

this study, no change in liver zinc was detected between saline or zinc-treated mice, and

no difference was seen among groups of zinc-treated mice. This is in agreement with data








collected in rats, where hepatic accumulation of gavaged zinc was not observed until nine

hours after dosing (McCormick et al. 1981). Since the liver can act rapidly to regulate

zinc metabolism, yet did not show an elevation in zinc concentration, it is unlikely that an

organ with less influence on zinc metabolism caused the differences seen in serum zinc.

In the time that has elapsed since the publication of the results found above

(Davis et al. 1998), several reports have been forwarded regarding the role of

metallothionein in zinc absorption. A number of these studies used a separate strain of

metallothionein knockout mice (Michalska and Choo 1993). Our findings regarding the

serum zinc response to the oral zinc dose was confirmed in these studies, i.e., knockout

mice had greater serum zinc levels than control mice over a range of oral doses that

began two orders of magnitude lower than in our studies (Coyle et al. 1999). Further,

their studies confirmed greater zinc accumulation in duodenum and jejunum of KO mice

than CK mice. These researchers also found lower accumulation of a radiolabelled dose

in tissues other than the intestine (liver, skin, muscle, kidneys and pancreas) at 4 h post

dose, which they interpreted as reduced absorption. However, KO mice lack the ability to

sequester zinc in several organs (as determined by the researchers mentioned above, as

well as by us in chapters 3, 4, and 5 of this dissertation), including the liver, which

renders measurement of tissue zinc accumulation unsuitable as an index of intestinal zinc

absorption in this model.

Within intestinal cells, higher intakes of zinc may be processed via a mechanism

that involves metallothionein. Since elevated intestinal metallothionein levels were not

associated with greater intestinal zinc accumulation, metallothionein does not seem to act

simply as a zinc sequestrant. Metallothionein may act as a zinc pool from which zinc is






27


highly available for transport back to the lumen, as suggested by Hoadley et al. (1988).

Without metallothionein, the KO mice may be unable to maintain a satisfactory

mucosa-to-lumen zinc flux. This might explain why KO mice have elevated zinc levels in

serum and intestine. Since metallothionein mRNA levels are also induced by zinc in

humans (Sullivan and Cousins 1997), we expect that metallothionein would affect zinc

absorption by the human intestine as well.














CHAPTER 3
REGULATION OF METALLOTHIONEIN EXPRESSION AND ZINC METABOLISM
BY DIETARY ZINC IN METALLOTHIONEIN TRANSGENIC AND
METALLOTHIONEIN KNOCKOUT MICE

Introduction

Homeostatic regulation of zinc metabolism by the dietary zinc supply is believed

to involve the protein metallothionein (as reviewed in Davis and Cousins 2000).

Metallothioneins are cysteine-rich, low molecular weight (6-7 kDa) metal binding

proteins that can bind up to seven atoms of zinc per molecule of protein (as reviewed in

Dunn et al. 1987). Metallothioneins are thought to provide a cellular zinc binding pool

that may be influenced by body zinc levels (as reviewed in Davis and Cousins 2000).

Redistribution of endogenous zinc is associated with stresses such as acute infection and

physical trauma. Redistribution is believed to involve IL-1, IL-6, and glucocorticoid

hormone mediated changes in zinc metabolism, all of which are linked to elevated

expression of metallothionein in the liver and other tissues (reviewed in Cousins 1989

and 1996). Metallothionein may also be involved in regulation of dietary zinc absorption

in the intestine, and affect accumulation of dietary zinc in the liver (reviewed in Davis

and Cousins 2000). Specifically, induction of intestinal metallothionein may reduce the

efficiency of zinc absorption during times of elevated zinc intake.

Previous studies firmly established the association of metallothionein induction

with cellular zinc accumulation and bodily zinc redistribution (reviewed in Davis and

Cousins 2000). Recent studies using metallothionein overexpressing transgenic mice (TG








mice) and metallothioein knockout mice (KO mice) have more directly analyzed the role

of metallothionein in zinc metabolism (Masters et al. 1994, Palmiter et al. 1993). Results

from these studies confirmed the role of metallothionein expression in protection against

severe zinc deficiency (Andrews and Geiser 1999, Dalton et al. 1996, Kelly et al. 1996).

Other studies confirmed the necessity ofmetallothionein expression for zinc

redistribution in response to immune stress, oxidative stresses and fasting (Davis et al.

submitted, Philcox et al. 2000, Rofe et al. 1996). Metallothionein's relationship to

suppression of zinc absorption is still under debate (Davis et al. 1998, Coyle et al. 1999).

Questions remain regarding metallothionein's role in tissue distribution of dietary

zinc. Previous studies from this lab found that metallothionein induction after acute

elevations in zinc intake was associated with accumulation of zinc in rat liver and

intestine. Whether metallothionein induction is responsible for this increase in hepatic

zinc, or whether it is only responding to the elevated hepatic zinc in not clear. Further,

the effect of metallothionein expression on zinc distribution in mice during chronic

exposure to elevated dietary zinc intakes is unknown. In these experiments we studied the

interrelationship of metallothionein expression and chronic exposure to a spectrum of

dietary zinc intake levels in metallothionein transgenic and metallothionein knockout

mice in order to monitor this mode of regulation more closely. The results suggest that

metallothionein expression affects tissue zinc accumulation only at highly elevated zinc

intakes. Metallothionein appeared to regulate its own expression, however. This might

have occurred through greater accumulation of cytosolic zinc. If so, metallothionein

expression may alter the expression of other zinc-regulated genes.








Materials and Methods

Animals

Metallothionein knockout and metallothionein transgenic mice used in this study

were derived from founder mice purchased from The Jackson Laboratory (Bar Harbor,

ME). The metallothionein overexpressing mice (designated TG mice) were generated in

C57BL/6 mice crossed with SJL mice (Palmiter et al. 1993), whereas metallothionein

knockout mice (designated KO) were generated in 129/SvCPJ mice (Masters et al. 1994).

C57BL/6 mice (designated CT) and 129S3/SvCPJ mice (designated CK) served as

controls, respectively. Only 7-9 week old male mice were used in these experiments.


Experimental Design

Mice were housed singly in stainless steel hanging cages with a 12 h light:dark

cycle. During experiments mice were given access to deionized water and semipurified

diet based on the AIN-76A formulation (AIN 77; Research Diets, New Brunswick, NJ)

with one of five zinc contents: 10, 50, 100, 200 and 300 mg zinc/kg diet (designated Zn0o,

ZnZn nloo, Zn2o0 and Zn30o, respectively). Initially, mice were given free access to the

Znl0 diet and deionized water for seven days. For seven days thereafter mice were given

free access to the Zn1o, Zn50, Zn100, Zn200o, or Zn300oo diet. All mice were killed between 8

AM and noon following completion of the dietary treatment. Care and treatment of the

mice received approval of the University of Florida Institutional Care and Use

Committee.

RNA Isolation and Northern Analysis

Total RNA was isolated from intestine and liver using TRIzol reagent (Life

Technologies, Gaithersburg, MD). Briefly, 50-100 mg of the proximal duodenum and








the liver were homogenized in 2 mL of TRIzol reagent. After addition of chloroform, the

RNA was processed and analyzed as described previously (Davis et al. 1998). Equal

quantities of RNA from mice of each group were pooled and subjected to northern

analysis (20 ptg total RNA per lane). Equal loading was confirmed by ethidium bromide

staining. Northern blot analyses were carried out using a rat metallothionein-1 cDNA

probe (Blanchard and Cousins 1996). This was radiolabeled with a12P-dCTP (Du Pont

NEN, Boston, MA) using the RTS RadPrime DNA Labeling System (Life Technologies)

as described previously (Blanchard and Cousins 1996). The metallothionein cDNA probe

hybridizes to both the normal metallothionein mRNA of the control and TG mice, and

the disrupted MT mRNA of the KO mice. Northern blots were also hybridized with a 3-

actin probe, and the 13-actin signal was used for normalization. Densitometry of the

autoradiographs was performed by scanning the film and measuring the relative intensity

using Intelligent Quantifier software (Bio Image, Ann Arbor, MI).


Analytical Methods and Statistical Analysis

Metallothionein was measured as described previously (Davis et al. 1998) using

the cadmium ('9Cd) binding assay (Eaton and Toal 1982). Total protein was measured

by the method of Lowry (1951). Serum zinc concentrations were measured by flame

atomic absorption spectrophotometry (AAS) (Hempe and Cousins 1989) after dilution

with deionized water. Tissue zinc was measured by AAS after sections of liver and

intestine were digested with HNO3/H2S04 (3/1) as described previously (Dunn and

Cousins 1989). Data were analyzed by two way ANOVA (2x5) with SigmaStat software

(Jandel Scientific, San Rafael, CA) to determine specific main effects and interactions

using genotype and dietary zinc as independent variables using. The Tukey-Kramer






32

multiple comparison test was used to determine significant differences between specific

groups (p < 0.05).


mg Zn/kg diet
MT- 1 mRNA

P-Actin mRNA

B

mg Zn/kg diet
MT- 1 mRNA

P-Actin mRNA


C


10 50 100 200 300


10 50 100 200 300


04004 *,4i0


10 50 100 200 300


30
<
I 25
* 20
CO.
^-15
Z
S10
E
H5


10 50 100 200 300


10 50 100 200 300
mg zinc/kg diet


Figure 3-1. Intestinal metallothionein mRNA in metallothionein knockout (KO),
knockout control (CK), transgenic (TG), and transgenic control (CT) mice that consumed
diets containing either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d. Equal amounts of total
RNA were pooled from intestine of 3-5 mice per group and analyzed by northern analysis
using a metallothionein-1 cDNA probe, with P3-Actin used for normalization. (A) KO and
CK mouse intestine. Although metallothionein-1 mRNA is present in KO mice, it is not
translatable and therefore does not give rise to metallothionein protein. (B) CT and TG
mouse intestine. (C) Graphical representation of metallothionein-1 mRNA expression as
a function of dietary zinc content.


DKO
*CK
OCT
OTG






I-y /









Results

Intestinal metallothionein mRNA was visibly upregulated by dietary zinc in CK

and CT mice fed the Zn0oo through Zn3oo diets (Fig. 3-1A,B). Dietary zinc induced

metallothionein mRNA more strongly and at lower zinc intakes in TG mice, but had little

effect on metallothionein mRNA in KO mice. The proportional increases in

metallothionein mRNA expression with dietary zinc intake were greater in mice with

greater metallothionein expression (TG > CT and CK > KO) (Fig. 3-1C). The

metallothionein mRNA expression pattern was similar in liver of KO, CK and CT mice

(Fig. 3-2A,B). Metallothionein mRNA expression in TG liver was very high at all dietary

zinc intakes, but relatively constant until the Zn300oo diet was consumed.

Intestinal metallothionein protein was measured in all genotypes (Fig. 3-3A,B).

The minimal values for metallothionein protein in KO mice (Fig. 3-3A) were shown to

be assay background only (Davis et al. 1998). Considering this, little metallothionein

expression was detected in CT or CK mice consuming the Zn10, Zn5o or Zn0oo diets (Fig.

3-3A,B). Expression of metallothionein protein was directly regulated by dietary zinc in

mice consuming the Zn200 and Zn300oo diets. Similar to the response of metallothionein

mRNA, intestinal metallothionein protein was induced at lower dietary zinc intakes and

to a greater extent in TG mice.

Intestine zinc concentrations were significantly affected by dietary zinc (p =

0.00008) in CT and TG mice. Intestine zinc contents appear to be associated with

intestinal metallothionein protein levels (Fig. 3-3A-D). Zinc content started to increase in

all mouse intestines at the same dietary zinc level that metallothionein induction









occurred. The increase in intestine zinc was greater in TG mice (except with the Zn300

diet), and no increase was seen in KO mice.

In contrast to the above results, the effects of dietary zinc on metallothionein

expression and zinc content in liver differed between the two control species (Fig. 3-4A-

D). Dietary zinc did not affect liver metallothionein levels in CK or KO livers (Fig. 3-

4A). Also, metallothionein levels in CK mice did not exceed the assay background seen

in KO mice. In contrast, TG mice had greater liver metallothionein values than CT mice

at all zinc intakes, but zinc did not induce metallothionein above basal levels until diets

containing 300 mg Zn/kg were consumed (Fig. 3-4B).

Liver zinc was not altered by dietary zinc or metallothionein expression except at

the highest dietary zinc intake (Fig. 3-4C,D). Liver zinc was not affected by dietary zinc

or genotype in CK and KO mice (Fig. 3-4C). Only the high dietary zinc concentration

(Zn300) elevated liver zinc, and only in TG mice (Fig. 3-4D).

A


mg Zn/kg diet

MT- 1 mRNA


13-Actin mRNA

B


CK
10 50 100 200 300

S, W3,
'^'^Mall


KO
10 50 100 200 300
................. ........ .. ."'


CT TG
mg Zn/kg diet 10 50 100 200 300 10 50 100 200 300

MT-I mRNA *gwm I

13-Actin mRNA
Figure 3-2. Liver metallothionein mRNA in metallothionein knockout (KO), knockout
control (CK), transgenic (TG), and transgenic control (CT) mice that consumed diets
containing either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d. mRNA was analyzed by
northern analysis as described in Fig. 3-1. (A) KO and CK mouse intestine. (B) CT
and TG mouse liver.








Serum zinc was affected by dietary zinc in all genotypes (Fig. 3-5A,B). Serum

zinc began to rise in CT and CK mice consuming Znjo0 and Zn200 diets, respectively, and


10 50 100 200 300


10 50 100 200 300


D" CT B
STG


b
CTr
bb


a, a
a
a a_ a
10 50 100 200 300


10 50 100 200 300


mg zinc/kg diet

Figure 3-3. Intestinal mucosal metallothionein protein and zinc in metallothionein
knockout (KO), knockout control (CK), transgenic (TG), and transgenic control (CT)
mice that consumed diets containing either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d.
Mucosal metallothionein in (A) CK and KO mouse intestine or (B) CT and TG mouse
intestine. Values found for KO mice reflect assay background only. Intestinal zinc in
(C) CK and KO mouse intestine or (D) CT and TG mouse intestine. Data are
presented as means SE of 3-6 mice/group. Statistical differences (p < 0.05) were
determined by two-way ANOVA followed by the Tukey-Kramer post test








continued to rise with increasing dietary zinc level. The effect of diet alone was very

significant (p = 0.0000005). No differences were seen between CK and KO mice. Serum

zinc of TG mice was slightly elevated comparedto CT mice consuming Zn10-Zn2oo diets.

When Zn300 diets were consumed, however, serum zinc was greater in CT mice than TG

mice.

Discussion

Metallothionein has long been implicated as a key biomolecule in regulation of

zinc homeostasis. Specifically, many lines of evidence point to metallothionein as a

component of the machinery involved in intracellular zinc accumulation (as reviewed in

Davis and Cousins, 2000). Here, we confirm that KO mice are unable to accumulate zinc

in the intestine in response to chronic exposure to elevated dietary zinc, as was found

previously in a different metallothionein KO mouse model (Tran et al. 1998). This report

extends this relationship to TG mice, which generally accumulated more intestinal zinc

than controls. These data show that zinc accumulation is affected by metallothionein

expression over a wide range of intestinal metallothionein contents. Further, our data

from TG mice show for the first time that the level of hepatic metallothionein expression

dictates the livers ability to accumulate zinc during chronic exposure to elevated dietary

zinc. This expands on the relationship previously seen after zinc injections in control and

KO mice, where KO mice were unable to accumulate zinc in the liver (Rofe et al. 1996).

If hepatic uptake of zinc is important during infection and trauma, metallothionein might

be crucial for processes that depend on this influx of zinc. Such a role has been suggested

for metallothionein expression in rat liver during regeneration and the acute phase

response (Arora et al. 1998, Dunn and Cousins 1989, Ohtake et al. 1978). A similar role








would be predicted in other metallothionein-expressing organs. As such, overexpression

of metallothionein may be beneficial to liver regeneration, or conditions associated with

the acute phase response.

We also discovered that induction of liver and intestine metallothionein mRNA

and protein occurred at lower dietary zinc intakes in animals with greater metallothionein


160-


SCK
DKO


0 r1 r.M r 20 F-
10 50 100 200 300


10 50 100 200 300 10
mg zinc/kg diet


Dl CT b
STG



a; a




10 50 100 200 300



L-CT D
STG b



a
aa aa a a aa

nomn


50 100 200


Figure 3-4. Liver metallothionein protein and zinc in metallothionein knockout (KO),
knockout control (CK), transgenic (TG), and transgenic control (CT) mice that consumed
diets containing either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d. Metallothionein in (A)
CK and KO mouse liver or (B) CT and TG mouse liver. Values found for KO mice
reflect assay background only. Zinc concentration in (C) CK and KO mouse liver or (D)
CT and TG mouse liver. Data are presented as means SE of 3-6 mice/group. Statistical
differences (p < 0.05) were determined by two-way ANOVA followed by the Tukey-
Kramer post test.









production (TG < CT and CK < KO; Fig. 3-1 & Fig. 3-2). Similar results were seen for

metallothionein protein (Fig. 3-3A,B & 3-4A,B). These data suggest that metallothionein

enhances its own expression, possibly by facilitating cellular zinc accumulation. In this

hypothesis, accumulation of cellular zinc in the metallothionein-zinc pool provides a

labile zinc pool that may interact with other pools (as illustrated in Davis and Cousins

2000). One such pool is the nucleus, where dietary zinc rapidly accumulates in a number

of organs (Cousins and Lee-Ambrose 1992). One of the biomolecules that the

metallothionein-Zn pool may interact with is the zinc-finger transcription factor MTF-1.

MTF-1 contains a zinc-binding site that may be sensitive to cellular zinc levels (as

reviewed in Andrews 2000). Zinc from the metallothionein-Zn pool may activate MTF-1,

resulting in subsequent transcriptional activation of the metallothionein gene through

numerous metal response elements (MREs) in the metallothionein gene promoter. This


40
35. CK A b DCT B
30 KO b H TG c
-ra,b,c b
25 a a a,b,c abc
a
20 a "
0 aa aa aab a
E15-
100

5

10 50 100 200 300 10 50 100 200 300
mg zinc/kg diet

Figure 3-5. Serum zinc in metallothionein knockout (KO), knockout control (CK),
transgenic (TG), and transgenic control (CT) mice that consumed diets containing
either 10, 50, 100, 200 or 300 mg Zn/kg for 7 d. Serum zinc in (A) CK and KO mice
or (B) CT and TG mice. Data are presented as means SE of 4-6 mice/group.
Statistical differences (p < 0.05) were determined by two-way ANOVA followed by
the Tukey-Kramer post test.








process would enhance production of metallothionein protein, which, after further zinc

binding, may fuel even greater metallothionein gene transcription. Cycling through this

pathway would amplify metallothionein levels, and as we observed, more amplification

would occur in mice with greater metallothionein expression.

The fact that metallothionein protein expression was elevated only with chronic

dietary zinc supplementation is in line with previous mouse experiments (Olafson 1983,

Tran et al. 1998). Tissue zinc levels were only elevated at high zinc intakes, also. Further,

KO mice maintained control over serum zinc, liver zinc and intestine zinc levels as well

as CK mice did. These results suggest that mice can maintain tight control of zinc

homeostasis over a large range of zinc intakes without metallothionein induction. Many

zinc transporter molecules have been identified recently, and it is likely that regulation of

the activity or expression of those proteins mediated zinc homeostasis over that range of

zinc intakes (reviewed in McMahon and Cousins 1998, Cousins and McMahon 2000).

The high correlation of zinc transporter 1 expression (Davis et al. 1998) and zinc

transporter-2 expression (Liuzzi et al. 2000) with metallothionein expression supports the

involvement of zinc transporters and metallothionein expression in zinc homeostasis.

Metallothionein appears to become an important mediator of cellular zinc homeostasis at

very high dietary zinc intakes, however (Davis et al. 1998). This is not to say that

metallothionein plays no role in zinc homeostasis at lower dietary zinc levels. It has been

shown that metallothionein expression provides partial protection against zinc deficiency

(Andrews and Geiser 1999, Dalton et al 1996, Kelly et al. 1996). It also is necessary for

redistribution of zinc during periods of stress (Cousins and Leinart 1988, Davis et al.

submitted, Dunn and Cousins 1989, Philcox et al. 1995). When considering the typical








range of dietary zinc intakes and in the absence of metallothionein-inducing stresses,

however, the effect of metallothionein expression on zinc homeostasis is minimal.

In conclusion, we found that metallothionein expression alters tissue zinc

accumulation, but only at highly supplemental zinc intakes. At lower intakes

metallothionein was not a factor. We found evidence that metallothionein expression

might act in a positive feedback loop to regulate its own expression, and that this occurrs

to a greater degree in mice with a greater number of MT gene copies. If this hypothesis is

true, metallothionein may also be involved in regulation of other zinc responsive genes.














CHAPTER 4
THE EFFECTS OF METALLOTHIONEIN GENE EXPRESSION AND
SUPPLEMENTAL DIETARY ZINC IN PROTECTION AGAINST
HEPATOTOXICITY IN METALLOTHIONEIN TRANSGENIC AND
METALLOTHIONEIN KNOCKOUT MICE

Introduction

Zinc and metallothionein are implicated in cellular defense against a number of

cytotoxic agents. There is evidence that supplemental zinc and overexpression of

metallothionein help protect cells and organisms from a number of stresses. For example,

administration of pharmacological zinc doses protects rodents from the toxicity of certain

metals and other chemicals, some of which cause oxidative stress (Blain et al. 1998,

Chvapil et al. 1973, Dhawan and Goel 1995, Powell et al. 1994). Similar results were

seen in cell culture (Coppen et al. 1988, Liu et al. 1991, Tate et al. 1999). The

mechanisms) through which zinc provides protection is uncertain. Zinc may protect

sulfhydryl groups from oxidation, may limit the redox reactive metal content of tissues,

or may elevate the activity of antioxidant enzymes (Coppen et al. 1988, Davis, C. D. et

al. 2000, Olin et al. 1995). Many believe that supplemental zinc provides antioxidant

protection through its powerful induction of metallothionein gene expression.

Preinduction of metallothionein by a number of metals (including zinc),

hormones, cytokines and other chemicals is associated with protection from the toxicity

of subsequent metal, chemical, and other stresses in cell culture and in vivo (Blain et al.

1998, Coppen et al. 1988, Kelley et al. 1988, Liu et al. 1991, Mello-Filho et al. 1988,

Moffat et al. 1996, Naganuma et al. 1985, Satoh et al. 1988, Schroeder and Cousins








1990). Several experiments in cell cultures transfected with metallothionein genes (Kaina

et al. 1990, Schwarz et al. 1995, Yao et al. 2000) and in cultures from metallothionein

transgenic and knockout mice (Lazo et al. 1995, Kondo et al. 1995, Wang et al. 1999,

Zheng et al. 1996) found similar results. Finally, a number of experiments in

metallothionein transgenic and metallothionein knockout mice came to similar

conclusion (Kang et al. 1997, Kang et al. 1999, Liu et al. 1995, Liu et al. 1998A, Liu et

al. 1999A, Masters et al. 1994, Michalska and Choo 1993, Rofe et al. 1998). The results

of several papers contradict the idea that metallothionein is universally protective,

however (DiSilvestro et al. 1996, Itoh et al. 1997, Liu et al. 1999A, Minami et al. 1999).

In some studies, metallothionein induction by zinc protected against the free

radical generating hepatotoxin carbon tetrachloride. These conclusions were reached

from in vitro studies, in vivo studies using nonspecific inducers of metallothionein, and

studies using pharmacological injections of zinc. While injections of zinc are known to

induce metallothionein and protect against oxidative stress in liver and cultured cells, it is

not clear whether supplemental dietary zinc mimics the protective effects of parenterally

administered zinc, and if so, whether the protection depends upon metallothionein

production. Conversely, it is not known whether the effect of metallothionein induction

on cytoprotection depends on the level of dietary zinc. Further, whether supplemental

zinc and metallothionein expression act additively or synergistically is uncertain.

Murine metallothionein knockout and metallothionein overexpressing models are

more direct models of the effects of metallothionein expression on cytotoxicity. Also,

zinc presented via the diet is more physiological than via injection. We determined

whether metallothionein overexpressing mice or metallothionein knockout mice had








altered sensitivity to carbon tetrachloride compared to identically treated controls. We

also determined whether supplemental dietary zinc reduced sensitivity to carbon

tetrachloride in these genotypes.

The results of the experiments explained herein confirm the importance of

metallothionein expression in protection against oxidative stress, but bring into question

the impact of supplemental zinc and/or elevated metallothionein expression in defense

against oxidative stress in vivo.


Materials and Methods


Animals

Metallothionein overexpressing mice (TG mice), C57BL/6 mice (CT mice),

metallothionein knockout mice (KO mice) and 129/SvCPJ mice (CK mice) were used in

these experiments. All experimental groups were age (8-11 wk) and sex matched. Mice

were housed singly in stainless steel hanging cages with a 12 h light:dark cycle. During

experiments, mice were given free access to deionized water and AIN-76A diets with

adequate or supplemental zinc content (10 or 500 mg zinc/kg diet, respectively; Research

Diets, New Brunswick, NJ). Care and treatment of the mice received approval of the

University of Florida Institutional Care and Use Committee.


Experimental Design

TG, KO and control mice were acclimated to diet containing 10 mg zinc/kg (Zn10)

and deionized water for 7 d. Following this acclimation period, one half of the mice were

switched to a diet containing 500 mg zinc/kg (Zns500) for 3 d while the other half remained

on the Zn10 diet. After the third day of dietary treatment, mice were injected with carbon








tetrachloride (CC4) in corn oil (20 tl/kg bw, i.p.) or corn oil alone between 8 AM and 10

AM, and animals were killed at 0 h, 12 h, 24 h, and 48 h post dose. Since mice in the 0 h

group did not receive injections, data at this time point represent the effects of diet and

genotype only.


Food Intake and Body Weight

Food intake was measured for the 3 d dietary treatment. Food intake intake is

expressed as mg/g body weight. Body weight was measured at the onset of the 3 d

dietary treatment.


Analytical Methods

..Cd was measured using a Packard Cobra II gamma spectrometer equipped with

a 3-inch crystal (Packard, Downers Grove, IL). Metallothionein protein was measured

by the cadmium (9Cd) binding assay (Eaton and Toal 1982). Total protein was

measured by the method of Lowry et al. (1951). Serum zinc concentrations were

measured by flame atomic absorption spectrophotometry (AAS) (Hempe and Cousins

1989). Tissue zinc was measured as described previously (Dunn and Cousins 1989).

Serum alanine aminotranferase (ALT) enzyme activity was measured

spectrophotometrically as the formation of pyruvate from alanine and a-ketoglutarate

using a commercial diagnostic kit (Sigma 505-P). The pyruvate formed is reacted with

2,4-dinitrophyenylhydrazine, forming a 2,4-dinitrophenylhydrazone derivative that can

be measured spectrophotometrically (Xmax = 505 nm). The absorbance is converted to

ALT activity units using a standard curve generated for pyruvate.








Total thiol groups were measured in liver homogenates spectrophotometrically

after treatment with 5,5'-dithio-bis(2-nitrobenzoic acid) (DTNB) (Jocelyn 1989).

Reaction of sulfhydryl groups with DTNB generates a yellow chromophore (Xmax = 412

nm, extinction coefficient of & = 13100 M' cm'). Nonprotein thiols were measured by

the same technique after first removing protein thiols by TCA precipitation (5% TCA).


Histological Analysis of Liver

Sections of liver were fixed in 10% buffered formalin, embedded in paraffin, and

stained with hematoxylin and eosin. These sections were analyzed visually for necrosis

and other signs ofhepatotoxicity (Khoo et al. 1996). Micrographs were obtained with a

Zeiss Axiovert S100 microscope (Carl Zeiss, Thomwood, NY) fitted with a CCD camera

for processing of digital images.


Statistics

Data were analyzed by ANOVA for a three way factorial design (2x2x2) to

determine significant main effects and interactions using genotype, dietary zinc and

oxidant treatment as independent variables (SAS, SAS Institute Inc. Cary, North

Carolina). The Tukey-Kramer multiple comparison test was used to determine significant

differences between specific groups. Serum ALT data were log transformed to obtain

homogeneous variances. Significance was established at p < 0.05.






46


Control Knockout
A 50
10+CO
1 1O+CC14
40-
S40- A 500+CO
| t A 500+CCI4
= 30-
20- - "



B 90 -- -- ------ ----------
120
10






E 600



300-


0 C I I I I I I I I I

C 30






10


I I I I I a__ _---_ -,-
0 12 24 36 48 0 12 24 36 48
Hours Hours

Figure 4-1. Indices of zinc homeostasis in metallothionein knockout (KO) and control
(CK) mice 0, 12, 24, or 48 h after injection of carbon tetrachloride (CC4) or corn oil
(CO). The mice had been fed either adequate dietary zinc (Zn10) or supplemental
dietary zinc (Zn500). (A) Serum zinc concentration (pmol Zn/L serum). (B) Liver zinc
(nmol/g liver). (C) Liver metallothionein (nmol/g liver). Data are presented as means
SE ofn = 3 (0 h) or n = 4-7 (12, 24, and 48 h) mice/group. Statistical differences (p
< 0.05) were determined by ANOVA for a three way factorial design, followed by the
Tukey-Kramer multiple comparison test.








Results

Food Intake

Diet consumption was similar between adequate zinc (Zn1o) and supplemental

zinc (Zn500) groups, and was also similar between TG and CT groups (data not shown).

Food consumption was statistically lower in KO than CK mice (p = 0.05), but the

difference was unlikely to have been biologically significant (0.46 mg *g bw' 3d' vs

0.48 mg g bw-' 3d-').


Zinc Status and Metabolism

Serum zinc, liver zinc and liver metallothionein. were measured at 0 h, 12 h, 24 h,

and 48 h post-dose (Fig. 4-1 & 4-2). Since mice in the 0 h group did not receive

injections, data at this time point represent the effects of diet and genotype only.

Knockout mice. The Zn5oo diet significantly increased serum zinc concentration

in both genotypes, but to a greater extent in control (CK) mice (Fig. 4-1 A). The only

significant effect of genotype was seen at 12 h after the CC14 treatment, when the serum

zinc concentration increased by 80% in KO mice only.

At Oh liver zinc content is significantly lower in KO mice than CK mice when fed the

Zn,0 diet, but not the Zno diet (Fig. 4-1B). Twelve hours later, CK mice fed the Zn5oo

diet and injected with CO (Zns0o+CO) had 25-45% more liver zinc than other CK groups

and significantly more (>100%) liver zinc than all KO groups. This effect may have been

due to the stress of the injections themselves, but all mice received injections at this time

point. The remaining CK groups had 35-90% more liver zinc than KO groups, but the

differences were not statistically significance. At 24 h and 48 h there was significantly

more liver zinc in CKs0o mice compared to all other genotype-diet combinations. The






48


only significant effect of CC14 treatment was at 24 h, when CC14-treated mice had lower

liver zinc values than CO treated mice. KO mice did not sequester zinc in the liver in


A


5L




B36
0
E








9
2







B36

C0
3 27





27
C.)

3 18

2-



C
22


HiIH
6

E- 11


> 5


Control


40-

30

20- o--3


10
00
00


00-


00


00-



20
T5



0
>5 -

0 -

15 1.


J


0 12 24 36 48 0 12 24 36 48
Hours Hours


Figure 4-2. Indices of zinc homeostasis in metallothionein transgenic (TG) and control
(CT) mice 0, 12, 24, or 48 h after injection of carbon tetrachloride (CC14) or corn oil
(CO). The mice had been fed either adequate dietary zinc (Zn10) or supplemental dietary
zinc (Zns500). (A) Serum zinc concentration (pmol Zn/L serum). (B) Liver zinc (nmol/g
liver). (C) Liver metallothionein (nmol/g liver). Data are presented as means SE ofn =
3 (0 h) or n = 4-7 (12, 24, and 48 h) mice/group. Statistical differences (P < 0.05) were
determined by ANOVA for a three way factorial design, followed by the Tukey-Kramer
multiple comparison test.


Transgenic

10o+CO
_~ 1 10+CCI4
o i o+cci4
A 500+CO
A 500+CC14






I I I






- l -- t '




S I I I I








response to the Zn5oo diet, CC4 treatment, or combination of the treatments.Liver

metallothionein levels depend on genotype and dietary zinc, and also are affected by

oxidant treatment and the stress of the injection (Fig. 4-1C). These values roughly mirror

hepatic zinc values. At 0 h, 150% more metallothionein is detected in CK5oo compared to

CKo mice. At 12 h after injection, CK50soo+CO mice had greater metallothionein values

than other mice, and CK50soo mice had greater values than CKio mice. Results are similar at

24 h. At 48 h, the metallothionein levels in the CK5soo+cc14 group increased, and both CK5oo

groups have five fold greater metallothionein levels than CKio groups. As expected, KO

mice did not express metallothionein. It appears that CC14 toxicity delayed the induction

of metallothionein in these mice.

Transgenic overexpressing mice. Zinc homeostasis was also altered by

metallothionein expression. Serum zinc was 80-100% greater in CT mice fed the Zn500oo

diet than Zno diet throughout the experiment (Fig. 4-2A). Serum zinc also rose in TG

mice fed the Zn5oo diet, but -50% less than that found in CT mice. At 12 h, there was a

trend toward decreasing serum zinc values after injection of vehicle alone in CT mice,

and both CCl4 and vehicle treated TG mice. In contrast, serum zinc values were

significantly greater in CT+CC4 mice at this time point. This is similar to the serum zinc

response of KO+CC4 mice. Throughout the time course of 48 h, TG mice exhibited

better control over the serum zinc concentration, especially when consuming the Zn5oo

diet. Liver zinc was greater in TG500o mice than CT5soo mice at 0 h (Fig. 4-2B). At 12 h,

however, TG50o mice had greater zinc values than all other mice, and TG+CCl4 mice had

significantly greater liver zinc than all other genotype-oxidant combinations. At 24 h the








TG500 mice had greater liver zinc values than all other diet/genotype combinations. Liver

zinc declined in all groups between 24 h and 48 h, but remained greatest in TG50o mice.

As with CK and KO mice, liver metallothionein values mirrored the liver zinc

values. TG500 mice had more liver metallothionein at 0 h than other diet-genotype

combinations. Metallothionein was elevated in both genotypes at 12 h after CC14

treatment. Metallothionein was elevated to its highest in all Zn5oo groups at 24 h after

injection, and were greatest when the injection contained CC4. At 48 h liver

metallothionein declined in all groups, but the TG50o and TGcc14 groups still had slightly

more metallothionein than all other genotype-diet and genotype-oxidant combinations,

respectively. Interestingly, CC4 alone had a minimal effect on liver metallothionein in

CT or TG mice unless combined with the Zn500 diet.


Control Knockout
10+CO
D 10+CC14 A
_ 3 500+CO

E i%
S IA -1\ '-
....\
2o' ,, " \ ;
B-a*1,


C I I I I I I I I I
0 12 24 36 48 0 12 24 36 48
Hours Hours
Figure 4-3. Serum alanine aminotransferase activity of metallothionein knockout (KO)
and control (CK) mice 0, 12, 24, or 48 h after injection of carbon tetrachloride (CC4) or
corn oil (CO). The mice had been fed either adequate dietary zinc (Zno) or supplemental
dietary zinc (Zn50o). These activities were measured spectrophotometrically and
expressed as log activity units/mL serum. Data are presented as means SE of three (0 h)
or 4-7 mice/group (12, 24, and 48 h). Data are presented as means SE ofn = 3 (0 h) or
n = 4-7 (12, 24, and 48 h) mice/group. Statistical differences (p < 0.05) were determined
by ANOVA for a three way factorial design, followed by the Tukey-Kramer multiple
comparison test.








Hepatotoxicity and Oxidative Stress

Hepatotoxicity was assessed by measuring serum alanine aminotransferase

enzyme (ALT) activity, and by histological analysis of liver sections for signs of damage

and necrosis. Measurement of liver nonprotein thiols (a pool comprised mainly of GSH

molecules) and liver total thiols served as measures of oxidative stress.

Knockout Mice. All mice had similar and normal serum ALT activity levels at 0

h (Fig. 4-3). At 12 h all mice had significantly elevated serum ALT activity, but the mean

ALT level (actual ALT units) was 6-12 times greater in KO mice than in CK mice. At 24

h and 48 h after CC4, the ALT levels had declined, but were still significantly greater in

CC14 treated mice. There were no genotype effects at the 24 h and 48 h points, and no

effects of dietary zinc at any time point.

Hematoxylin- and eosin-stained liver sections were similar in all groups injected

with corn oil (data not shown). Consistent with serum ALT activities, at 12 h there was

significantly more liver necrosis in KO mice (Fig. 4-4B) compared to CK mice (Fig. 4-

4A). The perivenous regions of the KO liver sections displayed coagulation necrosis.

Cells were less eosinophilic, displayed a general loss of morphology, and many contained

pyknotic nuclei. Hepatic sinusoids were collapsed and the hepatic plate arrangement was

disrupted. In contrast, little necrosis was visible in CK liver sections. These differences

were not seen at 24 h and 48 h, however. There were no differences between dietary

groups.

Oxidative stress was measured as nonprotein thiol levels (Fig. 4-5A). Nonprotein

thiols (NPT) are a thiol pool made up largely of glutathione, and have been shown to

decrease in instances of oxidative stress (Powers et al. 1998). Although glutathione








concentrations are subject to circadian rhythms, this variable was controlled for by

inclusion of corn oil-treated mice for all combinations of genotype and dietary zinc.

Interestingly, NPT levels at 0 h tended to be lower in Zns5oo00 groups than Zn10 groups, but

the difference was not quite significant (p = 0.06). At 12 h post injection, there was a


Figure 4-4. Light micrographs (200X magnification) of hematoxylin- and eosin-stained
liver sections from metallothionein knockout (KO) or control (CK) mice 12 h after
injection of carbon tetrachloride (CCI4). (A) CK mouse; (B) KO mouse. All mice were
fed diets containing 10 mg Zn/kg diet (Zn10). More severe coagulation necrosis was seen
around the central vein (cv = central vein) in KO mice (Panel B) compared to CK mice
(Panel A). Results were similar in female mice and mice fed diets containing 500 mg
Zn/kg diet (Zn50o). No histological changes due to genotype or diet were noted in mice
given an injection of the corn oil vehicle. Bar = 100 pm.









general depression of NPT, but no differences among groups. At 24 h NPT levels were

significantly lower in the KO+CC4 groups, while KO+CO groups returned to normal.

Interestingly, at 48 h CC14 treated mice have significantly more NPTs in liver than CO

treated mice. Total thiols (TT) are a measure of both protein and nonprotein thiols

combined, including the thiols of metallothionein. There was a trend toward greater TT

levels in KO mice than CK mice at 0 h (Fig. 4-5B). At 12 h post treatment, however,



Control Knockout
400

300-


200-
~.E
g 100-



18
-- --------
too -




C10+CO
o 6-
(2 0| 10+CCI
A 500+CO
A so500+CC14
0I I I I I 1 I I I1
0 12 24 36 48 0 12 24 36 48
Hours Hours

Figure 4-5. Indices of liver thiol homeostasis and oxidative stress of metallothionein
knockout (KO) and control (CK) mice 0, 12, 24, or 48 h after injection of carbon
tetrachloride (CC14) or corn oil (CO). The mice had been fed either adequate dietary zinc
(Zn,0) or supplemental dietary zinc (Zns500). (A) Liver nonprotein thiols (nmol/g liver)
were measured spectrophotometrically. Protein thiols had been removed by TCA
precipitation. (B) Liver total thiols (pmol/g liver) were measured spectrophotometrically.
Data are presented as means SE ofn =3 (0 h) or n = 4-7 (12, 24, and 48 h) mice/group.
Statistical differences (p < 0.05) were determined by ANOVA for a three way factorial
design, followed by the Tukey-Kramer multiple comparison test.








liver TT levels in KO mice drop, and are significantly lower than in CK mice. At 24 h

and 48 h no differences were detected.

Transgenic overexpressing mice. Serum ALT activity was similar and normal at

0 h in all groups (Fig. 4-6). Serum ALT rose sharply in all groups treated with CC14 at 12

h, and remained elevated through 48 h. There were no differences due to genotype. ALT

was slightly greater in Zn500 treated mice than ZnI0 treated mice at 48 h, however. Of

note, ALT activities do not return to basal levels, as was observed with CK and KO mice

(Fig. 4-3)

Histological analysis of hematoxylin- and eosin-stained liver sections revealed

significant necrosis in all CC14-treated mice, but there were no significant differences

between genotype (Fig. 4-7A,B). There was significant lymphocyte infiltration in the


Control Transgenic
10+CO A 500+CO
o 10+CC14 A 500+CC14

< -

S1 2- >-



0 e I I I I I
0 12 24 36 48 0 12 24 36 48
Hours Hours

Figure 4-6. Serum alanine aminotransferase activity of metallothionein transgenic
(TG) and control (CT) mice 0, 12, 24, or 48 h after injection of carbon tetrachloride
(CC4) or corn oil (CO). The mice had been fed either adequate dietary zinc (Zn10) or
supplemental dietary zinc (Zn500). These activities were measured
spectrophotometrically and expressed as log activity units/mL serum. Data are
presented as means + SE of three (0 h) or 4-7 mice/group (12, 24, and 48 h). Data are
presented as means SE ofn =3 (0 h) or n = 4-7 (12, 24, and 48 h) mice/group.
Statistical differences (p < 0.05) were determined by ANOVA for a three way
factorial design, followed by the Tukey-Kramer multiple comparison test.













































Figure 4-7. Light micrographs (200X magnification) of hematoxylin- and eosin-stained.
liver sections from metallothionein transgenic (TG) or control (CT) mice 12 h after
injection of carbon tetrachloride (CC4). (A) CT mouse fed the ZnIo diet; (B) TG mouse
fed the Zn1o diet; (C) CT mouse fed the Zn50o diet; (D) TG mouse fed the Zns0o diet; (E)
enlargement of the perivenous region from figure 4-7D. Significant lymphocyte
infiltration is seen in the area directly surrounding the central vein (cv) of mice of both
genotypes fed the Zn50o diet after receiving CC14 (C-E). Results were similar in female
mice. No histological changes were related to genotype were noted in mcie given an
injection of corn oil vehicle and fed either diet. Bar = 100pm (A-D) or 25 lim (E).








area directly surrounding the central vein of ce of both genotypes fed the Zn500 diet (Fig.

4-7C-E).

There were no significant differences in NPT between TG and CT mouse livers at

0 h (Fig. 4-8A). Twelve hours after injection, the NPT levels were higher in CCl4-treated

mice. This relationship reversed at 24 h, and NPT levels were greatest in CT+CO mice.

Similar results were seen at 48 h.

Total thiol levels were similar in all mouse groups at 0 h (Fig. 4-8B). At 12 h

there is a trend toward greater TT levels in TGs500+CC14 mice compared to others. This is

likely due to induction of hepatic metallothionein. At 24 h, CCl4-treated mice had

significantly greater TT levels than CO treated mice. Also, TG mice had significantly

greater TT levels than CT mice. These significance disappeared at 48 h, but there is still

a trend toward greater TT levels in TGcc4 mice compared to TGco mice. Since the thiols

of metallothionein would be included in the TT measurement, and metallothionein

induction was greatest at 24 h and 48 h, metallothionein induction alone may have

accounted for the increased TT in TG mice at these time points. These also are the time

points of NPT suppression (Fig. 4-8A).


Discussion

We examined the effects of supplemental dietary zinc in combination with

different levels of metallothionein gene expression on susceptibility to oxidative stress in

vivo. Previous research showed that supplemental zinc protected rat hepatocytes cultures

from various cytotoxic agents (Coppen et al. 1988, Schroeder and Cousins 1990). Zinc

treatment increased metallothionein gene expression, and it was suggested that

metallothionein protein was the mediator of the protection. Similar results were seen









when metallothionein expression was elevated by interleukin-6 and dexamethasone

(Schroeder and Cousins 1990). In this experiment, we used KO and TG mice as models

of different levels of metallothionein gene expression (Masters et al. 1994, Palmiter et al.

1993). By using two zinc intake levels, we were able to generate different hepatic zinc


Control


1 Z 300
o o^

00g 150
Z


0 I I I I
24-



18-



6-

0 t I I
0 12 24 36 48
Hours


Transgenic
10+CO
0 1 O+CC14
A 500+CO
A 500+CC14


0 12 24
Hours


36 48


Figure 4-8. Indices of liver thiol homeostasis and oxidative stress of metallothionein
transgenic (TG) and control (CT) mice 0, 12, 24, or 48 h after injection of carbon
tetrachloride (CC4) or corn oil (CO). The mice had been fed either adequate dietary
zinc (Zn10) or supplemental dietary zinc (Zns500). (A) Liver nonprotein thiols (nmol/g
liver) were measured spectrophotometrically. Protein thiols had been removed by
TCA precipitation. (B) Liver total thiols (pimol/g liver) were measured
spectrophotometrically. Data are presented as means SE ofn = 3 (0 h) or n = 4-7
(12, 24, and 48 h) mice/group. Statistical differences (p < 0.05) were determined by
ANOVA for a three way factorial design, followed by the Tukey-Kramer multiple
comparison test.


I I I I I


.4
0>

0
II


I I I I I








and metallothionein levels to examine their effects on the toxicity of the hepatotoxin

CC14. CC4 is metabolized to the trichloromethyl radical by the enzyme cytochrome

P4502E 1 in the perivenous region of the liver lobule (McGregor and Lang 1996, Sipes et

al. 1990; Wong et al. 1998). As a result, the perivenous region is most affected by this

radical species, which causes lipid peroxidation and inactivation of enzymes such as the

cytochrome P450 enzymes (McGregor and Lang 1996).

In several studies zinc pretreatment protected against CC14-induced hepatotoxicity

in vivo (Cagen and Klaassen 1979, Chvapil et al. 1973, Dhawan and Goel 1995, Liu et al.

1998). This effect is only seen when copious amounts of dietary zinc (1000 mg/kg diet)

or parenteral zinc are given, however. For example, consumption of diets containing 300

mg Zn/kg, twenty fold the requirement for the rat, did not protect against CCl4-induced

hepatotoxicity in rats (Khoo et al. 1996). A proposed mechanism for the protective effect

of pharmacological doses of zinc is that zinc induces metallothionein, and that

metallothionein is the real mediator of the hepatoprotection. Metallothionein has been

shown to covalently bind CC4 metabolites and decrease the amount of these metabolites

bound to other cellular proteins (as reviewed in Klaassen and Liu 1998). In this way

metallothionein may prevent CC4 from reaching some of its cellular targets. Although

this might explain some of the zinc related protection, large parenteral zinc doses also

protect against CCl4 toxicity in the absence of metallothionein expression (Itoh et al.

1997). This might be related to suppression of CCl4 bioactivation, since supplementation

has been shown to inhibit the activity of some cytochrome P450 enzymes (Bray et al.

1986, Coppen et al. 1988).








Since metallothionein expression has been inversely related to damage after an

oxidative insult, we expected that the damage in mouse livers in this experiment would

be inversely proportional to metallothionein expression (i.e., TG < CT and CK < KO). In

support of this, enhanced susceptibility of KO mice to CC14 toxicity was reported recently

(Liu et al. 1998B). Also, we assumed that mice consuming the Zn5oo diet would be

protected compared to those eating the Znj0 diet since supplementation would induce

more metallothionein protein and result in greater cellular zinc accumulation. Finally,

since metallothionein expression was thought to be key to zinc-related cytoprotection, we

expected no protection against hepatotoxicity by zinc supplementation in KO mice.

In support of our hypothesis we saw greater hepatotoxicity in KO mice compared

to CK mice (6-12 fold higher ALT values) at the 12 h time point. At later time points,

however, toxicity seemed equivalent in these genotypes. This is in spite of the fact that

KO mice had no metallothionein present in liver and were unable to sequester additional

hepatic zinc. Also, TG and CT mice did not differ in the level of hepatotoxicity produced

despite huge differences in hepatic metallothionein and zinc. Further, the nonprotein thiol

levels were not dramatically altered during the experiment, similar to results from

experiments with acetominophen in KO mice (Liu et al. 1999A).

To our knowledge, this is the first report where the effect of supplemental dietary

zinc on CC14-induced hepatotoxicity in mice was examined. Also, this is the first

assessment of the combination of supplemental dietary zinc and toxicity of any kind in

metallothionein overexpressing mice and metallothionein knockout mice. We show for

the first time using these models that neither supplemental dietary zinc nor

metallothionein overexpression alone protected against CCl4-induced hepatotoxicity in








mice. Further, no combination of metallothionein gene expression and either adequate or

supplemental dietary zinc provided protection, even though the levels of liver zinc and

liver metallothionein varied over a large range among groups.

These results argue against a direct antioxidant role for MT against CCl4 toxicity,

since antioxidant protection against hepatotoxicity would likely be dose dependent

(Tirmenstein et al. 1997, Yao et al. 1994). Instead, the data fit better in a plateau model,

where metallothionein expression was important up to a point (< the level in CT and CK

mice), but beyond this point further expression is not useful. This is more in line with a

metallothionein-specific function, such as regulation of tissue zinc accumulation and/or

intracellular zinc trafficking. Specifically, KO mice might be less protected against CC14

than CK mice because they are unable to regulate zinc homeostasis appropriately (Coyle

et al. 1995, Davis et al. 1998, Philcox et al. 1995).

The most marked differences between genotypes in this experiment were

alterations in zinc metabolism. In both the TG and the KO experiments, the animals with

the lowest metallothionein expression displayed the weakest control over serum zinc

levels 12 h after CC14 treatment. In fact, serum zinc rose in KO mice and declined in TG

mice. This is coincident with induction of hepatic metallothionein and elevation of

hepatic zinc in all but the KO mice at this time point. This is also coincident with greater

hepatotoxicity in KO mice.

Metallothionein's role in maintaining appropriate hepatic zinc levels might be

especially important under conditions of stress, including oxidative stress.

Metallothionein is induced during the acute phase response and during hepatic

regeneration, and is required for normal hepatic regeneration after partial hepatectomy








(Arora et al. 1998, Ohtake 1978). Also, zinc can be mobilized from metallothionein by a

number of oxidants or shifts in glutathione redox status (Berendji et al. 1997, Fliss and

Menard 1992, Tatsumi and Fliss 1994). These may be mechanisms for mobilization of

intracellular zinc during oxidative stress (Maret 1995). The end result may be enhanced

transfer of zinc from metallothionein to zinc-dependent proteins (Jiang et al. 1998). Zinc

mobilization may also affect gene expression.

If metallothionein acts as a zinc-donating molecule during oxidative stress, KO

mice might not be able to keep up with the demand for zinc incorporation into zinc-

dependent proteins. As a result, the level of functioning zinc-dependent proteins

produced during oxidative stress may not be sufficient in KO mice, and may explain why

KO mice were more adversely affected than CK mice. On the other hand, TG mouse

cells have more zinc available for donation to zinc-dependent proteins, but they may not

be able to produce those proteins rapidly enough to take advantage of the excess zinc. In

this case, the level of MT protein produced in CT mice is both necessary and sufficient,

and TG mice would not be better protected.

While other explanations for the results of this experiment exist, we can rule out

several. First of all, the results of this experiment are not likely affected by differences in

other antioxidants in these mice since the levels of other antioxidant enzymes and

molecules are reported to be similar between genotypes (Iszard et al. 1995, Kang et al.

1997, Lazo et al. 1995, Liu et al. 1999, Rofe et al. 1998, Wang et al. 1999, Wu and Kang

1998, Zheng et al. 1996). Also, bioactivation ofCC14 should be similar between

genotypes since the activity ofcytochrome P4502E1 is similar in TG and KO mouse

livers to their respective controls (Iszard et al. 1995, Itoh et al. 1997, Rofe et al. 1998).








Since these results are in opposition to some experiments done in rats, we cannot exclude

species difference as a confounding variable. However, it seems unlikely that a process

as basic as radical scavenging (antioxidant action), a simple oxidation-reduction reaction,

would differ between two rodent species. It should be noted that different strains of mice

were used in these experiments, and, therefore, the results are not likely due to

peculiarities of any individual inbred strain. However, we cannot rule out the possibility

that knockout mice are better protected than we expected due to some adaptive

mechanismss. For instance, altered gene expression has been reported in KO mice

(Kimura et al. 2000).

Two other research groups have published reports of CC14-induced hepatotoxicity

in KO mice, but no reports have been forwarded for TG mice (Itoh et al. 1997, Liu et al.

1998B). Those KO studies were designed to differentiate the hepatoprotective effects of

exogenous compounds (sakuraso-saponin and oleanic acid) from their ability to induce

metallothionein synthesis. Both groups used the same KO mouse model, injected 50 tl

CCl4/kg bw, and reported results for 24 h post-dose. Itoh and coworkers found no

difference in hepatotoxicity between genotypes as measured by plasma GOT activity. Liu

and coworkers found greater damage in KO mice as measured by serum ALT, serum

SDH, and histological analysis of hematoxylin- and eosin-stained liver sections. Itoh and

coworkers injected the dose subcutaneously, while Liu and coworkers injected the dose

intraperitonealy. Slower uptake of CC14 from the subcutaneous injections of Itoh and

coworkers might explain why no toxicity was seen at the 24h time course in their study.

Serum ALT values and histological analysis from the 12 h time point in this report match

well with the results of Liu and coworkers at 24 h. The differences in the time course of








hepatotoxcity between that experiment and ours may have been due to the smaller dose

(20 p.1 CCl4/kg bw) used in our experiments.

The lack of protection against oxidative stress by MT overexpression in this

experiment is in line with results from several other experiments that used models of

metallothionein gene overexpression. Early studies using transfection of the human MT-

2a gene into Chinese hampster ovary Kl-2 cells and several of tumor cells lines found no

resistance against free radical generating agents (Kaina et al. 1990, Kelley et al. 1988).

Further, TG mice were not resistant to adriamycin cardiotoxicity or y-irradiation

(DiSilvestro et al. 1996, Liu et al. 1999B). The existing evidence that metallothionein

overexpression alone protects against oxidative stress was found in transformed cells and

sheep pulmonary artery endothelial cells (SPAEC) transfected with metallothionein, and

with a second TG mouse strain that only overexpresses metallothionein in the heart. In

the former case, transfection of NIH 3T3 cells with MT-1 protected against nitric oxide

induced cytotoxicity (Schwarz et al. 1995). Viable cell determinations were not made

until 6 to 7 days after nitric oxide treatment, however, so it could be argued that the

difference in the number of cells remaining at that time was due to improved cell

recovery instead of radical scavenging. Protection against hyperoxia and tertiary-butyl

hydroperoxide was also seen in SPAEC transfected with mouse or human MT genes, but

the determination of viable cell numbers were not performed until 1-2 days after oxidant

exposure was initiated (Pitt et al. 1997). Again, it is difficult to separate the contributions

of antioxidation and cell recovery to cell survival.

In the case of the heart-specific metallothionein overexpressing TG mouse strain,

there is both in vivo and in vitro evidence that metallothionein protected against








oxidative stress, including ischemia reperfusion, hydrogen peroxide, and doxorubicin

treatment (Kang et al. 1997, Kang et al. 1999, Wang et al. 1999, Wu and Kang 1998).

While this provides convincing evidence for an antioxidant role for metallothionein in

cardioprotection, it should be noted that a very high level of overexpression (10-130-

fold) is needed to see these effects, as lower levels of overexpression (3-fold) didn't

protect against adriamycin cardiotoxicity (DiSilvestro et al 1996). Further, mouse heart,

an organ that metallothionein is not normally abundant in, has a much weaker antioxidant

capacity than mouse liver (as reviewed in Kang 1999). So while this model confirms that

metallothionein overexpression can protect mouse heart against oxidative stress, the

effects may be small since they were unmasked only when metallothionein expression

was astronomically high, in an organ that doesn't normally produce much

metallothionein, yet is highly susceptible to oxidative stress. Therefore, it cannot be

assumed that the same results would be seen in the liver.

The lack of protection against CCI4 hepatotoxicity by zinc supplementation (500

mg/kg diet) in any of the genotypes used in this experiment strongly suggests that the

required dietary zinc level for mice (10 mg/kg diet) provides as much protection as is

possible by dietary zinc. It also suggests that the hepatoprotective effects associated with

zinc injection are not readily reproduced with dietary zinc. Combined with data from zinc

deficiency studies we see a plateau affect of dietary zinc against oxidative stress in the

rodent model, just as we do with metallothionein. Zinc-deficient diets render rodents

more susceptible to oxidative stress, zinc-adequate diets alleviate this condition, but

supplemental dietary zinc provides no further protection.





65


The results of the experiments explained above confirm the importance of

metallothionein expression in protection against oxidative stress, but bring into question

the impact of supplemental zinc and/or elevated metallothionein expression in defense

against oxidative stress in mouse liver in vivo. Further, the protection against oxidative

stress appears to correlate with changes in zinc metabolism.














CHAPTER 5
EFFECTS OF METALLOTHIONEIN GENE EXPRESSION AND SUPPLEMENTAL
ZINC IN PROTECTION AGAINST OXIDATIVE STRESS IN PRIMARY
HEPATOCYTE CULTURES FROM METALLOTHIONEIN TRANSGENIC AND
METALLOTHIONEIN KNOCKOUT MICE

Introduction

As indicated in the previous chapter, metallothionein and zinc are implicated in

cellular antioxidant defense. In the previous experiment with metallothionein knockout

mice (KO) we found that metallothionein expression protected against carbon

tetrachloride-induced oxidative stress, but metallothionein overexpression in transgenic

mice (TG) did not provide further protection. Further, mice fed diets with supplemental

zinc (500 mg Zn/kg diet) were not protected compared to mice fed only the required zinc

intake (10 mg Zn/kg diet). These results are inconsistent with direct antioxidant activity

of metallothionein induction or supplemental zinc. To more directly assess the cellular

antioxidant functions of supplemental zinc and metallothionein, We studied the effects of

metallothionein expression and moderate zinc supplementation on tertiary-butyl

hydroperoxide-induced cytotoxicity in primary hepatocyte cultures from TG and KO

mice. Specifically, hepatocytes from metallothionein knockout mice can be used to

determine if zinc acts in cytoprotection independent of metallothionein by examining

whether these cells had altered sensitivity to tertiary-butyl hydroperoxide. We also

determined whether treating cells with zinc and/or dexamethasone and interleukin-6, all

of which induce metallothionein expression, influences sensitivity to tertiary-butyl








hydroperoxide in these genotypes. The results of this study provide further evidence that

metallothionein and supplemental zinc do not act directly as antioxidants.


Materials and Methods

Animal Model

Metallothionein knockout and metallothionein transgenic mice used in this study

were bred in house using founder mice purchased from The Jackson Laboratory, Bar

Harbor, ME. The metallothionein overexpressing mice (designated TG mice) were

originally generated in C57BL/6 mice crossed with SJL mice (Palmiter et al. 1993).

Backcrossing against C57BL/6 mice permit the use of C57BL/6 mice as controls

(designated CT mice). The metallothionein knockout mice (designated KO mice) were

generated in 129/SvCPJ mice crossed with C57BL/6 mice (Masters et al. 1994). These

mice were maintained on a 129/SvCPJ background. 129S3/SvImJ mice served as controls

(designated CK mice). All experiments used 7-11 wk old female mice. Mice were housed

in plastic box cages with a 12 h light:dark cycle. The mice were given access to tap water

and a standard rodent diet (Harlan Teklad 8604, Madison, WI) until they were used for

hepatocyte preparations. Care and treatment of the mice received approval of the

University of Florida Institutional Care and Use Committee.


Hepatocyte Isolation

All liver perfusions began between 8 AM and noon. In all experiments

hepatocytes were collected from mice of both genotypes (KO and CK or TG and CT) on

the same day. Mice were anaesthetized with sodium pentabarbital (60 mg/kg i.p.).

Hepatocytes were isolated using a two-step perfusion technique run retrograde from the








inferior vena cava, with perfusate allowed to flow out of the portal vein without

recirculation (Schroeder and Cousins 1991, Renton et al. 1978). Livers were first

perfused with a Ca-free buffered solution (8-10 mL/min), followed by a buffered solution

containing collagenase. After perfusion the liver was aseptically transferred to a sterile

cell culture hood. The liver was disrupted and the cells were liberated using a cell

scraper.


Hepatocyte Culture

The hepatocytes were suspended in wash medium (Williams Medium E with 10

mmol/L HEPES buffer and 10 mmol/L TES buffer) followed by centrifugation (50 x g

for 3 min) to separate viable cells from debris. The wash and centrifugation steps were

repeated two times. The final cell pellet was resuspended in culture medium (Williams

Medium E supplemented with 10% FBS, 100 nmol/L insulin, 100 units/mL penicillin,

100 tg/mL streptomycin). Cell number and viability were determined using a

hemacytometer and trypan blue exclusion. Preparations with greater than 85% viability

were used for experiments. Cells were added to type I collagen-coated 35 mm tissue

culture plates (0.5x 106 cells/plate in 2 mL medium) and allowed to attach (37C, 5%

C02). After a 2-3 h attachment period the culture medium was removed and replaced

with 1 mL of one of four different medium combinations: medium containing 4 jtmol

zinc/L, medium with added zinc (32 jtmol/L), medium with dexamethasone (1 gmol/L)

and interleukin-6 (100 units/mL), and medium with added zinc, dexamethasone and

interleukin-6. Cells were maintained in these culture conditions for 18-22 h prior to

addition of tertiary-butyl hydroperoxide (TBH). Carbon tetrachloride was also used in








some experiments, but its nonpolar nature and volatility led to problems with

reproducibility within and between experiments.

Cytotoxicity assays

After the culture period was complete the medium was again removed and

replaced with medium containing 0 to 500 gmol/L TBH for 30 to 150 min. At the end of

TBH treatment, cells and medium were prepared for determination of lactate

dehydrogenase (LDH) activity. Briefly, medium was centrifuged (13000 x g) to remove

unattached cells and debris, and then stored at -20C for up to 48 h before assay. Cells

were removed in 1% triton X-100, disrupted by repeated passage through 200 jLL pipette

tips, and centrifuged (13000 x g). The supernatant was diluted with 4 volumes of Tris

buffer (100 mmol/L 2-amino-2-(hydroxymethyl)-1,3-propanediol, pH 7.5) and stored at -

20C for up to 48 h before assay. LDH activity was assayed spectrophotometrically

using a commercial kit (Sigma LD-L) by incubating aliquots of medium or cell extracts

(125 pL) with reagent solution (375 pL) containing lactate (50 mmol/L) and NAD (7

mmol/L) in buffer (pH 8.9). LDH in the sample catalyzes the reaction of lactate and

NAD, resulting in the production of pyruvate and NADH. Formation of NADH results in

an increase in absorbance at 340 nm. The rate of increase in absorbance is directly

proportional to LDH activity in the sample. The percent LDH leakage is calculated as the

LDH activity of the medium as a percentage of the sum of the LDH activities of the

medium + the cells (Jauregui et al. 1981).

Cell viability was also determined by the ability of live cells to convert 3-[4,5-

dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT, Sigma) to a colored

product (Denizot and Lang 1986). In these experiments, cells were cultured on collagen








coated 96 well plates (3 x 104 cells/well in 200 iL WME) and treated as described above

for 35 mm plates. At the end of the culture period the medium was replaced with medium

containing 1-50 ptmol/L TBH and incubated for 120 min. This lower range of TBH

concentrations was used in these experiments to correct for the greater ratio of medium

volume-to-cell number. Medium containing TBH was then replaced with serum-free

culture medium containing MTT (2.4 [imol/L) for 3 h. MTT is a tetrazolium dye that is

converted to an insoluble purple formazan through cleavage of the tetrazolium ring by

dehydrogenase enzymes of live cells. Cultures were washed and then solubilized in 100

iL acidic isopropanol (40 pimol/L HC1 in absolute isopropanol). After shaking the wells

were analyzed (A560-A650o). Data are presented as percent of activity of untreated cultures.


Analytical Methods

For intracellular glutathione analysis cells were homogenized in 1% (39 mmol/L)

sulfosalicylic acid (2-hydroxy-5-sulfobenzoic acid; SSA) and placed on ice (> 20 min) to

precipitate proteins, which were then removed by centrifugation (13000 x g). The

supernatants were frozen overnight (-20C). Total glutathione (oxidized + reduced) was

measured spectrophotometrically by the glutathione (GSH) reductase-recycling assay

using a microtiter plate reader (Baker et al. 1990). Briefly, supernatants were diluted with

buffer (100 mmol/L sodium phosphate, 1 mmol/L EDTA, pH 7.5) and reacted with 5,5'-

dithiobis-(2-nitrobenzoic acid) (DTNB; 150 pmol/L) in the presence of NADPH (100

pmol/L) and GSH reductase (1.0 units/mL). Reaction of GSH with DTNB generates

oxidized glutathione (GSSG) and the highly colored 5-thio-2-nitrobenzoic acid anion,

which can be measured at X = 410 nm. GSH is regenerated from GSSG by glutathione

reductase (using reducing equivalents from NADPH), which allows the reaction between








GSH and DTNB to continue at a linear rate. GSH concentration is determined from a

standard curve using GSSG. '9Cd (1.35 GBq/nmol; Isotope Product Laboratories,

Burbank, CA) was measured using a Packard Cobra II gamma spectrometer (Packard,

Downers Grove, IL). Metallothionein protein content of the cells was measured prior to

treatment with oxidant by the cadmium ('"9Cd) binding assay (Eaton and Toal 1982).

Briefly, cells were lysed in 10 mmol/L Tris buffer containing protease inhibitors (0.1

mmol/L phenylmethylsulfonylfluoride, 1.2 1tmol/L leupeptin and 1.5 p.mol/L pepstatin

A). Homogenates were centrifuged (10000 x g), the supernatant was boiled and

centrifuged again (10000 x g), and the resulting supernatant was incubated with 'gCd.

Unbound '"Cd was removed using hemoglobin. "9Cd bound to metallothionein was

measured by y-counting, and converted to moles of metallothionein (Davis et al. 1998).

Cell zinc was measured by atomic absorption spectrophotometry after cells were

solubilized in an aqueous solution of sodium dodecyl sulfate (7 pmol/L) and sodium

hydroxide (400 mmol/L) as previously described (Schroeder and Cousins 1991). Total

protein was measured by the method of Lowry et al. (1951).


Statistics

Data were analyzed by ANOVA for a three way factorial design (2x2x2) to

determine significant main effects and interactions among genotype, medium zinc and

dexamethasone/Il-6 treatment (SAS, SAS Institute Inc. Cary, North Carolina). The

Tukey-Kramer post hoc test was used to determine significant differences between

specific groups when interactions were significant (p < 0.05).








Results

Hepatocyte metallothionein expression was directly related to MT genotype and

was inducible by zinc treatment, dexamethasone and 11-6 (dex/Il-6) treatment, and the

combination of these treatments in all but KO cells (Fig. 5-1). Also, induction by these

treatments was greater in TG cells than either control strain. Further, metallothionein

induction by the combination of zinc treatment and dex/Il-6 treatment was synergistic.

Cell zinc concentrations were lower in KO cultures than CK cultures under all

treatment conditions (Table 5-1). As with metallothionein, cell zinc was increased in CK

cultures in response to zinc treatment and dex/I1-6 treatment, and was increased

synergistically by these two treatments combined. Cell zinc was not affected by these

treatments in KO cultures, however.

Lactate dehydrogenase (LDH) enzyme leakage was used as a measure of

cytotoxicity. Leakage was not greatly different between CK and KO cultures after

treatment with TBH (350 pmol/L or 450 pmol/L) when cells were not previously

exposed to MT inducers (Fig. 5-2A). Zinc pretreatment modestly elevated LDH

leakagefrom CK mouse hepatocytes treated with 450 pimol/L TBH, but did not alter

toxicity in KO cultures. Treatment with dex/I1-6 slightly increased LDH leakage in KO

cells in response to 450 ptmol/L TBH, but increased toxicity in CK cells by three fold. It

should be noted that neither zinc treatment nor dex/I1-6 treatment increased LDH leakage

in the absence of TBH. The combination of zinc treatment with dex/I1-6 treatment was

associated with a similar level of TBH- induced toxicity as dex/IL-6 treatment alone.

Similar trends were seen in experiments with hepatocytes from TG and CT mice.

(Fig. 5-2B). LDH leakage was already greater in TG cultures than CT cultures without








exposure to metallothionein-inducing agents. Pretreatment with zinc slightly elevated

LDH leakage in CT cultures when exposed to 450 .mol/L TBH, but nearly doubled

leakage from TG cells. Treatment with dex/Il-6 enhanced leakage even further in both

genotypes, but again the increase was greatest in TG cultures. Toxicity was 5-fold greater

in TG cultures pretreated with dex/Il1-6 and exposed to 350 pmol/L TBH than identically

treated CT cultures. Similar to the KO experiment, the combination of zinc treatment and

dex/Il-6 treatment resulted in TBH-induced toxicity similar to that seen with dex/Il-6

treatment alone.

5000 -d
control d
4500 -
SZn
4000 -
w Dex/II-6
3500 -
0 Zn+ Dex/jII-6
0 3000 -
WW
S2500

S2000
0
S1500 c

1000 c

500 ac b kl
0 aa a.

KO CK CT TG

Figure 5-1. Metallothionein content of metallothionein knockout (KO), knockout control
(CK), metallothionein transgenic (TG), and transgenic control (CT) mouse hepatocytes.
Cells were cultured in Williams Medium E (WME) containing either 4 imol zinc/L
(control), 32 tmol Zn/L (Zn), 1 pmol dexamethasone/L and 100 units Il-6/mL (Dex/Il-6),
or 32 lmol Zn/L, 1 pmol dexamethasone/L and 100 units Il-6/mL (Zn + Dex/Il-6) for 20
h prior to harvest. MT was quantified for KO hepatocytes cultured in control or zinc
medium only. Data are means SEM (n = 4-8 cultures). Significant differences (p <
0.05) were determined using ANOVA for a 2x2x2 factorial design, followed by the
Tukey-Kramer multiple comparison test.








Table 5-1. Cell zinc concentration in hepatocyte cultures from metallothionein knockout and control mice after treatment with zinc or
dexamethasone and 11-6abc'

Cell zinc (jAmol/g protein) ANOVA Results

Genotype Control Zn Dex/Il-6 Zn & Dex/Il-6 p


CKd 12 lw,x 13 0" 18 + Oy 24 0z
KOe 101w 11 +0w 101w 11 0w
Genotype x Zinc x Dex/Il-6 0.0095
Genotype x Zinc 0.0002
Genotype x Dex/I1-6 0.0001
Zinc x Dex/I1-6 0.0005
Genotype 0.0001
Zinc 0.0001
Dex/iIl-6 0.0001


a Cell zinc values are means SEM, n = 6.
b Zinc treatment was 32 gimol/L for 20 h.
c Dexamethasone and 11-6 treatment (Dex/II-6) was 1 tmol/L Dex and 100 units Il-6/mL for 20 h.
d CK, hepatocyte cultures from control mice.
' KO, hepatocyte cultures from metallothionein knockout mice.
wxyz Values followed by different superscript letters are significantly different.









Cell viability measured with the MTT assay revealed results similar to

those seen with the LDH assay. Viability was greater in untreated or zinc-treated KO


A30

25

S20

S15

S 10

5

0




100
B 90
80
70
U
S60
S50
S40
S30
20
10
0


&? V
& x^ px
&- I?


Figure 5-2. Tertiary-butyl hydroperoxide-induced cytotoxicity. Cytotoxicity was
measured as Lactate dehydrogenase leakage from metallothionein knockout (KO),
knockout control (CK), metallothionein transgenic (TG), and transgenic control (CT)
mouse hepatocytes after treatment with tertiary-butyl hydroperoxide (TBH). Hepatocytes
were cultured as in figure 5-1 for 20 h, then treated with 0, 350, or 450 jimol/L TBH for
2 h. (A) LDH leakage from CK and KO mouse hepatocytes. (B) LDH leakage from CT
and TG mouse hepatocytes. Data are means SEM (n = 5-6 cultures from a
representative experiment). Significant differences (p < 0.05) between groups treated
with TBH were determined using ANOVA for a 2x2x2 factorial design, followed by the
Tukey-Kramer multiple comparison test.








cultures than CK cultures when exposed to 10 imol/L TBH, and in all KO cultures

compared to CK cultures at 15 gmol/L TBH (Table 5-2). All TG cultures had lower

viability than CT cultures at 10 p.mol/L TBH (Table 5-3). Nearly all cells from these

genotypes were dead at 15 jtmol/L TBH.

Total glutathione was measured after the culture period but before the addition of

TBH to determine if glutathione status was affected by genotype, medium zinc and/or

dex/I1-6 treatment (Table 5-4). Glutathione concentrations were 25-55% lower in CK

cultures than KO cultures, depending on the culture condition. Dex/Il-6 treatment

reduced cellular GSH in both genotypes, but to much lower levels in CK cultures. Zinc

treatment alone, or in combination with dex/I1l-6 had little affect on glutathione levels.


Discussion

Previous work from this lab found that zinc supplementation protected against

oxidative stress in primary cultures of rat hepatocytes (Coppen et al. 1988, Schroeder and

Cousins 1990). Protection was credited, in part, to metallothionein induction by zinc.

Induction of metallothionein with hormones and cytokines was also correlated with such

protection (Schroeder and Cousins 1990). We recently found that metallothionein

expression protected mice against carbon tetrachloride heptotoxicity, but neither

supplemental dietary zinc nor metallothionein overexpression provided further protection

in metallothionein knockout and metallothionein transgenic mice (Davis et al.

submitted). These results were inconsistent with general antioxidant functions for

metallothionein and supplemental dietary zinc, and are in opposition to a number of

previously published reports (Liu et al. 1998A, Liu et al. 1999A, Rofe et al. 1998). With








Table 5-2. Cell viability after tertiary-butyl hydroperoxide exposure of hepatocyte
cultures from metallothionein knockout and control mice previously treated with zinc
and/or dexamethasone and 11-6abc

Cell Viability (% of untreated cultures) ANOVA Results

Genotype TBH Control Zn Dex/Il-6 Zn &
imol/L Dex/Il-6 p value


5 61 2
84 1









10 54 3
904


15 111
32 2


65 2
70 1









52 2
73 3


91
28 1


CKd
KOe


a Cell viability values are means + SEM, n = 10-11.
b Zinc treatment was 32 gmol/L for 20 h.
' Dexamethasone and 11-6 treatment (Dex/I1-6) was 1 gmol/L Dex and 100 units Il-6/mL
for 20 h.
d CK, hepatocyte cultures from control mice.
e KO, hepatocyte cultures from metallothionein knockout mice.


63 2 82 1
62 2 63 2
Genotype x Zinc x Dex/Il-6
Genotype x Zinc
Genotype x Dex/Il-6
Zinc x Dex/Il-6
Genotype
Zinc
Dex/Il-6

66 2 72 2
66 3 62 2
Genotype x Zinc x Dex/I1-6
Genotype x Zinc
Genotype x Dex/Il-6
Zinc x Dex/iIl-6
Genotype
Zinc
Dex/Il-6

121 181
27 2 26 2
Genotype x Zinc x Dex/Il-6
Genotype x Zinc
Genotype x Dex/Il-6
Zinc x Dex/I1l-6
Genotype
Zinc
Dex/Il-6


CK
KO


CK
KO


0.8090
0.0001
0.0001
0.0001
0.0625
0.0741
0.0655



0.4920
0.0020
0.0001
0.0030
0.0001
0.0435
0.8038



0.5875
0.0380
0.0016
0.0063
0.0001
0.7010
0.5909








the experiments described herein we attempted to determine whether metallothionein

induction and supplemental zinc could act as cellular antioxidants against TBH-induced

toxicity.

We used primary hepatocytes cultures from TG and KO mice as a cellular model

of stably altered metallothionein expression, independent of metallothionein inducing

agents. In addition we assessed the effects of zinc and/or dexamethasone and IL-6

treatment on cytotoxicity in these cells, which have inherently different abilities to

produce metallothionein. In these experiments we found a consistent, direct relationship

between cellular metallothionein content and susceptibility to TBH toxicity. This

relationship occurred whether metallothionein was elevated due to genotype, inducing

agents (zinc or dexamethasone and 11-6 treatment), or their combination. Further,

dexamethasone and 11-6, powerful metallothionein inducers and protectors against

hepatocyte toxicity in rats, increased sensitivity to TBH toxicity in these mice. To our

knowledge, this is the first report that demonstrates an inverse relationship between

metallothionein expression and susceptibility to an oxidative stress.

Although the results of this study were unexpected, possible explanations can be

proposed. Antioxidant defense against the toxicity of tertiary-butyl hydroperoxide

involves consumption of considerable quantities of glutathione (GSH) by the antioxidant

enzymes glutathione peroxidase and phospholipid glutathione peroxidase (Rush et al.

1985). In addition, glutathione transferase enzymes may detoxify genotoxic products of

lipid peroxidation by conjugatng them to glutathione (Hubatsch et al. 1998). As such,

disturbances in GSH metabolism may enhance TBH toxicity. For example, hepatocyte

glutathione levels decline when cysteine availability is low (Wang et al. 1997). In our









Table 5-3. Cell viability after tertiary-butyl hydroperoxide exposure of hepatocyte
cultures from metallothionein transgenic and control mice previously treated with zinc
and/or dexamethasone and I1-6 a'bc

Cell Viability (% of untreated cultures) ANOVA Results

Genotype TBH Control Zn Dex/Il-6 Zn &
L.tmol/L Dex/Il-6 p value


5 82 5 79 2
76 6 84 5


10 58 3x
48 1y


15 181
14 1


48 2x'y
33 2z


15 1
130


CTd
TG e


0.1038
0.8780
0.1500
0.2906
0.1553
0.7956
0.3624



0.0415
0.7802
0.0054
0.0007
0.0001
0.0001
0.0001


78 2 80 3
94 8 87 5
Genotype x Zinc x Dex/I1-6
Genotype x Zinc
Genotype x Dex/Il-6
Zinc x Dex/II-6
Genotype
Zinc
Dex/I1-6

53 3xy 48 2x'y
29 2z 31 Iz
Genotype x Zinc x Dex/Il-6
Genotype x Zinc
Genotype x Dex/I1-6
Zinc x Dex/Il-6
Genotype
Zinc
Dex/Il-6

212 161
140 130
Genotype x Zinc x Dex/Il-6
Genotype x Zinc
Genotype x Dex/Il-6
Zinc x Dex/I1-6
Genotype
Zinc
Dex/Il-6


a Cell viability values are means SEM, n = 9-11.
b Zinc treatment was 32 tmol/L for 20 h.
C Dexamethasone and 11-6 treatment (Dex/I1-6) was 1 gmol/L Dex and 100 units 11-6/mL
for 20 h.
d CT, hepatocyte cultures from control mice.
e TG, hepatocyte cultures from metallothionein transgenic mice.
X'y.z Means with different superscript values are significantly different.


0.5971
0.0204
0.0434
0.5886
0.0001
0.0021
0.1061








experiments we found that as metallothionein accumulation increased, cellular

glutathione decreased (Table 4). Since cysteine residues make up 33% of the total amino

acid content of metallothionein protein, induction of MT protein may consume a

significant amount of the cellular cysteine pool. Consequently, induction of

metallothionein by dex/Il1-6 treatment may have depleted glutathione levels and enhanced

TBH toxicity by competing for cysteine needed for GSH synthesis. Further, depletion of

cysteine pools might have inhibited the GSH regeneration during the toxic insult.

Alternatively, excess metallothionein may form mixed disulfides with oxidized

glutathione (GSSG) generated during TBH-induced oxidative stress (Brouwer et al. 1993,

Chai et al. 1994). Mixed disulfide formation may remove glutathione from its

regenerative pathway, delaying or preventing regeneration of reduced GSH needed for

antioxidant protection and other cellular processes (Gilbert 1995, Meister 1995).

The reduced sensitivity of KO cultures to TBH toxicity might also be explained if

there is a difference in the peroxidizability of cellular membranes between genotypes.

For instance, testes of zinc-deficient rats have a lower peroxidizable fatty acid level, and

are resistant to peroxidation in vitro (Oteiza and Keen 1996). If cells from KO mice act

similar to zinc-deficient tissues due to the lack of the intracellular ZnMT pool, they may

also have lower levels of the most peroxidizable fatty acids, and gain resistance to TBH

toxicity.

Several cell culture experiments have shown that treatment with metallothionein-

inducing agents is associated with protection against oxidants (Coppen et al. 1988,

Schroeder et al. 1990). Similar results were found in cells transfected with

metallothionein genes (Pitt et al. 1997, Schwarz et al. 1995) or cells from TG and KO








mice (Lazo et al. 1995, Rofe et al. 1998, Wang et al. 1999, Zheng et al. 1996). Of the two

studies that used primary hepatocyte cultures from KO mice, one showed only modest

protection against TBH (Zheng et al. 1996). The other report showed significant

protection by metallothionein expression against paracetamol toxicity, but only when

cultures were derived from fed mice; no differences in toxicity were seen between

genotypes in cultures from fasted mice (Rofe et al. 1998). Further, none of those studies

investigated whether inducing metallothionein prior to oxidant treatment affects toxicity.

Several other studies failed to find any protection by metallothionein expression (up to

166-fold) against the free radicals produced by x-radiation, bleomycin, or doxorubicin


Table 5-4. Cellular glutathione concentrations in hepatocyte cultures from metallothionein
knockout and control mice after treatment with zinc or dexamethasone and 11-6abc'

Cell glutathione (jjmol/g protein) ANOVA

Genotype Control Zn Dex/I1-6 Zn & Dex/Il-6 p value


CKd 23 +3 24+2 111 111
KOe 31 3 40 2 25 3 23 2

Genotype x Zinc x Dex/I1-6 0.1066
Genotype x Zinc 0.3360
Genotype x Dex/iIl-6 0.7097
Zinc x Dex/iIl-6 0.0605
Genotype 0.0001
Zinc 0.1832
Dex/Il-6 0.0001



a Cellular glutathione values are means SEM, n = 6.
b Zinc treatment was 32 imol/L for 20 h.
c Dexamethasone and 11-6 treatment (Dex/Il-6) was 1 pmol/L Dex and 100 units 11-6/mL
for 20 h.
d CK, hepatocyte cultures from control mice.
e KO, hepatocyte cultures from metallothionein knockout mice.








treatment (Kaina et al 1990, Kelley et al. 1988). Instead, they found protection to be more

consistent against alkylating agents. Our results contribute further evidence against a

cellular free radical scavenging function for metallothionein.

Supplemental zinc (32 tmol/L) did not protect against oxidative damage in these

experiments. In fact, zinc treatment resulted in a consistent increase in LDH leakage in

all genotypes (Fig. 5-2). Other studies using mouse hepatocytes also found no protection

by zinc treatment (50 gmol/L and 100 pmol/L) against free radical generators (Tezuka et

al. 1995, Rofe et al. 1998). This is in contrast to several studies using zinc-supplemented

(48 ptmol/L) rat hepatocytes (Coppen et al. 1988, Schroeder et al. 1990). Zinc induced

metallothionein in rat and mouse hepatocytes, but zinc only protected rat cultures. Since

rat hepatocytes were protected and mouse hepatocytes were not, we cannot rule out

species difference as a confounding variable in these experiments. For example, mouse

hepatocytes are significantly more sensitive to TBH-induced toxicity than rats, possibly

due to a greater peroxidizable lipid content of cellular membranes (Rush et al. 1985).

Since GSH is an important substrate for detoxification ofperoxidized lipids by

phospholipid glutathione peroxidase and glutathione transferase enzymes, hepatocytes

from mice may be more sensitive to glutathione depression than hepatocytes from rats if

lipid peroxidation is greater in mice after TBH treatment. It seems unlikely, however,

that chemical properties as basic as free radical scavenging (a simple oxidation-reduction

reaction) and zinc-thiol binding would differ between two rodent species. It should also

be noted that hepatocytes from two different mouse strains were used in our experiments.

As such, it is unlikely that the results presented here were due to peculiarities of any

particular mouse strain. Further, levels of other antioxidants do not differ among these








mouse genotypes (Iszard et al. 1995, Lazo et al. 1995). Instead, these results show that

supplemental zinc does not provide consistent protection against oxidative stress, and in

this case exacerbated toxicity.

Another interesting result was that dex/Il -6 treatment, which protected rat

hepatocytes cultures from oxidative stress (Schroeder et al. 1990), enhanced

susceptibility to TBH in both mouse strains in this experiment. The enhanced

susceptibility was closely related to the magnitude of metallothionein expressed after that

treatment, and was not altered by zinc-treatment. This is in opposition to the report of

Rofe and coworkers (1998), who found that dexamethasone (1 pmol/L) protected CK

mouse hepatocytes against paracetamol toxicity when administered at the same time as

paracetamol. KO hepatocytes were not protected by dexamethasone treatment,

suggesting that metallothionein production was necessary for that effect. Metallothionein

was not induced before paracetamol exposure, however, which may explain why their

results were different than ours. In support of this, glutathione levels were similar

between genotypes before paracetamol exposure in their experiments. They also found

that supplemental zinc nearly tripled metallothionein levels in dexamethasone treated CK

cultures, yet inhibited dexamethasone-mediated protection against paracetamol (Rofe et

al. 1998). This suggests that overproduction of metallothionein may be counterproductive

even when metallothionein induction occurs during paracetamol exposure.

In conclusion, these results argue that supplemental zinc and/or metallothionein

do not protect hepatocyte cultures against tertiary-butyl hydroperoxide-induced oxidative

stress, but rather enhance the toxicity. Further, preinduction of metallothionein enhanced

the toxicity. Further research is required to determine the mechanism involved, but it is






84


likely related to depressed glutathione levels in hepatocytes after metallothionein

induction. If so, induction of metallothionein prior to oxidant exposure may be

counterproductive if protection against that oxidant relies on glutathione availability.














CHAPTER 6
CONCLUSIONS


The research reports bound together in this dissertation provide new insights into

the biochemical and physiological functions of metallothionein and zinc. The foci of this

research fall into two categories:

(1) the effects of metallothionein expression and dietary zinc intake in zinc

absorption, distribution, and intracellular zinc trafficking in TG and KO mice,

and

(2) the effects of supplemental zinc, metallothionein expression, and their

combination in defense against oxidative stress.

The discussion that follows brings together data from the four separate research

reports in an effort to provide a unified model of the biological roles of metallothionein

and zinc, and their interaction.


Zinc and Metallothionein in Zinc Absorption and Metabolism


Metallothionein Expression and Zinc Absorption

The results outlined in Chapter 2 provide further evidence for the theory that

induction of intestinal metallothionein is a least part of a mechanism for controlling the

flux of zinc from the intestinal lumen to the general circulation. Specifically, serum zinc

was elevated to a greater level in mice with lower levels of metallothionein expression

after an acute oral zinc dose (TG < CT & CK < KO). Similar results were found in a









separate metallothionein knockout strain over a range from normal to supplemental zinc

intake (Coyle et al. 1999, Coyle et al. 2000). Contrary to our hypothesis, however,

intestinal metallothionein induction did not result in greater zinc accumulation in the

intestine, suggesting that metallothionein's role is not simply to sequester zinc in the

intestine. It may be that metallothionein acts to suppress zinc absorption by enhancing

zinc flux back toward the lumen, as suggested by Hoadley and coworkers (1988). It

should be stressed, however, that the effects of metallothionein on zinc absorption seen in

Chapter 2 can be interpreted only with reference to an acute oral dose of zinc.


Metallothionein Expression and Zinc Metabolism

The results of Chapters 3 through 5 provide insight into the overall impact of

metallothionein expression on zinc distribution, accumulation and trafficking at the

cellular and whole body levels. Data from Chapters 3 and 4 (only 0 h data from Chapter

4) display the effects of dietary zinc intake (ranging from 10 mg/kg diet to 500 mg/kg

diet) on metallothionein expression. Conversely, we also saw how the level of

metallothionein expression affects intestinal, hepatic and serum zinc concentrations.

Dietary zinc induced intestinal and hepatic metallothionein protein expression, but highly

supplemental dietary zinc intakes were required (> 200 mg/kg diet). At lower intakes

metallothionein was undetectable. Further, metallothionein was overexpressed in TG

mouse liver at all zinc intakes, yet did not affect hepatic zinc accumulation until diets

containing thirty-fold the requirement were consumed. At 500 mg Zn/kg diet, however,

hepatic zinc accumulation was directly related to hepatic metallothionein induction in all

genotypes. These results are directly in line with the role proposed for metallothionein in

zinc metabolism; i.e., metallothionein is required for zinc accumulation in tissues.









However, within the range of dietary zinc intakes that we studied, metallothionein

expression only altered tissue zinc accumulation under conditions of dietary zinc excess.

The situation might be very different under conditions of stress and zinc deficiency. For

example, Philcox and coworkers (2000) reported that metallothionein expression inhibits

intestinal zinc loss during stress (starvation and immune stress associated with

lipopolysaccharide injection), and prevents body zinc loss during zinc deficiency.

The results of Chapter 4 also show that hepatic zinc accumulation during stress is

dependent on metallothionein expression. This had been shown in a separate KO mouse

model after immune stress (Philcox et al. 1995). Our results extend this finding to the

level of metallothionein expression observed in TG mice consuming supplemental zinc

and subsequently treated with carbon tetrachloride. We also report that hepatic

metallothionein expression after exposure to stress was, in part, dependent upon the level

of zinc consumed in the diet. In Chapter 5 we report that intracellular zinc accumulation

in cultured hepatocytes after exposure to excess extracellular zinc, dexamethasone and

interleukin-6, and their combination was also dependent on the level of metallothionein

expression. This was also reported recently in another KO mouse model (Coyle et al.

1995).

Evidence was found implicating metallothionein in autoreguolation of its own

gene expression (Chapter 4). Metallothionein expression was stimulated at lower dietary

zinc intakes in mice with more copies of the metallothionein gene (TG > CT and CK >

KO). This finding could have larger implications, as it suggests that metallothionein

expression may also regulate expression of other zinc-responsive genes (Fig. 6-1). At this

time we speculate that metallothionein acts to regulate gene expression through one of








two mechanisms: metallothionein may act as a labile zinc pool that can release zinc for

use in zinc finger transcription factors (e.g., metal response transcription factor-1) that act

in upregulation of zinc-responsive genes, or metallothionein may directly donate zinc to

such proteins (Maret 2000, Zeng et al. 1991). Either way, these results suggest that

metallothionein may be involved in zinc metabolism at dietary zinc levels that do not

cause detectable changes in tissue zinc accumulation.

Before leaving this subject, it should be noted that although KO mice have altered

zinc metabolism, this is only seen under conditions of stress or supplemental zinc intake

(Chapters 2-5). For example, serum zinc levels are maintained accordingly in KO mice





SapoMTF-1 ^^
/ V ^ ~MT gene ^\

-MTF-1
r^~~Free 'V MT mRNA +/
Zn


\~nM ^- apoMT B ^ MTF-1



^^^^^ apoMTF-1 I



Figure 6-1. Autoregulation of metallothionein gene expression. Metallothionein might
regulate expression of its own gene through one of two pathways: (A) zinc released from
ZnMT enhances the availability of free zinc for incorporation into transcription factors,
such as MTF-1, and/or (B) ZnMT directly donates zinc to MTF-1. Either scenario might
promote the DNA binding activity of MTF- 1 by activating a key zinc finger within this
transcription factor protein.









allowed to acclimate for several days to diets ranging in zinc content from 10 to 200

mg/kg. In fact, serum zinc concentrations were lower in KO mice than CK mice

consuming diets with 300 and 500 mg Zn/kg. Obviously, KO mice can compensate for

the loss of metallothionein expression over a wide range of zinc intakes. Altered

expression of one or more zinc transporters may explain this adaptation response in KO

mice (Liuzzi et al. 2000, McMahon and Cousins 1998).


Zinc and Metallothionein in Defense Against Oxidative Stress


In Vivo

Metallothionein expression protected CK mice against carbon tetrachloride-

toxicity, confirming metallothionein's role in protection against oxidative stress.

Metallothionein overexpression provided no further protection, however, with or without

added dietary zinc. Supplemental zinc alone was not protective either. These results are

inconsistent with a general (direct) antioxidant role for either zinc or metallothionein.

Instead, the result may be more indicative of the general need for rapid zinc accumulation

in the liver; a process that depends on metallothionein expression (chapters 3-5). We

speculate that the KO mice are more susceptible to oxidative stress because they lack the

capacity for rapid and sustained hepatic zinc accumulation. This zinc may be required for

numerous proteins (enzymes and transcription factors) involved in gene expression, as

well as other proteins involved in cellular metabolism. Since excess dietary zinc and

elevated metallothionein expression provide no additional protection, it may be that

adequate zinc intake and "normal" metallothionein expression are all that are needed to

maintain these processes. The liver may not be capable of synthesizing proteins quickly









enough to take advantage of the excess intracellular zinc provided via the diet or elevated

metallothionein levels. Alternatively, greater production of these proteins may not be

helpful.


In Vitro

Cytotoxicity studies in primary hepatocyte cultures from KO and TG mice were

undertaken to determine the protective roles of zinc and metallothionein against oxidative

stress in a simpler, more easily manipulated model. We found that induction of

metallothionein prior to tertiary-butyl hydroperoxide-treatment resulted in heightened

cytotoxicity; a result exactly opposite of that hypothesized, and opposite of what had

been reported previously with other oxidants. This relationship between metallothionein

expression and cytotoxicity was true whether metallothionein was elevated due to

genotype, zinc, dexamethasone and interleukin-6, and any combination of these factors.

Further, the magnitude of cytotoxicity was proportional to the magnitude of

metallothionein protein induced.

We found that metallothionein expression was inversely related to cellular

glutathione levels. Since the first line of defense against the toxicity of tertiary-butyl

hydroperoxide is the enzyme glutathione peroxidase, which relies on glutathione for

reducing equivalents, we reason that the depletion of glutathione is responsible for the

enhanced cytotoxicity observed with metallothionein induction. Not all of the effect of

dexamethasone and interleukin-6 treatment on glutathione was dependent on

metallothionein expression, however, since glutathione was also reduced to some extent

by this treatment in cultures from KO mice as well.








At least two hypotheses have been generated to explain this phenomenon (Fig. 6-

2). The first hypothesis (the author's) is that metallothionein induction depletes the

available cellular cysteine pools. Evidence for this includes the fact that cysteine residues

make up 1/3 of the total amino acid residues of both metallothionein and glutathione.

Also, the absolute cysteine content of the metallothionein pool (after induction of

metallothionein by the various treatments used in this experiment) is comparable to the

cysteine content of the glutathione pool. Further, the glutathione content of primary

hepatocyte cultures is highly dependent on the available cysteine content of the culture

medium (Wang et al. 1997).

The second hypothesis (of Dr. Cousins) is that the metallothionein induced by

various treatments physically interacts with glutathione, and thereby interferes with the


MT


Cysteine




TBOOH GSH 4 w NADP


TBOH ;S C GSSG IsH : NADPH


MT

Figure 6-2. Mechanisms by which metallothionein might cause glutathione depletion.
Metallothionein might interfere with glutathione regeneration through physical
interactions with oxidized glutathione. Alternatively, induction of metallothionein
might consume enough cysteine so as to interfere with glutathione synthesis.









use and/or availablility of glutathione as a substrate for glutathione peroxidase. Evidence

for this viewpoint includes the fact that glutathione forms mixed disulfides with a number

of intracellular sulfhydryl-containing proteins, as well as the fact that several putative

glutathione binding sites have been reported in the metallothionein molecule (Brouwer et

al. 1993, Maret 2000). This hypothesis includes observations from in vitro experiments

where glutathione peroxidase and reduced glutathione release zinc from metallothionein

and cellular oxidants cause release of cellular glutathione (oxidized and reduced). There

is no reason to believe that the mechanisms outlined in these hypotheses would be

mutually exclusive. If metallothionein does affect hepatic glutathione levels in vivo, this

relationship might explain hepatic glutathione depression seen after treatment with a

certain hormones, or under some stress conditions.

In summary, the findings of these studies agree with the previously proposed

theories regarding the role of metallothionein in zinc metabolism. The results herein

refine this model by constraining it to conditions of highly excessive zinc intake. A more

novel finding was that metallothionein may regulate its own expression by providing a

labile intracellular zinc pool that may feedback positively on metallothionein gene

expression.

Metallothionein expression protected against oxidative stress in vivo, but

metallothionein overexpression and supplemental dietary zinc provided no further

protection. We interpret this to mean that there is a threshold of zinc intake and

metallothionein expression that is protective, but above which is not helpful. Normal

metallothionein expression combined with the required zinc intake meet this threshold

value in mice. These results are not consistent with direct antioxidant roles for