..____

LIEMRY
Michigan State
University

 

 

————

OVERDUE FINES:
25¢ per day per ite-

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from c1 rcuht1on records

 

 

 

BRAIN IRON'IN THE RAT:
DISTRIBUTION, SEX DIFFERENCES, AND EFFECTS OF SEX HORMONES

By

Joanna Marie Hill

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology

1981

J
.r n

’7
r’

I ,l'
.3

6//

ABSTRACT
BRAIN IRON IN THE RAT:

DISTRIBUTION, SEX DIFFERENCES,
AND EFFECTS OF SEX HORMONES

By

Joanna Marie Hill

Although the brain contains relatively large amounts of
iron, and iron deficiency alters behavior, little is known
about those factors which affect brain iron or the role of
iron in the brain. Sex hormones are responsible for sex
differences in many aspects of iron metabolism throughout
the body.

The purposes of this study were to: (l) localize iron
deposits in the rat brain; (2) determine if a sex difference
exists in brain iron stores; (3) determine the effects on
brain iron levels of natural events in which sex hormones
fluctuate (e.g. estrous cycle and pregnancy); and (4)
determine if exogenous estrogen alters the effects of ovari-
ectomy and castration on brain iron levels.

Brain iron was localized by histochemical methods and
direct measurement of iron concentrations of high-iron areas
(pooled globus pallidus and substantia nigra) and lower iron
areas (cortex) of the brain, as well as the serum and liver
were made by spectr0photometry.

This study has determined that brain iron: is unevenly
distributed in the rat brain; occurs in different cellular

and extracellular compartments in different parts of the

Joanna Marie Hill

brain; and increases with age. Brain iron fluctuates

during the estrous cycle, rising to the highest levels
during proestrus. During the first third of pregnancy
brain iron concentration rises and although the level falls
later, pregnancy does not deplete brain iron. There is a
sex difference in the brain iron concentration between males
and females in the proestrus stage of the estrous cycle,

and ovariectomy and castration have different effects on
brain iron levels.

The results of this study suggest that: the pattern of
iron distribution may be related to the participation of iron
in the metabolism of peptides; brain iron accumulation is
influenced by ovarian hormones; and iron plays a role in

neuroendocrine regulation.

ACKNOWLEDGEMENTS

Grateful acknowledgment is made to my coadvisers
Dr. J.I. Johnson and Dr. R.C. Switzer for their interest,
guidance and constructive criticisms throughout all
stages of this study. Also, the helpful comments of my
committee members Dr. M. Balaban, Dr. C.D. Tweedle and
Dr. S.T. Kitai are sincerely appreciated.

It is a pleasure to acknowledge Dr. P.D. MacLean,
Chief, Laboratory of Brain Evolution and Behavior, NIMH
not only for his enthusiasitc support but also for the
laboratory space, equipment and supplies he made available
to me.

I wish to thank Mrs. J. Bupp for her editorial com-
ments and competent typing of the dissertation and Mr. R.
Harbaugh for invaluable technical assistance with surgical
procedures and the care and breeding of animals.

I thank also my husband, Jim for his assistance with
the statistical analysis and for the patience, understand-
ing and support I received from him and my daughters,
Andrea and Katherine, without which this study would not

have been possible.

ii

TABLE OF CONTENTS

 

PAGE
LIST OF TABLES ........................................ iv
LIST OF FIGURES ........................................ v
INTRODUCTION ........................................... 1
LITERATURE REVIEW ...................................... 4
MATERIALS AND METHODS ................................. 37
HISTOCHEMISTRY ...................................... 37
Animals ............................................ 37
Treatment Groups ................................... 37
Preparation of Tissue .............................. 40
Staining techniques ................................ 43
Analysis of Data ........... , ....................... 49
SPECTROPHOTOMETRY ................................... 50
Animals ............................................ 50
Treatment Groups ................................... 50
Collection of tissue for the Spectrophotometric
Measurement of Iron ................................ 59
Requirements for the Spectrophotometric

Measurement of Iron ................................ 61
Solutions for Iron Spectrophotometry ............... 63
Determination of Serum Iron ........................ 65
Preparation of Liver Extract ....................... 66
Determination of Liver Iron ........................ '68
Preparation of Brain Extract ....................... 69
Determination of Brain Iron ........................ 72
Analysis of Data ................................... 74
RESULTS ............................................... 75
HISTOCHEMISTRY ...................................... 75
SPECTROPHOTOMETRY .................................. 108
DISCUSSION ........................................... 134
SUMMARY .............................................. ‘157
BIBLIOGRAPHY ........................................ 160

iii

TABLE

LIST OF TABLES

PAGE

The Distribution of Non-haemin Iron in
Different Parts of the Human Brain Autopsy
Cases, 30-100 Years of Age ....................... 17

Determination of Iron in the Subcellular
Fractions of Different Brain Areas ............... 27

Summary of Iron Measurement Study ................ 58

Analysis of Variance and Tukey's Test of Iron
Measures of Estrous Cycle and Sex Difference
Data ............................................ 111

Analysis of Variance with Regression of Iron
Measures of Pregnancy Data ...................... 116

Analysis of Variance with Orthogonal Contrasts

of Iron Measures of Control Males (CM), Cast-

rated Males (CAST), Castrated Males with

Estrogen Implants (CAST+EST), Intact Females

in Estrus (ESTF), Ovariectomized Females (OVX)

and Ovariectomized Females with Estrogen

Implants (0VX+EST) .............................. 122

The Distribution of Monoamines and GABA in Iron
Concentrating Areas of the Rat Brain ............ 136

The Distribution of Peptides and Iron Concent-
rating Areas of the Rat Brain ................... 139

iv

LIST OF FIGURES

FIGURE

1.

Distribution of iron. Perl's-DAB stain for
iron, no counterstain. Stained areas are
accumulations of iron. Parasagittal view
of a 52 week old female rat. Magnification

X 4.5 ........................................

Distribution of iron in the forebrain at
the level of the anterior commissure in a
32 week old female rat. Perl's-DAB, no
counterstain. Stained areas are accumul-

ations of iron. Magnification X 5.4 .........

Distribution of iron in the forebrain at
the level of the anterior commissure in a
32 week old male rat. Perl's-DAB, no
counterstain. Stained areas are accumula—

tions of iron. Magnification X 5.4 ..........

Distribution of iron in the forebrain at the
level of the globus pallidus in a 32 week
old female rat. Perl's—DAB, no counter-
stain. Stained areas are accumulations of

iron. Magnification X 5.4 ...................

Distribution of iron in the forebrain at the
level of the globus pallidus in a 32 week
old male rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4 ..........................

Distribution of iron in the forebrain at the
level of the thalamus in a 32 week old
female rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4 ..........................

Distribution of iron in the midbrain at the
level of the substantia nigra in a 32 week

old female rat. Perl's-DAB, no counterstain.

Stained areas are accumulations of iron.

Magnification X 5.4 ..........................

V

PAGE

..76

.78

..78

..78

..78

..80

..80

LIST OF FIGURES -— continued

FIGURE PAGE

10.

ll.

12.

13.

14.

Distribution of iron in the midbrain at the
level of the substantia nigra in a 32 week
old male rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4 ............................. 80.

Photomicrograph of iron accumulation in area
postrema and dorsal to the central canal

Darker staining areas are accumulations of iron.
Perl's-DAB counterstained with thionin.
Magnification X 100 ............................. 83

Photomicrograph of iron accumulation in the
subfornical organ and choroid plexus.

Darker staining areas are accumulations of iron.
Perl's-DAB counterstained with thionin.
Magnification X 100 ............................. 83

Photomicrograph of iron-filled fibers in the
lateral edge of the optic tract. Iron = brown.
Perl's-DAB counterstained with thionin.
Magnification X 200 ............................. 86

Dark-field photomicrograph of iron—filled
tanycytes and clumps of granules in the
ventro-medial hypothalamus. Iron = white.
Perl's-DAB counterstained with thionin.
Magnification X 120 ............................. 86

Dark-field photomicrograph of iron-filled
structures in the ventro-medial hypothalamus

and arcuate area. Iron = white. Perl's-DAB
counterstained with thionin. Magnification

X 120 ........................................... 88

Dark-field photomicrograph of iron—filled
structures in the ventro-medial hypothalamus

and median eminence. Iron = white. Perl's-DAB
counterstained with thionin. Magnification

X 120 ........................................... 88

vi

LIST OF FIGURES -- continued

FIGURE PAGE

15.

l6.

l7.

l8.

19.

20.

21.

22.

Photomicrograph of iron distribution in the
ventral pallidum, islands of Calleja and

olfactory tubercle. Iron = brown. Perl's-DAB
counterstained with thionin. Coronal view.
Magnification X 82 .............................. 90

Photomicrograph of iron-filled cells and

fibers in the globus pallidus. Iron = brown.
Perl's-DAB counterstained with thionin.
Magnification X 900 ............................. 90

Photomicrograph of iron-filled cells and

fibers in the ventral pallidum. Iron = brown.
Perl's-DAB counterstained with thionin.
Magnification X 476 ............................. 93

Photomicrograph of iron-filled cells and
fibers in the substantia nigra. Iron = brown.
Perl's-DAB counterstained with thionin.
Magnification X 476 ............................. 93.

Photomicrograph of small neuron-like iron-

filled cells in the globus pallidus.

Iron = brown. Perl's-DAB counterstained with
thionin. Magnification X 900 ................... 95

Dark-field photomicrograph of bundles of fine
iron-filled fibers in the globus pallidus.

Iron = white. Perl's-DAB counterstained with
thionin. Magnification X 1000 .................. 95

Photomicrograph of "cascades" of iron—filled
fibers seen in the ventral pallidum and

olfactory tubercle. Iron a brown. Perl's-DAB
counterstained with thionin. Magnification 9
X 476 ........................................... 7

Dark-Field photomicrograph of iron in bouton-
like structures on or in lateral septal neurons.
Iron a white. Perl's-DAB counterstained with
thionin. Magnification X 226 ................... 97.

vii

LIST OF FIGURES -- continued

FIGURE PAGE

23.

24.

25.

26.

27.

28.

29.

Dark-field photomicrograph of iron in

bouton-like structures on or in bed nucleus

of the stria terminalis neurons. Iron = white.
Perl's-DAB counterstained with thionin.
Magnification X 226 ........................... 100

Photomicrograph of iron in bouton-like

structures on or in ventral pallidal neurons.
Iron = brown. Perl's-DAB counterstained

with thionin. Magnification X 900 ............ 100

Dark-field photomicrograph of fine grains of

iron in cells of the supraOptic nucleus.

Iron = white. Perl's-DAB counterstained with
thionin. Magnification X 216 ................. 103

Dark-field photomicrograph of fine grains of
iron in cells of the suprachiasmatic nucleus

Iron = white. Perl's-DAB counterstained with
thionin. Magnification X 200 ................. 103

Iron concentration of liver, pooled globus
pallidus and substantia nigra, cortex and

serum of control males and females in each

of the four stages of the estrous cycle.

Mean values i l S.E.M., see ANOVA tables

for sample sizes .............................. 110

Iron concentration of liver, pooled globus
pallidus and substantia nigra, cortex and

serum every four days throughout pregnancy

and 15-24 hours postpartum. Mean values i

l S.E.M., see ANOVA tables for sample sizes...115

Iron concentration of pooled globus pallidus

and substantia nigra of control males (CONT d),
castrated males (CAST 6), castrated males

with estrogen implants (CAST 6+EST), females

in estrus (ESTRUS 9), ovariectomized females

(OVX Q), and ovariectomized females with

estrogen implants (OVX o+EST). Mean valuesi

l S.E.M., see ANOVA tables for sample sizes...121

viii

LIST OF FIGURES -- continued

FIGURE PAGE

30.

31.

32.

33.

Iron concentration of cortex of control males

(CONT 6), castrated males (CAST 6), castrated
males with estrogen implants (CAST 6+EST),

females in estrus (ESTRUS 9), ovariectomized
females (OVX g), and ovariectomized females

with estrogen implants (OVX 9+EST). Mean

values i l S.E.M., see ANOVA tables for sample
sizes ........................................... 125

Iron concentration of liver of control males

(CONT 6), castrated males (CAST 6), castrated

males with estrogen implants (CAST 6+EST),

females in estrus (ESTRUS 9), ovariectomized
females (OVX ), and ovariectomized females

with estrogen implants (OVX 9+EST). Mean

values $.1 S.E.M., see ANOVA tables for sample
sizes ........................................... 128

Iron concentration of serum of control males

(CONT 6), castrated males (CAST 6), castrated
males with estrogen implants (CAST 64EST),

females in estrus (ESTRUS 9), ovariectomized
females (OVX 9), and ovariectomized females

with estrogen implants (OVX 9+EST). Mean
values.i.l S.E.M., see ANOVA tables for sample
sizes ........................................... 130

Body weight of control males (CONT 6), castrated
males (CAST 6), castrated males with estrogen
implants (CAST 64EST), females in estrus
(ESTRUS 9), ovariectomized females (OVX a),
ovariestomized females with estrogen imp ants

(OVX 9+EST). Mean values i 1 S.E.M., see

ANOVA tables for sample sizes ................... 133

ix

INTRODUCTION

Iron deficiency is the most prevalent nutritional
disorder faced by the p0pulations not only of "under-
developed" nations but also of nations, such as the United
States, which enjoy a varied and abundant food supply (WHO,
1968; Ten State Nutritional Survey, 1970; Nutrition Canada,
1973; Garby, 1973; Kessner and Kolk, 1973; HANES, 1974). An
excessive accumulation of iron in the body tissues also
afflicts a significant fraction of the human p0pulation
(Sheldon, 1927; MacDonald, 1964; Pollycove, 1972). Abnormal
iron levels are accompanied by a variety of both physical
and behavioral disorders (see Pollitt and Leibel, 1976;
Oski, 1979; Dooling et al., 1974; Goldberg and Allen, 1980).

It is known that iron is stored in relatively large
amounts in the brain, chiefly in the basal ganglia (wuth,
1923; Tingey, 1937; Cumings, 1948, 1968; Hallgren and
Sourander, 1958; Musil, Haas and Waurschinek, 1962;
Harrison, Netsky and Brown, 1968) and that iron deficiency
significantly alters the behavior of man and animals (Glover
and Jacobs, 1972; Pollitt and Leibel, 1976; Oski, 1979;
Weinberg, Levine and Dallman, 1979; Williamson and Ng,
1980). However, there has been little work on those factors
which directly affect iron metabolism of the brain either in
iron deficiency or in the iron replete condition.

Sex hormones influence behavior, and these hormones,

chiefly estrogen, are responsible for a sex difference in

l

2

iron metabolism with effects on the uptake, levels of
storage, and the mobilization of iron (Steenbock, Semb and
Van Donk, 1936; Smith and Otis, 1937, Otis and Smith, 1940;
Widdowson and McCance, 1948; Kaldor and Powell, 1957;
Dubrunquez and Lederer, 1963; Murray and Stein, 1968;
Bj¢rk1id and Helgeland, 1970; Linder et a1., 1973; Cook,
Hershko and Finch, 1973; Hershko and Eilon, 1974; Planas,
1976). Sex hormones are known to act directly on hormone-
sensitive areas of the brain regulating the secretion of
gonadotrophins and thereby influencing behavior (see reviews
by Young, 1961; Flerko, 1966; Lisk and Barfield, 1975;
McEwen, 1981). However, neither a sex difference in brain
iron levels nor any effect of fluctuating levels of sex
hormones on brain iron levels has been reported. Deter-
mining if a sex hormone such as estrogen influences brain
iron levels will provide suggestive evidence that this
hormone plays a role in brain iron metabolism. Secondarily,
this evidence will suggest either that brain iron is influ-
enced by an internal environmental factor, related to sex
difference, which constantly affects other iron storage
areas, or that the brain is "spared" from such a sex dif-
ference. Also, the potential that the availability of iron
may fluctuate in the brain tissues in response to changing
hormone levels, as in the female reproductive cycle, has
significance in the study of any aspect of brain iron

metabolism.

3

The purpose of this study is to: (1) localize iron
deposits in the rat brain, (2) determine if a sex difference
exists in brain iron stores both quantitatively and in
location, (3) determine the effects on brain iron levels of
natural events in which sex hormones fluctuate, e.g.,
estrous cycle and pregnancy, and (4) determine if exogenous
estrogen alters the effects of ovariectomy and castration on
brain iron levels.

Both histochemical and spectrOphotometric determination
of brain iron are made. Not only does the histochemical
identification of iron in the brain function to localize
iron deposits, but also, the intensity of the reaction can
be used as an indication of the amount of iron present in
different brain areas. Spectrophotometry provides an
accurate quantitative measure of iron, and the tissues
analyzed by this method include the serum, liver, high iron
areas of the brain (pooled globus pallidus and substantia
nigra), and a lower iron area (cortex). The spectrophoto-
metric determination of serum and liver iron along with
brain iron serves two purposes. It indicates not only those
animals which may have anemia or some defect in iron metab-
olism and should therefore be excluded from the study, but
also demonstrates whether the experimental manipulations
influence iron levels in the brain in the same direction and

magnitude as occurs in the serum and liver.

LITERATURE REVIEW

Three hundred years ago, Syndenham recognized the use
of iron as a specific remedy for chlorosis (Fairbanks et a1,
1971), now recognized as iron deficiency anemia, a common
disorder among young women of menstrual and child—bearing
age. Iron was thus the first trace element known to be
essential. Since that time, the study of iron metabolism
has been an active field of research. Much of the data
has accumulated from the study of man as many aspects of
iron exchange can be studied from blood and small samples
of liver and bone marrow which can be collected without
lethal effects. Next to man, the laboratory rat has been
the animal upon which most iron metabolism research has
been centered.

Despite the vast amount of research in the field of
iron metabolism, in some areas our knowledge is incomplete:

- The mechanism of absorption of iron from food in the
intestine is not completely understood. There is some dis-
agreement as to the way in which the intestine regulates
uptake of iron and what factors are important in the bio-
availability of iron in food.

- The mechanisms of iron transfer from blood to cell,
cell to cell, and between compartments within a cell are
unclear. The realtive importance of various oxidizing and
reducing agents in this process is disputed.

- Lastly, a question deserving more attention: what is
4

5
the function of the large amount of iron stored in the
brain--what is its role in brain metabolism and behavior?
A review of the literature specifically dealing with

iron in the brain will follow a brief outline of iron

metabolism in mammals.

Iron in Biological Tissues

Iron has two special properties which will make it a
vital constituent of every mammalian cell as well as an
essential element in biological processes throughout nature:
iron can exist in two stable oxidation states (Fe++, Fe+++),
and it can form many complexes. In biological systems,
almost all iron is involved in processes related to oxygen
metabolism.

Iron is present in tissues in two forms, heme and non-
heme compounds. The heme compounds include hemoglobin (O2
carrier), myoglobin (O2 carrier), the heme enzymes catalase
and peroxidase (protect cells from the harmful by-products
of oxygen, superoxide ion and hydrogen peroxide), the cyto-
chromes (electron transport), cytochrome oxidase (terminal
oxidase), xanthene oxidase (purine metabolism and iron
mobilization) and tryptophan dehydrogenase (tryptophan
oxidation). The non-heme compounds include transferrin
(iron transport), ferritin (iron storage), hemosiderin (iron
storage), aldehyde oxidase (indole amine degradation),
tyrosine dehydrogenase (catecholamine synthesis), succinic
dehydrogenase (Kreb's citric acid cycle), NADH-dehydrogen-
ase, monoamine oxidase, ribonucleotide reductase, aconitase,
aldolase, lipoxidase, and others (Freiden, 1974; Subcommit-
tee on Iron, 1979; Yasunobo, Mower, and Hayaishi, 1975).

In an iron replete man, iron is distributed in the

following way:

65% total body iron in hemoglobin

3% total body iron in myoglobin

13% total iron in ferritin

12% total iron in hemosiderin

<l% total body iron in transferrin

<l% total body iron in known iron enzymes

5% total body iron in unknown compounds
(taken from Frieden and Osaki, 1974)

At birth, the total body iron concentration of a mammal
is increased by the elevated red blood cell mass and the
presence of appreciable storage iron in the liver; however,
during suckling, body iron reserves are depleted (Lintzel,
Rickenberger and Schairer, 1944; Widdowson and McCance,
1948, 1951; Keen and Hurley, 1980). At weaning the iron
stores begin to accumulate and by 16 weeks of age, the rat
has a threefold increase in liver iron (Cook, Hershko and
Finch, 1973). The rate of accumulation of iron begins to
slow at about 32 weeks (Kaldor and Powell, 1957) and a low
rate of increase is maintained throughout life (Bjorklid and

Helgeland, 1970: Kaldor and Powell, 1957; Widdowson and
McCance, 1948).

Iron Exchange

The general pathways of iron exhange are well known.
External exchange occurs largely through the intestinal
tract. Iron attached to chelates is absorbed by the mucosa
of the duodenum and jejunum, where it is trapped by iron
binding proteins and made available to the transfer system
(Gitlin and Cruchaud, 1962; Charlton et a1., 1965). The
binding capacity of the protein is regulated by the iron
needs of the individual. Some absorption also occurs by
diffusion, governed by the amount of iron in the gut lumen
(Thompson and Valberg, 1980). Only a small portion of
dietary iron is available to the mucosa, as much intestinal
iron forms polymerized iron hydroxide which is not absorb-
able (Huebers and Rummel, 1975; Thompson and Valberg, 1980).
In rats, the dietary intake of iron is about 100 times that
of man (when expressed on a per kilogram basis), making
absorption the primary route for meeting increased iron
requirements (Cook, Herschko and Finch, 1973).

Iron excretion occurs primarily through the feces by
exfoliation of the intestinal mucosa. In man, very little
iron is lost this way (McCance and Widdowson, 1937);
however, in the rat this represents an important route of
excess iron excretion (Cheney et a1., 1967).

Within the body, iron is transported by plasma trans-
ferrin. About 90% of the body iron is in the blood, bone
marrow, liver, and Spleen (the erythron). The group next

highest in iron concentration includes the kidney, heart,

9

skeletal muscles, pancreas, and brain (Cheney et a1., 1967;
Keen and Hurley, 1980). Newly absorbed iron is almost
immediately taken up by bone marrow where it is incorporated
into heme in the formation of red blood cells which soon
after appear in the blood.

In the reticuloendothelial cells, primarily in the
spleen, iron from effete erythrocytes is released, stored in
ferritin or hemosiderin, or transported by transferrin to be
used in the formation of new red blood cells, to the liver
to be stored in hepatocytes or to other body tissues to meet
their individual needs (Cheney et a1., 1967; Cook, Hershko
and Finch, 1973). In man, with a limited capacity for
absorption and slight loss through excretion, much of the
iron for heme formation comes from the recycling of iron
released through the breakdown of senescent red blood cells
(Noyes et a1., 1964). In the rat, however, absorption and
excretion play a greater part in maintaining iron balance
(Cheney et a1., 1967; Cook, Hershko and Finch, 1973) than in

man .

10

Sex Differences in Iron Metabolism

Sex differences brought about by gonadal hormones are
apparent in several aspects of iron metabolism. As first
reported by Steenbock et a1. (1936) and later confirmed by
others (Widdowson and McCance, 1948; Kaldor and Powell,
1957; Bjorklid and Helgeland, 1970; Cook, Hershko and
Finch, 1973; Linder et a1., 1973; Hershko and Eilon, 1974),
female rats have liver iron stores two to three times
greater than male rats. A similar sex difference also
occurs in the total body iron (Otis and Smith, 1940; Kaldor
and Powell, 1957), spleen iron (Widdowson and McCance, 1948;
Kaldor and Powell, 1957), kidney iron (Linder et a1., 1973),
serum iron (Cook, Hershko and Finch, 1973; Hershko and
Eilon, 1974), but not heart iron (Linder et a1., 1973).

Female rats have more liver ferritin (iron storage
protein), more apoferritin, a faster rate of liver ferritin
turnover (Bjorklid and Helgeland, 1970; Linder and Munro,
1973; Linder et a1., 1973; Hershko and Eilon, 1974), greater
liver iron exchange (Hershko and Eilon, 1974), absorb more
iron from their diet (Otis and Smith, 1940; Murray and
Stein, 1968; Cook, Hershko and Finch, 1973: Hershko and
Eilon, 1974), have a greater plasma iron turnover, and a
larger percentage saturation of transferrin (Hershko and
Eilon, 1974) than male rats.

Iron stores are low at weaning and the above sex dif-
ferences begin to emerge, under the influence of sex

hormones, only after sexual maturity at 6 to 7 weeks of age

ll

(Widdowson and McCance, 1948; Kaldor and Powell, 1957;
Bjorklid and Helgeland, 1970; Cook, Hershko and Finch,
1973).

Ovariectomy causes reductions in liver iron (Steenbock
et a1., 1936; Widdowson and McCance, 1948; Dubrunquez and
Lederer, 1963; Hershko and Eilon, 1974), plasma iron turn-
over, and serum iron (Hershko and Eilon, 1974). Castration
results in increases in liver iron (Widdowson and McCance,
1948; Dubrunquez and Lederer, 1963; Hershko and Eilon,
1974), plasma iron turnover, and serum iron (Hershko and
Eilon, 1974). The replacement of the appropriate sex
hormones to gonadectomized rats results in a return to the
values found in normal intact animals (Hershko and Eilon,
1974).

It is clear from the above studies that both testicular
and ovarian hormones influence iron exchange. These
hormones exert their effects through erythrOpoiesis, fer-
ritin synthesis, ferroxidase activity of ceruloplasmin, and
iron absorption by the intestinal mucosa. Erythropoiesis is
enhanced by testosterone and repressed by estrogen (Kennedy
and Gilbertson, 1957). The higher serum and liver iron
values of the female could, in part, be explained by the
inhibition of erythropoiesis. With less iron involved in
hemoglobin formation, more is available for serum and
storage. Although iron can induce the synthesis of the iron
storage protein ferritin (Linder et a1., 1973), estrogen can

directly stimulate ferritin synthesis and therefore increase

12

iron storage (Bjorklid and Helgeland, 1970). Estrogens have
a stimulatory effect on the concentration of ceruloplasmin,
a ferroxidase, which brings about mobilization of iron from
iron stores (see Planas, 1973). Elevated levels of cerulo-
plasmin could thus cause a rise in serum iron. The influ-
ence of estrogen on the rate of absorption of dietary iron
by females is probably the most important factor in the
establishment of greater iron reserves. Erythropoiesis,
ferritin synthesis, and ceruloplasmin have their primary
affect in altering the internal distribution of iron in

storage and erythroid compartments.

13

Iron in Pregnancy

During pregnancy and lactation, increased iron stores
in the female may be a means of providing iron to offspring.
During pregnancy, not only do sex hormones change, but also
the iron needs of the female are greatly increased. Preg-
nancy depletes iron stores (Steenbock et a1., 1936; Widdow-
son and McCance, 1948; Hershko, Cohen and Zajicek, 1976),
causes an increase in iron absorption (Manis and Schacter,
1962; Hershko, Cohen and Zajicek, 1976), and increases
plasma iron turnover (Hershko, Cohen and Zajicek, 1976).
However, Hershko, Cohen and Zajicek (1976) report that the
depletion of liver iron occurs only in the last 5 to 6 days
of pregnancy in the rat, and that there is actually a 24%
increase in hepatic iron stores the first 15 days of preg-
nancy. In the rat, progesterone increases until day 15 of
pregnancy (Sato and Henkin, 1973); the increase in iron
absorption and hepatic iron storage coincides with the high
levels of this ovarian hormone. The precipitous drop in
storage iron during the last few days of pregnancy could be
from a combined effect of a drop in sex hormone stimulated
absorption and the vastly increased needs for iron for
growth of the young. On day 21 of pregnancy, only 21% of the
initial storage iron remains (Hershko, Cohen and Zajicek,
1976). Liver iron stores remain low during lactation
(Widdowson and McCance, 1948), and even 12 weeks after
weaning her young, a mother rat's hepatic iron stores still

do not match that of an unmated female of the same age

l4

(Widdowson and McCance, 1948).

Several authors caution against extrapolating the sex
difference in iron metabolism found in the rat to other
species. Their reason stems from the observations of
Widdowson and McCance (1948). In this study, a sex dif-
ference in liver iron was found in poultry, mice, and rats,
but not in guinea pigs or rabbits. The authors, however,
do not report the age since sexual maturity of the guinea
pig nor the reproductive condition of the rabbits. Since
age and reproductive condition can influence sex differences
in liver iron storage levels, the reported lack of a sex
difference in the guinea pigs or rabbits must remain in
question.

In humans a sex difference in serum and liver occurs,
but it is in the opposite direction to that found in the
rat: females have lower levels than do males. The monthly
loss of iron through menstruation (averaging about 40 ml
of blood and thus 20 mg of iron, Beaton, 1974) and the
iron stress of pregnancy combined with the rather low
intestinal absorption of iron in humans probably accounts
for the fact that premenopausal women have lower iron
stores than men (Charlton et a1., 1970). It has been
estimated that in some populations, 50% of the women of
reproductive age have no iron stores (see Frieden and

Osaki, 1974).

15
Regional Distribution and Quantification of Brain Iron in
the Human Brain

The presence of iron in the brain, detected by histo-
chemical means, was first reported by Zaleski (1886),
followed in 1915 by Guizzetti. Guizzetti recognized a
characteristic pattern of iron distribution in the brains of
various mammals and humans. The globus pallidus and sub-
stantia nigra were consistently found to have stainable
iron depostis although the iron reaction was not equally
intense in all species.

The first systematic study of the distribution of iron
in the brain was conducted by Spatz (l922a,b). The immer-
sion of thick macroscopic sections of human brain tissue
into potassium ferricyanide (Turnbull's blue) revealed
a characteristic pattern of iron distribution and allowed
Spatz to subdivide the centers of the central nervous sys-
tem into four groups according to the amount of iron
present.- Group one, having the most intense iron reaction,
included the globus pallidus and substantia nigra. The red
nucleus, caudate, putamen, subthalamus, and dentate nucleus
formed the second group. The third group included the
cerebral cortex, anterior thalamus, mammillary body,
midbrain tectum, cerebellar cortex, and central grey
matter of the third ventricle and gave a week iron reaction.
The fourth group, in which no staining was observed, was
comprised of the medulla oblongata, grey matter of the

Spinal cord, spinal and sympathetic ganglia and the white

l6

matter of the central nervous system.

In microscOpic sections, iron appears as a fine granu-
lar deposit in the neuropil, oligodendrocytes, and nerve
cells, mainly located in the globus pallidus and substantia
nigra (Spatz, 1922a; Muller, 1922).

WOllemann (1951) measured the intensity of the iron
reactions with photometric means and obtained results in
agreement with Spatz's observations. As it is not known
whether the histochemical methods stain all the iron present
in tissue, quantification of iron by chemical means has been
made on human brain tissue by several workers (WUth, 1923;
Tingey, 1937; Cumings, 1948; Hallgren and Sourander, 1958;
Sundermann and Kempf, 1961; Musil et a1., 1962; Courville et
a1., 1963; Harrison et a1., 1968). Although the results
from these studies show that a definite parallelism exists
between chemical and histochemical measurements (see Table 1
below), the actual amounts reported vary considerably.
Differences in extraction techniques, analytical methods,
iron contamination, presence of blood in the tissue, and
state of health and age of the individuals account for some
of this variation. Hallgren and Sourander (1958) report
that the iron content of the globus pallidus and substantia
nigra is of the same magnitude as that of the liver, but
that the amount of iron in the whole brain equals about one-
fifth the content of an iron replete liver.

That iron increases with age has been shown histochem-

ically (Guizzetti, 1886; Spatz, 1922a) and biochemically

17

Table l. The Distribution of Non-haemin Iron in Different

Parts of the Human Brain Autopsy Cases, 30-100 Years of Age

 

mg iron/100 g
fresh weight

 

Globus pallidus 21.30i3.49
Red nucleus 19.48i6.86
Substantia nigra 18.4616.52
Putamen .32i3.43
Dentate nucleus .35:4.86
Caudate nucleus 9.28:2.14
Thalamus 4.76:1.16
Cerebellar cortex 3.35i0.87
Motor cortex* 5.03i0.88
Occipital cortex 4.55i0.67
Sensory cortex 4.32:0.58
Parietal cortex 3.81:0.67
Temporal cortex 3.13:0.57
Prefrontal cortex 2.92i0.41
Frontal white matter 4.24i0.88
Medulla oblongata 1.40:1.16
Meninges 1.02:0.29
Liver 13.44i9.36

 

9540-100 years of age
(from Hallgren and Sourander, 1958)

l8

(Tingey, 1937; Hallgren and Sourander, 1958; Sundermann and
Kempf, 1961). Hallgren and Sourander measured the iron in
81 human brains of various ages and showed that different
centers of the central nervous system accumulate iron at
different rates. Some areas reach maximal levels that are
maintained, and others accumulate iron throughout life.

They report that the iron content of the globus pallidus and
substantia nigra increases rapidly the first 2 decades of
life and no further increase occurs after age 30, whereas
the caudate and putamen do not reach maximal levels until
age 50—60. Within the cerebral cortex, motor cortex has the
most iron, followed by occipital, sensory, and parietal
cortex. Lower values are found in the prefrontal and
temporal areas. Only the medulla oblongata showed no
increase of iron content with age. These authors also found
that, except for cases of anemia when brain and liver iron
values were both depressed, the brain iron levels do not
correlate with liver iron values.

Contrary to Hallgren and Sourander's results, Sunder-
mann and Kempf (1961) report that the iron content of the
globus pallidus, caudate putamen, and red nucleus increases
throughout life, whereas a maximum is reached by 30-40 years

in the cortex and the thalamus.

19

Brain Iron and Neurological Disorders

The excess accumulation of iron in the body tissues in
hemochromatosis is found also in brain iron stores (Sheldon,
1927; Cammermeyer, 1947). Increased brain iron in the
globus pallidus also has been reported in Hallervorden-Spatz
disease (Spatz and Metz, 1926; Meyer, 1958; Dooling et a1.,
1974) accompanied by disturbances in gait. General iron
metabolism appears normal; however, brain iron shows a
slower than normal turnover (Szanto and Gallijas, 1966).
Goldberg and Allen (1980) have shown that brain iron
accumulation is not a specific effect, for Hallervorden-
Spatz patients also have increased brain copper, zinc,
calcium, and manganese. Although iron is greatly increased
in the globus pallidus, non-pallidal areas (caudate, putamen,
substantia nigra, cortex, and hippocampus) actually show a
decrease in iron compared to normal brains. Histochemically
demonstrable iron is evident in the plaques and areas of
neurofibrillary change seen in the cortex of Alzheimer's
patients (Goodman, 1953). However, Hallgren and Sourander
(1960) were unable to find an increase in total iron levels
in the affected cortical areas. Wilson's disease is a
disorder of c0pper metabolism, and slightly increased brain
iron has been reported (Cumings, 1948). Although Spatz and
Metz (1926) report increases in basal ganglion iron in cases
of Huntington's chorea, Courville et al. (1963) were unable
to find a consistent difference in the brain iron levels

between normal and Huntington's patients. Increased brain

20

iron is reported in cases of Kaschin-Beck's disease, a
disorder caused by increased dietary iron (Hsaing, 1941);
however, the relative distribution follows normal patterns
(Spatz, 1922a). In dementia paralytica, brain iron depostis
are increased (Spatz and Metz, 1926; Tingey, 1937). Freeman
(1930) has reported a slight decrease in the brain iron of
patients“ with schizophrenia, and increased amounts of brain
iron have been demonstrated in the globus pallidus of
psychotic patients (Spatz and Metz, 1926; Strassman, 1945).

The epileptiform discharges occurring after head
trauma and hemorrhagic infarction have been traced to the
deposition of hemoglobin iron in the neuropil (Rubin and
Willmore, 1980). The damaging effects of iron are due to
its catalytic role in lipid peroxidation. The peroxidation
of lipids is a chain reaction caused by the hydroxyl radical
rearranging lipid bonds and destroying membranes. The
superoxide radical (027) formed from the reaction of ferrous
iron (Fe2)and molecular oxygen (02) reacts with H202 to
produce the very reactive hydroxyl radicals (OH’). Many in
2329 systems are present to prevent lipid peroxidation
either by preventing the reduction of iron (e.g., ceru-
loplasmin, transferrin), trapping or breaking down peroxides
GX-toc0pherol, catalase and peroxidase), or dismutating the
superoxide radical (superoxide dismutase) (Halliwell, 1979).
Regardless of these preventative systems, any neurological

disorder in which iron and oxygen can interact the

21
destruction of biological membranes by lipid peroxidation

can result.

22
Brain Iron in Laboratory Animals

There are no reports in the literature of the histo-
chemical localization of iron in the brains of non-human
mammals except for the reference to a "weak iron reaction"
in the brains of animals by Guizzetti (1886) and Spatz
(1922a). According to these authors, animal brain tissue
contains less iron than human tissue, and the quantity
parallels the evolutionary stage of the animal. The mouse
brain reportedly gives no iron reaction, whereas in the
rabbit, cat, dog, and monkey, a positive reaction of in-
creasing intensity is found in the globus pallidus and
substantia nigra. In the monkey brain, iron is visible in
the dentate nucleus but not the red nucleus. An increase
in iron with age in mammal brains in also reported by these
authors.

Reports of the chemical determination of brain iron
concentration in non-human mammals are also sparse. The
studies of Dallman and coworkers on the effects of iron
deficiency on brain iron of the developing rat (Dallman,
Siimes and Manies, 1975; Dallman and Spirito, 1977;
Weinberg, Dallman and Levine, 1979; Weinberg, Levine and
Dallman, 1979) provide some information.

Included in these studies are measurements of control
animal non-heme iron concentration in whole rat brain from
birth to about 11 weeks of age. The values given should
represent "normal“ brain iron concentrations for the

Sprague-Dawley rat. After the removal of the blood from

23

brain tissue, total non-heme iron is only slightly less than
total iron; only a small percentage of the brain iron is in
heme compounds (Rafaelson and Kofod, 1969). Total non-heme
iron in the whole brain is reported to be:

age (wks) ugFejg wet wt.

 

female rats:

(Dallman, Siimes and Manies, 1975) 3 9
4 9.3
6 9.8
10 10
ll 11
male rats:
(Dallman and Spirito, 1977) birth 5
1 5.3
2 5.13
3 6.11
5 7.0
8 6.4
male and female rats:
(Weinberg, Dallman and Levine, 1979)
(Weinberg, Levine and Dallman, 1979) 4 6-7 (f)
6-6.5 (m)
8 7.39 (f)
7.09 (m)

Investigation of the development of brain iron in the
male rat shows that brain iron concentration is low at birth
and increases only slightly during development from 5 ug/g
wet weight at birth to about 6 ug/g wet weight at 8 weeks.
However, as the brain is in a period of rapid growth during
this time, the total iron in the brain is actually increas-
ing. A fourfold increase in non-heme iron occurs between 3
days and 3 weeks of age, and a 40% increase occurs from 3 to
8 weeks, the peak uptake of iron by the brain occurring
between days 12 and 18. The iron compounds in the brain of

the adult male have an extremely slow rate of turnover

24

(Dallman and Spirito, 1977).

The values obtained from the measurement of iron in
male brains is fairly consistent from experiment to exper-
iment; however, the amounts of iron reported for the female
brain are higher in the Dallman, Siimes and Manies (1975)
study than in the others. Slight changes in the procedures
of iron measurement, diet, source or strain of experimental
animal may explain this difference. There was no statis-
tically significant sex difference reported in the brain
iron of control animals in any of the studies; however, in
all cases, the female brain had a higher concentration of
iron/g wet weight than the male. As most data were from
young animals, a statistically significant sex difference
may not yet have developed. In one study group in which the
diet of pregnant rats was supplemented with iron from 18
days of pregnancy through lactation, 4 week old female pups
had accumulated significantly more brain iron than male
littermates (Weinberg, Dallman and Levine, 1979). Other
reports (Rafaelson and Kofod, 1969; Williamson and Ng, 1980)
give values of between 10 and 11 ugFe/g wet weight for the
whole brain non-heme iron of adult rats--age and sex un-
reported. Unperfused rabbit brain contains 20 ugFe/g wet
weight non-heme iron (Hanig and Aprison, 1967).

The pattern of development of brain iron in the mouse
(Keen and Hurley, 1980) differs considerably from the devel-
0pment of brain iron in the rat reported by Dallman and

Spirito (1977). At birth, the mouse brain total iron

25
concentration is about 17 ugFe/g wet weight, drops by day 5

to about 7 ugFe/g wet weight, remains constant until about 3
weeks of age, and then increases until at 8 weeks the level
at birth, 17 ugFe/g wet weight, is reached (Keen and Hurley,
1980).

Although a systematic study of the iron concentration
in separate areas of the brain has not been performed on
non-human mammals, the following values have been reported

in the adult rat--age and sex unreported (Youdim and Green,

1977):
cerebellum .075i.008 ug/mg protein
brain stem .052:.008 ug/mg protein
caudate .092:.008 ug/mg protein
amygdala .058i.005 ug/mg protein
cortex .050i.003 ug/mg protein
hypothalamus .076 ug/mg protein

Fractionation of brain tissue in 0.32 M sucrose yields
a nuclear fraction Pl’ a crude mitochondrial fraction P2,
and a microsomal supernatant fraction. The percentage of
total non-heme iron in each fraction as reported for the rat
(Youdim and Green, 1977) and for the dog (Hallgren and

Sourander, 1958) is as follows:

rat dog
Nuclear fraction P1 8% 14%
Crude mitochondrial fraction P2 55% 46%

Microsomal and supernatant P3 37% 40%

26

Rajan, Colburn and Davis (1976) resuspended the crude
mitochondrial fraction, separated it into bands on Ficoll
gradients, and identified the components with electron
microscopy as well as measured the total iron in each. In
whole brain of the 6 week old male rat analyzed in this
manner, 82% of the iron was a myelin fraction, 6% in mito-
chondrial fraction, and 12% in synaptosomal fraction.

Similar subcellular fractionation of various areas of
the rat brain (see Table 2) show no consistent pattern in

the distribution of iron among different compartments.

27
Table 2. Determination of Iron in the Subcellular Fractions

of Different Brain Areas

 

 

 

Subcellular
Brain Areas Fractions ug/mg Protein
Brain Stem Myelin 0.15:0.02
Synaptosome 2.03il.92
Mitochondria 0.40i0.04
Synaptic vesicles 1.79:1.09
Membrane 4.70:0.86
Striatum Myelin 2.17i2.0
Synaptosome 1.30:0.04
Mitochondria 0.87:0.15
Membrane 8.0 i2.l
Cerebral Cortex Myelin 5.7312.01
Mitochondria 6.52il.50
Membrane 0.18i0.05
Synaptic vesicles 1.4liO.32
Hypothalamus Myelin 1.19:0.35
Synaptosome 2.76:0.80
Mitochondria 2.09i1.33
Cerebellum Myelin 5.81:3.39
Mitochondria -----

Note: The values shown in the Table represent each of the
average of four determinations of the tissue samples
prepared from different rat brain at various times.

(from Rajan, Colburn and Davis, 1976)

28

Forms of Iron in the Brain

As in the other tissues of the body, brain iron occurs
in both heme and non-heme compounds. When blood, and there-
fore hemoglobin, is removed from the brain, the non-heme
component is practically equal to total iron (Rafaelson and
Kofod, 1969).

Ferritin and the known iron enzymes comprise only a
portion of the measurable iron in the brain; the greater
part of the brain iron occurs in yet unidentified form(s).
Ferritin is a specialized tissue protein designed for iron
storage (Harrison et a1., 1974) in which iron is stored as a
hydrous ferric oxide phosphate complex within the protein
shell (Drysdale, 1976). Ferritin stains positively with
iron stains (Strassman, 1945; Deizel, 1955), appears dis-
persed within the cell (Westcott et a1., 1953), is water
soluble, and is heat stable to 70°C (Weinfeld, 1964). Only
about one-third of the human brain non-heme iron is in this
form (Hallgren and Sourander, 1958) and only 15-25% in whole
rat brain (Dallman, Siimes and Manies, 1974), whereas 60—70%
of liver non-heme iron is stored in ferritin (Kaldor, 1958;
Cook, Hershko and Finch, 1973). Sundermann and Kempf (1961)
report nearly identical patterns of distribution and change
with age for brain total iron and ferritin.

Hemosiderin, the other iron storage compound of the
liver, spleen, and bone marrow has been identified histo-
chemically in damaged brain tissue (Strassman, 1945).

Hemosiderin appears as yellow-brown deposits in unstained

29
tissue and stains positively for iron (Strassman, 1945;
Subcommittee on Iron, 1979). It is considered to be an
aggregate form of ferritin, is insoluble in water, and
contains lipids and carbohydrates as well as protein
(Deizel, 1955; Weinfeld, 1964; Subcommittee on Iron, 1979).
The intracellular granular iron deposits of the globus
pallidus and substantia nigra do not contain lipids or
carbohydrates, as determined histochemically, and therefore
apparently do not contain hemosiderin (Deizel, 1955).

The remaining known iron compounds are enzymes.
Although not present in large quantities, the following
enzymes are of importance in brain chemistry because of
involvement either directly or indirectly in neurotrans-
mitter metabolism.

The iron containing enzymes aconitase and succinic
dehydrogenase are related to the metabolism of four of the
eight most abundant compounds in the brain (occurring in
concentrations > 3 mM). The pathway for the synthesis of
these compounds--glutamate, y-aminobutyric acid (GABA),
glutamine, and glutathione--is derived directly from Kreb's
cycle via a shunt, sometimes called the GABA shunt, occur-
ring between a-ketoglutarate and succinate (see Cooper,
Bloom and Roth, 1978). Aconitase catalyzes reactions
required for the synthesis of‘a—ketoglutarate, and the
conversion of succinate to fumarate is catalyzed by suc—
cinate dehydrogenase. It has been estimated that the GABA

shunt may account for 10-40% of the total brain metabolism

30

(see Cooper, Bloom and Roth, 1978).

The peroxides formed from the oxidation of monoamine
neurotransmitters are degraded by catalase, an iron contain-
ing enzyme. The greatest concentrations of catalase are
found in the norepinephrine-rich hypothalamus and the
dopamine-rich substantia nigra (Brannan et a1., 1981).

Tyrosine hydroxylase is the initial and rate limiting
enzyme in the synthesis of the catecholamine neurotrans-
mitters norepinephrine (NE) and dopamine (DA) (Sourkes,
1972). In the brain, the enzyme is associated with the
synaptosome fraction (Cooper, Bloom and Roth, 1978).

The exact relation of iron to the activity of monoamine
oxidase (MAO) is not clearly understood; however, the
maintenance of normal MAO levels requires iron (Symes,
Missala and Sourkes, 1971). MAO is an important enzyme in
the degradation of NE, DA, and serotonin.

Iron is essential for the catalytic activity of trypto-
phan hydroxylase, the initial and rate limiting enzyme in
the synthesis of serotonin (Kuhn et a1., 1980). Aldehyde
oxidase, a key enzyme in the degradation of serotonin, is
also an iron requiring enzyme (Bray, 1975).

As well as having a role in the synthesis and degrada-
tion of biogenic amines, iron may be an important factor in
chelate formation in the storage and transport of catechola-
mines. Rajan, Colburn and Davis (1971) have reported
significant amounts of iron, and other metals, in the sub-

cellular fractions of rat brain which are known to have high

31

concentration of catecholamines. The molecular structure
and functional groups associated with the catecholamines is
compatible with these molecules having strong metal chela-
tion potentials. Ferrous iron specifically enhances the
binding of serotonin to serotonin-binding protein in synap-
tosomes and serotonergic tracts (Tamir, Klein and Rapport,
1976) and is involved in serotonin receptor mechanisms
(Lehmann, personal communication).

The fact that iron is involved in so many different
aspects of many different neurotransmitters probably ac-
counts for the fact that the distribution of iron does not
correspond exactly to the known distribution of any neuro-
transmitter.

Rafaelson and Kofod (1969) claim a small amount of iron
is likely in a complex attachment to lipids in the brain.
Perhaps this is in relation to iron's structural role in
many membranes and membrane bound proteins (Jacobs and
Worwood, 1974). In brain tissue treated with prussian blue,
iron appears in the neuropil (Spatz, 1922a), perhaps in
association with myelin. In addition, Rajan, Colburn and
Davis (1976) found that over 80% of the iron in resuspended
crude mitochondrial fraction from whole rat brain is in the
myelin subfraction. In this regard, it is noteworthy that
(l) the peak uptake of radio iron by the rat brain-—days 12
to 18 postnatally (Dallman and Spirito, l977)--coincides
with the beginning of CNS myelination--10 to 12 days post-

natally (Norton and Poduslo, 1973); (2) both myelin

32

formation and iron accumulation continue late into adulthood
(Hallgren and Sourander, 1958; Norton and Poduslo, 1973);

(3) both iron in the male rat brain (Dallman and Spirito,
1977) and myelin (Davison and Peters, 1970) are stable,
showing very slow turnover rates; and (4) in Dallman, Siimes
and Manies' (1974) study in which a persistent depression of
brain iron occurred following short-term iron deficiency in
the rat pup (see below), the iron deprivation occurred
during postnatal days 10 to 28, the period of the initiation

of myelin formation and its rapid proliferation.

33
Iron Deficiency, the Brain and Behavior

Iron deficiency is characterized by anemia, stunted
growth, epithelial changes, gastrointestinal abnormalities,
abnormal lipid metabolism, weakness, increased susceptibil-
ity to infection, anorexia, amenorrhea, menstrual irreg-
ularity, and alterations in behavior (Sherman, 1978; Sub-
committee on Iron, 1979; Oski, 1979).

The behavioral changes in patients with iron deficiency
include irritability, apathy, listleness, fatigue, lack of
ability to concentrate, pagophagia (pathological craving for
ice), and in children, inattention, hyperactivity, and
decreased scholastic performance (Webb and Oski, l973a,b;
Cantwell, 1974; Pollitt and Leibel, 1976).

In studies of laboratory animals, the behavioral
effects of iron deficiency include decreases in total
activity, a reversal of diurnal rhythm (Glover and Jacobs,
1972), and changes in maze learning and forced exercise
performance (Edgerton et a1., 1972; Pollitt and Leibel,
1976), and general responsiveness to environmental stimuli
and learned task performance (Weinberg, Levine and Dallman,
1979; Weinberg, Dallman and Levine, 1979).

The behavioral effects of iron deficiency are thought
to result from altered metabolism due to decreased amounts
of iron available for the brain iron enzymes involved in
cellular oxidative functions and neurotransmitter metabolism
(Pollitt and Leibel, 1976; Mackler et a1., 1978; Youdim et
a1., 1980).

34

Following a brief period of iron deficiency in the
young rat (days 10 to 28), a deficit in brain iron persists
even to the adult stage (Dallman, Siimes and Manies, 1974).
The behavioral changes in general responsiveness, reac-
tivity, and avoidance learning seen in these animals also
persist in rehabilitated adult rats (weinberg, Dallman and
Levine, 1979). These experiments suggest that severe iron
deficiency during development induces irreversible changes
in the brain.

Although in iron deficiency the activity of the iron
enzymes cytochrome C, cytochrome oxidase, succinic dehydro-
genase, and monoamine oxidase are reduced in various body
tissues (Dallman and Schwartz, 1968; Symes, Missala and
Sourke, 1971; Youdim et a1., 1980), in the brain, neither
the ability to carry out oxidative phosphorylation nor the
levels of these enzymes appear to be affected (Mackler et
a1., 1978; Youdim et a1., 1980). However, the activity of
aldehyde oxidase, a key enzyme in serotonin degradation, is
significantly reduced in the brain of iron deficient rats,
and concentrations of serotonin and total 5-hydroxyindole
compounds are reported to be elevated (Mackler et a1.,
1978). The increased concentration of serotonin can cause
drowsiness, inattentiveness, and a decreased ability to
learn (Douglas, l974)-—symptoms often associated with iron
deficiency.

Youdim and associates found that, except for a decrease

in serotonin, no changes occurred in the activities of

35
enzymes related to catechole or indole amine metabolism nor
in the level of dopamine in the brain tissue of iron defi-
cient rats (Youdim et a1., 1980). However, the hyperac-
tivity caused by drug induced increases in serotonin and
dopamine was depressed in iron deficient animals even
though, under these conditions, the amounts of these puta-
tive neurotransmitters were alike in normal and iron-
deficient rats. That the decreased activity was due to a
decreased post-synaptic response (i.e., at the receptor
level) is supported by the fact that putative agonists to
dopamine and serotonin also cause a depressed behavioral
response (Youdim et a1., 1980), and a further study (Ash-
kenazi, Ben Shachar and Youdim, 1980) has demonstrated a 50%
decrease in dopamine receptors in the caudate of iron
deficient rats.

Iron is important in many aspects of brain function,
with involvement in the catecholamine and serotonin neuro-
transmitter systems especially evident. Catecholamine and
serotonin systems are sex hormone sensitive (Kalra et a1.,
1972; Kalra and McCann, 1973) and are involved in the
sexual differentiation of the brain (Arai and Gorski, 1968)
and in the central nervous system regulation of gonado-
trophin secretion (Ganong, 1975; McCann and Moss, 1975).
The availability of iron could thus affect the central
nervous system regulation and coordination of reproductive

functions.

36

In view of the prevalence of iron deficiency in the
world and its possible long term effects on brain neuro-
transmitter systems, the effects of hormones and other
factors on the iron accumulating abilities of the brain
merit investigation.

It is the purpose of this study to (1) localize iron
deposits in the rat brain; (2) determine if a sex difference
exists in brain iron stores; (3) determine the effects on
brain iron levels of natural events in which sex hormones
fluctuate (e.g., estrous cycle and pregnancy); and (4)
determine if exogenous estrogen alters the effects of

ovariectomy and castration on brain iron levels.

MATERIALS AND METHODS

Both histochemical and spectrophotometric evaluation of
brain iron are made. The histochemical techniques provide
qualitative determinations of brain iron by localizing
deposits and indicate by the intensity of the stain the
amount of iron present. SpectroPhotometry yields a quanti-

tative measure of iron within specific areas of the brain.

HISTOCHEMISTRY

Animals

Intact Sprague-Dawley (SD) rats were obtained from the
Animal Breeding Center, National Institutes of Health, Beth-
esda MD; castrated and ovariectomized animals purchased
from Taconic Farms, Germantown, NY, were delivered the day
after surgery. 'The animals were kept in groups of two to
four in standard plastic laboratory cages and provided
with water and Purina Rat Chow ad libitum. The light
cycle was 8 hrs light and 16 hrs dark, the dark period

beginning at 1330 hrs. All animals were allowed to adjust
to our labaoratory conditions for at least 3 weeks before

being experimentally utilized.

Treatment Groups

 

The histochemical investigation was composed of four
separate studies, each designed to answer specific questions

about brain iron deposits. Below is an outline of each of

37

38

the studies with a description of the purpose and the

number, age, and sex of the animals. Any special treatments

are included with the description of the group concerned.

Study 1 - to determine (1) location of iron in the rat

brain, (2)

and (3) if

brain iron.

4.

4.
4.
4.
3.
l

3

12-18
12—18
32 wk
32 wk
43 wk
52 wk

if brain iron increases with age,

there is a sex difference in

wk old virgin females

wk old virgin males

old retired breeder females
old retired breeder males
old retired breeder females

old retired breeder female

Study 2 - to determine if pregnancy causes visible

changes in brain iron either in location or

amount.

These animals can be compared with

the control animals of the same age in

Group 1.

2, 18 wk old 7 days pregnant females

4, 18 wk old postpartum females

3, 32 wk old 16 days pregnant females

About 1 hr before the dark period of the light cycle, a

female was placed in a cage with a sexually mature male. If

lordosis and mating were observed, the pair remained togeth-

er for 1 to 2 hrs,

after which the female was removed.

39

Thus, the time of insemination is known within about 2 hrs.

Study 3 - to determine if ovariectomy causes visible
changes in brain iron either in location or
amount and to determine if estrogen causes a
visible change in brain iron compared with
untreated ovariectomized animals.
2, 18 wk old females, ovariectomized at 10 wks
2, 32 wk old females ovariectomized at 28 wks
2, 18 wk old females, ovariectomized at 10 wks,
received estrogen implant at 17 wks
3, 32 wk old females ovariectomized at 28 wks,
received estrogen implant at 30 wks

Estrogen implants were prepared by heating a few grams
of B-estradiol 3-benzoate (Sigma Chemical Company, St.
Louis, MO) in a small beaker. When the powder melted, a
thin wire was dipped into the liquid, removed, and the
coating allowed to air dry. This process was repeated
several times until a pellet about the size of a grain of
rice formed. Any wire protruding beyond the hormone “grain"
was then removed.

The animal was anesthetized with an intraperitoneal
injection (0.05 cc/100 g body weight) of Nembutal (sodium
pentobarbital 60 mg/ml, Abbott Laboratories, N. Chicago,
IL). The back of the neck was shaved and then washed with a
1:750 solution of Zephiran (Winthrop Laboratories, New York,

NY). A 3 cm incision was made longitudinally through the

40

skin on the back of the neck. The implant was broken in
half to remove the wire and then inserted under the skin.
Before closing the incision with two or three clips, the
wound was packed with Furacin (Eaton Veterinary Labora-

tories, Norwick, NY).

Study 4 - to determine if castration causes any visible
change in brain iron, either in location or
amount and to determine the effect of estro-
gen on the brain iron of castrated males.
These animals can also be compared to the
females of similar age in Group 3.

2, 32 wk old males castrated at 30 wks
2, 32 wk old males castrated at 28 wks,
received estrogen implant at 30 wks

The estrogen implant was made and inserted in the same

manner as in Study 3 animals above.

Preparation of Tissue
All animals were sacrificed between 0800 and 1300 hrs.

The animals were first weighed and then anesthetized with an
intraperitoneal injection of Nembutal (0.1 cc/100 g body
weight). The chest cavity was opened and the animal per-
fused intracardially with 50 ml 0.9% saline, followed by 250
ml of 10% formalin in 0.9% saline. Immediately upon comple-
tion of perfusion, the brain was removed from the skull and

placed in a solution of 10% formalin in 10% alcohol. The

41

tissue remained in this solution until cut. To prevent any
possible metal contamination of the tissue, great care was
exercised to avoid touching the brain with surgical instru-
ments.

Brains were cut frozen, with a sliding microtome. Most
were cut at 50 pm in the coronal plane; however, a few were
cut in the sagittal or horizontal plane and a few cut at 100
or 150 pm. The knife was coated with silicone stopcock
grease (Dow Corning, Midland, MI) to prevent metal contami-
nation from the knife.

Although many procedures are available for the local-
ization of iron in biological tissues (e.g., Pearse, 1961;
Humason, 1979), Perl's reaction (prussian blue) is found to
be most effective on the rat brain tissue used here. A
modification of a recently discovered intensification of the
Perl's reaction with diamino benzidine (DAB) was used on
some tissue (Nguyen-Legros, et a1., 1980).

Perl's reaction localizes ferric iron (Fe+++) by
forming blue-colored ferric ferrocyanide when tissue is
acidified in the presence of potassium ferrocyanide. The
acid (HCl) serves to remove some of the Fe+++ from the
proteins to which it is bound. The Turnbull blue reaction
for ferrous iron (Fe++) (Humason, 1979) produces no visible
reaction in rat brain tissue suggesting that most iron in
the rat brain is in the ferric form.

The intensification of the Perl's reaction with DAB is

based on the oxidation of benzidine by H202 in the presence

42
of a suitable catalyst (Nguyen-Legros et a1., 1980).
Oxidized benzidine compounds form blue or brown colored
precipitates. Many compounds can catalyze the oxidation
including hemoglobin, myoglobin, peroxide (Tietz, 1976), and
ferric ferrocyanide (Nguyen-Legros et a1., 1980). Here the
ferric ferrocyanide of the Perl's reaction catalyzes the
oxidation of benzidine by H202, forming a brown precipitate.
Since Perl's reaction is blue, Neutral Red was used as a
counterstain (only sections 50 pm or less were counter-
stained). However, when the intensification step with DAB

was used, thionin was the counterstain.

43

Staining Techniques

Perl's stain for ferric iron
Acid wash all glassware; avoid the use of metal
instruments.

Solutions

 

2% Hydrochloric acid (HCl):

To 2 m1 concentrated HCl add deionized water
to the 100 m1 mark.

2% Potassium ferrocyanide (K4Fe(CN)6):

To 2 g potassium ferrocyanide add deionized
water to the 100 m1 mark.

Neutral Red

7.5 g Neutral Red (Chroma-Gesellschraft)

2.5 g Safranin O (Chroma-Gesellschraft)

Add deionized water to the 1000 m1 mark.
Staining;procedure

1. Rinse tissue 10-20 min in deionized H20.

2. Mix 1:1 freshly prepared 2% H01 and 2%
potassium ferrocyanide; heat quickly to
56°C.

Immerse tissue and agitate for 3 min.
Rinse tissue in deionized H20 - 10 min.

Mount on subbed microscope slides.

C‘U‘l-l-‘UO

When slides are dry, immerse in 70% alcohol
30 sec.

7. Distilled H20 30 sec.

8.

9.
10.
ll.
12.
13.
14.

44

Immerse in Neutral Red 1 min.

Rinse in H20 1 min.

70% alcohol 30 sec.

95% alcohol - 2 changes 15 sec each.
Absolute alcohol - 2 changes 30 sec each.
Xylene - 2 changes at least 3 min each.

Mount with Permount.

Iron - blue, Background - pink

45

Perl's DAB for ferric iron

 

Solutions

 

2% HCl and 2% Potassium ferrocyanide are made the
same as in Perl's reaction.
0.2 M monobasic sodium phosphate
To 27.6 g monobasic sodium phosphate add
deionized water to the 1000 m1 mark.
0.2 M dibasic sodium phosphate
Add 28.4 g dibasic sodium phosphate (rapidly
stirring with a magnetic stirrer) to 800 m1
deionized water; when dissolved add deionized
water to the 1000 m1 mark.
0.1 M phosphate buffer pH 7.4
To 190 ml of 0.2 M monobasic sodium phosphate
add 180 m1 of 0.2 dibasic sodium phosphate;
add deionized water to the 2000 m1 mark.
Diamino benzidine DAB (Sigma Chemical Company,
St. Louis, MO)
Dissolve 0.05 g DAB in 100 ml 0.1 M phosphate
buffer pH 7.4; filter.
1% Hydrogen Peroxide (H202)
To 3.3 ml of 30% H202 add deionized water to
the 100 m1 mark.
DAB is a possible carcinogen; it is handled with care:
used glassware, leftover solution, and other material can
apparently be decontaminated by immersion in a solution of

laundryibleach.

 

46

Staining procedure

1.

3
4.
5
6

React tissue as in Perl's stain steps 1
through 4 (heating of Perl's solution to 56°C
can be omitted as satisfactory reaction is
obtained with solutions at room temperature;
this also results in less tissue damage).
Immerse tissue in DAB solution kept cool with
ice on rotary table 20 min.

Add 1 ml of 1% solution of H202 20 min.

Rinse in deionized water 10 min.

Mount on subbed slides.

When dry, counterstain with thionin pH 4.5.

The above DAB procedure was performed on tissue without

pretreatment with Perl's and also on tissue in which the

ferric iron had been reduced to the ferrous form with mercap

toacetic acid (1.0%)., In neither of these situations was

DAB staining apparent.

47

*
Thionin pH 4.5

 

Solutions

 

Formalin alcohol
To 40 m1 commercial formaldehyde add 460 ml
95% ethanol.
Chloroform ether alcohol
Mix together 25 ml 95% ethanol
25 ml ether
200 ml chloroform
Thionin pH 4.5
Mix together 12.5 ml of 1% aqueous thionin
solution (Fisher Scientific Co., Fair
Lawn, NJ).
56.5 ml 1.0 M acetic acid
42.5 ml 1.0 ml sodium acetate
138.5 ml deionized water
Acid alcohol
To 5 m1 concentrated HCl add 295 ml 70%
ethanol.
Acetic acid alcohol
Add 1 m1 glacial acetic acid to 500 ml 95%
ethanol.

Staininggprocedure

 

l. 95% alcohol 10 min.

2 Formalin alcohol, 5 min.
3. 95% alcohol 3 min.

4

Chloroform, ether, alcohol 10 min.

\OWVO‘UI

10.
ll.
12.
l3.
14.

15.
16.
17.
18.
19.
20.

48

95% alcohol 3 min.
Acid alcohol 5 min.
70% alcohol 3 min.
Deionized H20 2 min
Thionin stain (longer as stain ages) 3 min.
Deionized H20 wash.

Deionized H20 2 min.

70% alcohol 3 min.

95% alcohol 3 min.

Acetic acid alcohol, watch closely; check
every 2-3 min.

95% alcohol 3 min.

100% alcohol 3 min.

100% alcohol 3 min.

Xylene 3 min.

Xylene 3 min.

Mount with Permount.

Iron - brown/black, Background - blue

* R.C. Switzer, unpublished technique.

49

Analysis of Data

 

The histochemical localiZation of iron is useful in
determining the distribution of iron in the brain and
permits within-section determinations of the relative
concentration of iron among different areas of the brain.
However, most of the tissue in this study was stained with
Perl's method alone as the intensification procedure was
discovered only after most of the material had been pro-
cessed. At best, the Perl's method produces a light stain
in rat brain tissue and accurate determinations of small
between-section differences are difficult to make. Slight
differences in the thickness of the section or in the
processing of tissue could result in greater differences
than the treatments themselves. The judgement of differ-
ences in intensity of stain in the histochemistry results
section include only large relative differences and have the
above mentioned limitations.

The determination of quantitative differences in brain
iron among treatment groups in this study is based primarily
on spectrophotometry because it is an accurate, reproducible

and sensitive method of measuring iron concentration.

50

SPECTROPHOTOMETRY

Animals

The Sprague-Dawley (SD) rats used in the pregnancy
study were obtained from the National Institutes of Health
laboratories. They arrived 11 to 14 wks of age and were
allowed to adjust to our laboratory conditions for at least
3 wks before mating. All other rats were bred in our own
facility, the Laboratory of Brain Evolution and Behavior,
Poolesville, MD, from breeding stock purchased from the
National Institutes of Health. The animals were kept in
groups of two to four in laboratory cages and provided with
water and Purina Rat Chow ad_libitum. The light cycle was
8 hrs light and 16 hrs dark, the dark period beginning at
1330 hrs.

Treatment Groups

The spectrOphotometric determination of iron was
performed on the serum, liver, and both high iron (globus
pallidus + substantia nigra) and lower iron (cortex) areas
of the brain. The investigation was organized into three
studies each designed to answer specific questions about
brain iron. The outline below describes the purpose of the
study and the number, age, and sex of the animals. Any

special treatment is included in the description.

51

Study 1 - to determine (1) if the levels of brain iron
change throughout the estrous cycle when
endogenous levels of sex hormones fluctuate,
and (2) if a sex difference occurs in brain
iron levels. The stage of estrus is
determined by vaginal smear.
10, 12 wk old virgin female rats is estrus
10, 12 wk old virgin female rats in metestrus
10, 12 wk old virgin female rats in diestrus
10, 12 wk old virgin female rats in proestrus
10, 12 wk old virgin male rats

Sexual maturity in rats is reached between 6 and 8 wks;
delaying sacrifice until 12 wks of age should allow sex
hormone related changes to occur. The stage of estrus of
the females was determined by vaginal smear.

Vaginal Smear

A cotton-tipped applicator was inserted into the vagina
and then rolled onto a subbed microsc0pe slide, air dried
for 10 sec, and then immersed into a 10% solution of for-
malin. The smear was stained with Harris' Hematoxylin (see
below).

Vaginal smears were analyzed and the stage of estrus
determined according to the description by Turner (1961).
Vaginal smears contain polymorphonuclear leukocytes and
epithelial cells. In the proestrus stage of the estrus
cycle, smears contain nucleated epithelial cells, singly or

in sheets. Estrus smears are characterized by masses of

52
cornified epithelial cells. Metestrus follows estrus and in
vaginal smears some leukocytes are found among the epi-
thelial cells. Diestrus smears contain large numbers of

leukocytes and few epithelial cells.

Study 2 - to determine if brain iron levels change
during pregnancy, a period when liver iron
is raised initially, then depleted, and
maternal sex hormone levels change.

7, 19 wk old female rats, 4 days pregnant,
lst pregnancy

7, 19 wk old female rats, 8 days pregnant,
lst pregnancy

7, 19 wk old female rats, 12 days pregnant,
lst pregnancy

7, 19 wk old female rats, 16 days pregnant,
lst pregnancy

7, 19 wk old female rats, 20 days pregnant,
lst pregnancy

7, 19 wk old female rats, 12-20 hrs post partum,
lst litter

Pregnant females were obtained by placing a female in a

cage with a sexually mature male at about 1 hr before the
dark period of the light cycle. If lordosis and mating were
observed, the pair remain caged together for l to 2 hrs,
after which the female was removed. Thus, the time of

insemination was known within about 2 hrs.

53

As much as possible, the pregnancies were allowed to

occur so that at the day of perfusion, all animals were 19

wks i 6 days of age.

Study 3 - to determine (1) if ovariectomy or castration

10,

10,

10,
10,

affects brain iron levels, and (2) if 3 wks
treatment with estrogen increases or de-
creases iron levels compared with control
animals and ovariectomized or castrated
animals.
12 wk old female rats ovariectomized at 4 wks
of age
12 wk old female rats ovariectomized at 4 wks
of age and implanted with estrogen at 9 wks
of age
12 wk old male rats castrated at 4 wks of age:
12 wk old male rats castrated at 4 wks of age;

and implanted with estrogen at 9 wks of age

The intact males and females from Study 1 are the

control animals to which the above treatments can be com—

pared.

Gonadectomy occurred at 4 wks of age, before the onset

of sexual maturity. The estrogen implants were left in for

3 wks prior to sacrifice in the expectation that 3 wks is

long enough to effect estrogen—dependent changes in iron.

54

Ovariectomy

Four week old female rats were anesthetized with an
intraperitoneal injection of Nembutal (0.05 cc/100 g body
weight). The lower back was shaved and then washed with a
1:750 solution of Zephiran. A 3 cm incision was made
longitudinally through the skin of the back with a scalpel
and the incision was kept open with a retractor. The
Opening was moved over an ovary and a small (0.7 mm) inci-
sion was made through the body wall at the level of the
ovary. The ovary was picked up with forceps and pulled
through the opening. The blood vessels were cauterized
(Codman neurocoagulator) and the ovaries cut from the uterus
with fine scissors. The incision in the body wall was
closed with one or two stitches. The retractor was moved
over the other ovary which was removed in a similar fashion.
Before closing the midline incision with clips, the wound

was packed with powdered Furacin.

Castration

Four week old male rats were anesthetized with an
intraperitoneal injection of Nembutal (0.05 cc/100 g body
weight). The scrotum was shaved and then washed with a
1:750 solution of Zephiran. A 1 cm incision was made along
the base of the scrotum through the scrotal skin and a small
opening made at the base of each cremasteric pouch with fine
scissors. By exerting a slight pressure on the upper

scrotum, the testes were pushed out of the openings of the

55
cremasteric pouches. The blood vessels were cauterized and
the testis cut from the vas deferens with scissors. The
wound was packed with powdered Furacin and the incision

closed with two or three stitches.

Estrogen implant preparation

Silastic tubing (Dow Corning, Midland, MI) .078 in ID x
.125 in OD was cut into 10mm lengths and one end closed off
with a dab of Silastic Medical Adhesive (Dow Corning,
Midland, MI). When the adhesive was dry the tube was
filled with beta-estradiol 3 benzoate - about 0.01 g (Sigma
Chemical Co., St. Louis, MO) - and the other end of the
tubing closed off with adhesive. The implant was soaked in
a 0.9% saline solution for 24 hrs before being placed in the
animal. Implants prepared in this manner permit a relatively
steady rate of diffusion of hormone (Ciaccio, L.A., personal

communication).

Insertion of implant

Nine week old ovariectomized females and castrated
males were anesthetized with intraperitoneal injection of
Nembutal (0.05 cc/100 g body weight), The back of the neck
was shaved and then washed with a 1:750 solution of Zephi-
ran. A 3 cm incision was made longitudinally through the
skin on the back of the neck. The implant was inserted
under the skin. Before closing the incision with two

or three clips, the wound was packed with Furacin.

56

Harris' Hematoxylin

 

 

Solutions
Hematoxylin
Hematoxylin crystals 5.0 gm
Alcohol, 95% 50.0 cc
Ammonium or potassium alum 100.0 gm
Distilled water 1000.0 cc
Mercuric oxide 2.5 gm

Dissolve the hematoxylin in the alcohol, the alum in
the water by the aid of heat. Mix the two solutions.
Bring the mixture to a boil as rapidly as possible and
then remove from the heat and add the mercuric oxide.
Reheat the solution until it becomes a dark purple,
about 1 min, and promptly remove the container from the
heat and plunge it into a basin of cold water. The
solution is ready to use when cool. Add 2-4 cc of
glacial acetic acid to 100 cc of solution if desired.
Acid Water

Distilled water 1000 cc
Hydrochloric acid, concentrated 10 cc

Eosin-Phloxine Solution
Stock Eosin

Eosin Y, water soluble 1.0 gm
Distilled water 100.0 ml

Stock Phloxine

Phloxine B 1.0 gm
Distilled water 100.0 ml

Working Solution

Stock Eosin 100.0 m1
Stock Phloxine 10.0 ml
Alcohol, 95% 780.0 ml
Glacial acetic acid 4.0 m1

57

Make up working solution as needed. Working solution

should be changed at least once a week.

Staining Procedure

 

l.

2
3.
4

10.
11.

Results:

Wash with tap water.

Harris' hematoxylin for 10 min.

Rinse in tap water.

Differentiate in acid water - 3 to 10 quick dips.
Check the differentiation with the microscope--
nuclei should be distinct and the background very
light or colorless.

Wash in running tap water for 5 min.

Stain with eosin from 15 sec to 2 min depending on
the age of the eosin and the depth of counterstain
desired.

70% alcohol.

95% alcohol.

Absolute alcohol - at least 2 changes.

Xylene - 2 changes

Mount in Permount.

cornified epithelial cells - pink

all other cells - blue

58

Table 3. Summary of Iron Measurement Study

number age in
of wks at
animals sacrifice
Females
intact proestrus 10 12
estrus 10 12
metestrus 10 12
diestrus 10 12
day of pregnancy
at sacrifice

pregnant 7 4 l9
7 8 19
7 12 19
7 16 19
7 20 19
7 postpartum 19

age in age in

wks of wks of

gonadectomy implant
ovariectomized (OVX) 10 4 12
OVX + estrogen implant 10 4 9 12

Males

intact 10 12
castrated (CAST) 10 4 12

CAST + estrogen implant 10 4 9 12

59

Collection of TisSue for the Spectr0photometric Measurement

of Iron

All animals were sacrificed between 0800 and 1300 hrs.
They were first weighed and then anesthetized with an
intraperitoneal injection of Nembutal (0.1 cc/100 g body
weight). The chest cavity was opened and a 3 cc syringe
equipped with a 20 gauge needle was filled with blood from
the base of the heart. The needle was removed from the
syringe and the blood poured into an acid-cleaned test tube
and refrigerated, undisturbed, until the following day.
After the blood sample was taken, the animal was perfused
intracardially with 200 ml of saline to remove blood from
the tissues. The brain and a portion of the liver were
removed, placed in separate plastic weighing dishes, and
frozen on a block of dry ice. When frozen, the dish was
covered over with aluminum foil, sealed in a plastic bag,
and kept frozen (-20°C) for up to 12 wks. During perfusion,
the uterus was removed from intact, ovariectomized, and
estrogen treated females, its diameter measured, and, where
present, the ovaries were inspected for the presence of
follicles. In pregnant animals, the number of embryos was
recorded except at 4 days of pregnancy at which they were
too small to be seen. However, the uterus size was recorded

for the 4 day pregnant rats.

60

The morning following perfusion, the blood sample, now
clotted, was spun for 10 min at 3000 RPM and the serum
pipetted off into acid-cleaned screw-top vials and frozen

for up to 12 wks.

61

Requirements for the Spectrophotometric Measurement of Iron

Before the iron present in biological tissues can be
measured by spectrophotometric means the cells must be
separated and broken open; the iron is then removed from
proteins; the proteins are precipitated; ferric iron is
reduced to the ferrous form; copper is complexed to prevent
interference with iron measurements; and ferrous iron is
reacted with a chromogen to form a colored complex.

The procedure used here is based on the serum iron
method of Zak, Baginski and Epstein (1980), modified for
measurement of tissue extracts. The extraction of iron from
tissue is based on the method of Weinfeld (1964) in which
only the storage iron compartment, non-heme iron, is ex-
tracted, the heme compounds being resistant to acid hydro-
lysis. In the rat brain the non-heme iron is practically
identical to total iron when the blood has been removed from
the tissue (Rafaelsen and Kofod, 1969).

The tissue was broken up by homogenization, and a cool,
heat, cool cycle promoted cell lysis. It has been deter-
mined by Weinfeld (1964) that complete extraction of non-
heme iron from tissue occurs by hydrolysis in a 2.8N HCl
solution at 90°C for 60 min. For this reason, here, to two
parts tissue homogenized in iron-free water, one part of
8.5N HCl was added, giving a strength of 2.8N. In the
determination of serum iron, the removal of iron from the

transferrin was accomplished by mercaptoacetic acid

62
(Fielding and Ryall, 1970; Zak, Baginski and Epstein, 1980)
and trichloracetic acid (Tietz, 1976). Trichloroacetic acid
has the added advantage of precipitating proteins (Tietz,
1976; Zak, Baginski and Epstein, 1980).

Since most chromogens react with ferrous ions, ferric
iron must be reduced. Of available reductants, mercapto-
acetic acid, used here, has the added advantage of com-
plexing copper, thus eliminating the possibility of a
copper-chromogen complex interfering with iron-chromogen
measurements.

The chromogen 2,4-bis(5,6-diphenyl-l,2,4-triazin-3—y1)
pyridine tetrasulfonate (BDTPS), was chosen because it is
very sensitive and stable. At a pH between 2 and 6, ferrous
iron forms a magenta—colored complex with a peak maximum at
565nm where it has a molar absorptivity of 32,000 (Zak,
Baginski and Epstein, 1980; G. Frederick Smith, 1980).

The pH of the final solution can be adjusted by using a
30% sodium acetate solution in serum sample determinations.
However, in the determination of brain and liver iron, BDTPS
was dissolved in a saturated sodium acetate solution and the
pH of the brain extract further adjusted by the addition of
saturated sodium acetate to the test solution.

The intensity of the colored complex was measured with
a Beckman DU spectrophotometer. The sample was measured
against a reagent blank and the amount of iron determined by
comparing the reading to that of a known concentration of

iron, the iron standard.

63

Solutions for Iron Spectrophotometry

 

To prevent iron contamination, acid wash all glassware
with nitric or hydrochloric acid, avoid the use of metal
instruments, use only distilled or deionized water for
making solutions and rinsing glassware, and purchase metal
free reagents.

Iron Standard Solution

0.0010 mg Fe/ml (1.0 ug/ml) purchased from G.
Frederick Smith Chemical Co., Columbus, Ohio.

Trichloroacetic Acid - CC13COOH
To 1.1 lb metal—free trichloroacetic acid (G.
Frederick Smith Chemical Co., Columbus, OH) add
metal—free water to the 500 ml mark. This makes a
100% solution.

Mercaptoacetic Acid - (Thioglycolic acid) HSCHZCOOH
98% solution purchased from.A1drich Chemical Co.,
Milwaukee, Wisconsin.

Precipitating/Reducing Solution
To 20 ml of trichloroacetic acid solution add 1 ml
mercaptoacetic acid and dilute to the 100 ml mark
with metal-free water.

Saturated Sodium Acetate - NaCZH302'3H20

To metal—free water stirred with a magnetic
stirrer add sodium acetate (G. Frederick Smith
Chemical Co., Columbus, OH) until it no longer

goes into solution. Decant off the saturated

64

solution.

30% Sodium Acetate - NaC2H302- 3H2

To 30 gms sodium acetate add metal-free water to

O

the 100 ml mark.
Color Reagent in Saturated Sodium Acetate - 2,4-
bis(5,5-diphenyl,1,2,4-Triazin-3-yl)pyridine
tetrasulfonic acid, tetra sodium salt (DBTPS)
Dissolve 200 mg DBTPS (G. Frederick Smith Chemical
Co., Columbus, OH) in 5 ml metal-free water; add
saturated sodium acetate solution to the 100 m1
mark.
Color Reagent in 30% Sodium Acetate - DBTPS
Dissolve 200 mg DBTPS in 5 m1 metal—free water;
add 30% sodium acetate solution to the 100 m1

mark.

65

Determination of Serum Iron*

Serum samples were brought to room temperature. 500 pl
of serum sample, 500 pl of metal-free water, and 1 m1 of
precipitating/reducing agent were pipetted into an acid-
cleaned test tube, mixed well by inversion, and allowed to
stand 10 minutes. The sample was spun at 3000 RPM for 10
minutes, and a 1 m1 aliquot of the supernatant pipetted into
a 2 m1 autoanalyzer cup. (These cups have been found to be
quite free of iron.) Standard and blank solutions were
prepared by pipetting 500 pl each of metal-free water and
iron standard into separate cups and adding 500 pl preci-
pitating/reducing agent to each. The serum sample, reagent
blank, and iron standard solution were each treated with 300
pl BDTPS in 30% sodium acetate. After 5 minutes, the absor-
bance of the standard and of the sample were determined at
565 nm against the reagent blank solution.

Serum iron in ugFe/ml =

Absorbance of sample

ABSOrbance of standard X dilution factor

Dilution factor = 2 because serum was diluted in an
equal volume of metal-free water.

Example:

Absorbance of sample = 0.30

Absorbance of standard (1 ul/ml) = 0.22

:19 X 2 = 2.73 ugFe/ml

.22
*(Zak, Baginski and Epstein, 1980)

66

Preparation of Liver Extract

 

The liver sample was thawed, weighed, and placed in an
acid-cleaned homogenizing tube to which was added an equal
weight of iron-free water. The liver sample was homogenized
by a motor-driven teflon homogenizer until a fine suspension
was formed. A 0.2 gm sample of the homogenate was placed
into a preweighed acid-cleaned test tube, and 0.8 gms of
iron-free water were added. The tube was then corked and
refrigerated for at least 30 minutes.

To promote cell lysis, the tissue was heated and then
recooled. The cork was first replaced with a wad of clean
cotton wool, and the sample heated in a dry bath 10 minutes
at 90°C and then cooled in an ice bath 10 minutes. The
sample was reweighed and metal-free water added to correct
for any water loss through evaporation. 500 pl of 8.4 N HCl
were added, the weight recorded, and the sample heat ex-
tracted in a dry bath at 90°C for 60 minutes. After cooling
to room temperature, the sample was weighed and metal-free
water added to correct for any water loss through evapor-
ation. The sample was then spun 10 minutes at 3000 RPM and
filtered. Filters were prepared by placing a plug of
Whatman #2 filter paper in the tapered tip of a Pasteur
pipette. The sample was drawn up through the filter with a
pipet aid. The tip below the plug of filter paper can be
broken off if desired.

67
The sample was transferred to an acid-cleaned test

tube, corked, and refrigerated until analyzed.

68
Determination of Liver Iron

 

50 p1 of liver extract, 450 p1 metal-free water, and
500 pl precipitating/reducing agent were pipetted into an
autoanalyzer cup. Blank and standard solutions were pre-
pared by pipetting 500 pl metal-free water and 500 pl iron
standard into separate autoanalyzer cups and adding 500 p1
of precipitating/reducing solution to each. All solutions
were allowed to sit for 10 minutes after which 300 pl BDTPS
in saturated sodium acetate were added. After 10 minutes
the absorbances of the standard and sample were determined
at 565 nm against the reagent blank solution.

Liver iron in pg/gm wet weight:

 

Absorbance of sample X extraction X sample
dilution dilution
Absorbance of standard factor factor

Extraction dilution factor = 15 because the iron from
0.1 gm of liver was diluted in 1.5 gms of solution.
Sample dilution factor = 10 because 50 pl of sample was
diluted in 450 p1 metal-free water.
Example:
Absorbance of sample = .40
Absorbance of standard = .22 (1 pl/ml)

.40

——- X 15 X 10 = 272.7 nge/g wet weight
.22

69

Preparation of Brain Extract

 

Each frozen brain was cut coronally on a cryostat into
270 pm slabs which were placed on clean microscope slides,
warmed only long enough to promote adherence and them
refrozen. The knife was coated with silicone stopcock
grease and the excess wiped off. This prevented metal
contamination of the tissue from the blade. While viewing
the slabs with a dissecting microsc0pe, the desired areas,
globus pallidus, substantia nigra, and cortex, were removed
with a microcoring device (Palkovits, 1973) which removed
cylinders of tissue approximately 0.7 mm in diameter and
270 pm thick. The microcoring device was prepared from a 19
gauge hypodermic needle from which the pointed end was
ground off making a flat-ended cylinder. The tip of the
cylinder was sharpened on a grinding stone and wiped clean.
To prevent metal contamination of the sample, the needle
was coated with silicone stopcock grease and the excess
wiped off. The needle was attatched to a micrOpipetting
tube with a mouthpiece. A small wad of disposable wiping
tissue was placed in the tubing to prevent saliva from
entering the sample; this plug was changed frequently.
The desired sample was cored out of the slab of tissue on
the microscope slide with the sharpened tip of the needle
and then blown out of the microcoring needle into an acid-

cleaned homogenizing tube.

70

The microscope slide was kept frozen during coring by
placing it on a metal block kept cold in an alcohol/dry ice
bath.

The cores were collected in preweighed acid-cleaned
homogenizing tubes and cores from the globus pallidus and
substantia nigra of a brain placed in one tube, the cortex
sample into another. It was necessary to pool the globus
pallidus and substantia nigra in order to obtain a suf-
ficiently large tissue sample. The boundaries of the
globus pallidus and substantia nigra were determined using
as a reference and guide a set of Perl's-DAB stained,
coronally cut sections of rat brain. The cortex sample
came from the area dorsal to the globus pallidus.

The homogenizing tube was reweighed after coring and
the weight of tissue collected determined.

150 pl of metal-free water were pipetted into the
homogenizing tube and the spherical ground glass homogenizer
rotated by hand until the tissue was finely ground. Another
150 pl of metal-free water were pipetted down the walls of
the tube and the homogenizer to wash adhering tissue into
the solution. The sample was transferred into an acid-
cleaned test tube, weighed, corked, and refrigerated for at.
least 30 minutes.

To promote cell lysis, the tissue was heated then
recooled. The cork was first replaced with a wad of clean
cotton wool and the sample heated in a dry bath at 90°C for

10 min then cooled 10 min in an ice bath. The sample was

71
reweighed and metal-free water added to correct for any
water loss through evaporation. 150 pl of 8.4 N HCl were
added; the weight recorded; and the sample heat extracted in
a dry bath at 90°C for 60 min. After cooling to room
temperature, the sample was weighed and metal-free water
was added to correct for any water loss through evaporation.
The sample was spun 10 min at 1000 RPM and filtered. The
extract was placed in an acid-cleaned test tube and refrig-

erated until analyzed.

72

Determination of Brain Iron

100 p1 of brain extract, 100 pl saturated sodium
acetate, 300 p1 metal-free water, and 500 pl precipitat-
ing/reducing agent were pipetted into an autoanalyzer cup.
Blank and standard solutions were prepared by pipetting 500
pl metal-free water and 500 p1 iron standard into separate
autoanalyzer cups and adding 500 p1 of precipitating/re-
ducing solution to each. All solutions were allowed to sit
for 10 min after which 300 pl BDTPS in saturated sodium
acetate are added. After 10 min the absorbances of the
standard and the sample were determined at 565 nm against
the reagent blank solution.

Brain iron in pg/gm wet weight:

 

Absorbance of sample X extraction sample
Absorbance of standard dilution diIUtion
factor factor

The extraction dilution factor of a sample is the number of
times the iron in a sample is diluted by the added iron-free
water and HCl. Tissue collected usually weighs between
0.010 and 0.012 gms and the iron in this tissue was diluted
in 450 pl of liquid. For 0.012 gms of tissue, the extrac-
tion dilution factor is 37.5. Sample dilution factor equals
5 because 100 pl of extract was diluted with 100 p1 of

saturated sodium acetate and 300 pl metal-free water.

73

Example:

Absorbance of brain sample = 0.06

Absorbance of iron standard (1.0 pl/ml) = 0.22
Lgé X 37.5 x 5 = 51.1 nge/g wet weight
.22

74

Analysis of Data

 

Analysis of variance was used to compare the means of
estrous cycle, sex difference, pregnancy, effects of gonad-
ectomy, and effects of estrogen implants data. All data met
the assumption of variance homogeneity, as tested by maximum
variance divided by minimum variance. Tukey's test was
performed on the combined estrous cycle and sex difference
data in order to compare all possible pairs of estrous cycle
stages and to compare the male with each stage of the
estrous cycle. Regression analysis coupled with analysis of
variance was done on pregnancy data. Designed contrasts,
comparing day 4 of pregnancy to the other days of pregnancy,
were also performed. Orthogonal contrasts tested the
effects of gonadectomy and estrogen implants data. For ease
of computation, contrasts were performed among treatment
groups with equal replication. Excess replicates were
deleted on a random basis. In those instances where samples
were deleted because of unequal replication, analysis of
variance was performed on the total sample size, and it was
found that there was little or no change in the means or
standard errors of the groups, nor in the F values and
associated probabilities in the analysis of variance.
Statistical methods are those of Gill (1978) and Bliss
(1967).

RESULTS

HISTOCHEMISTRY

The Perl's-DAB treated tissue was chosed for the
photography in this section because it intensified the
Perl's reaction manyfold and permits the use of thin sec-
tions. A 40 um section stained with DAB demonstrates the
presence of iron, even in areas of the brain in which the
iron concentration is low, and gives a detailed resolution
‘of the location of iron at the cellular level. DAB has the
added advantage of being highlighted by dark-field micros-
copy. Thus, we have a method which not only yields a clear
view of iron-containing structures for light-field micro-
scopy, but also in which interference from counterstaining
can be removed with dark-field microSCOpy. The photo-
graphs presented here were taken with both dark-field
and light-field illumination. The intensification step
does not demonstrate the presence of iron in any area in
which the Perl's method alone does not indicate that iron
is present and no staining was apparent with DAB without

pretreatment with ferrocyanide.

Study 1 a. Localization of iron.in the rat brain.

Iron is unevenly distributed in the brain, occurring in
many different areas and in differing amounts (See Figures
1 through 8), and is present in different types of cells

and structures in these areas.

75

76

.m.q x cowumowmwcwwz .umn oHoEom nHo xooB
mm m mo Sow> Hmuuwwmmmumm .aonfi mo mGOHumHSEDoow ohm mmon pocfimum

.cflmumuoucsoo on .coufl How awmum m<muw.HHom .couw mo coaunnwnumwn .H onswwh

msemamnu Hmuuco> mamas mwucmumnSm measofiaaoo Hoanmnsm

 

oaouonau huouowmao

Escaaamm Hmuuco>

 

 

.Ilmnmaosu >uo>HHo uoauonnm

/ mfimfiufifi Hangmm

7

msoaona umaaocawomoucm

mseaaama apnoea
mnoaosc

coawusaloumwsmo msmaona Hmawsmwfiuu Hmawmm

/

uomuu muouomeo maoaosc HoHHonouoo

 

TIIIIIIm5H50HHHoo uoauowdH

can: snouummao i ‘

_-.~— . .

g___

 

xouuoo msaamooamfim

Figure 2.

Figure 3.

Figure 4.

Figure 5.

77

Distribution of iron in the forebrain at the

level of the anterior commissure in a 32 week
old female rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

Distribution of iron in the forebrain at the
level of the anterior commissure in a 32 week
old male rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

Distribution of iron in the forebrain at the
level of the globus pallidus in a 32 week
old female rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

Distribution of iron in the forebrain at the
level of the globus pallidus in a 32 week
old male rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

78

Figure 2. . ' FEMALE

lCortex
Cingulate cortex
Lateral septal nucleus
Caudate—putamen nucleus
Bed nucleus of the
stria terminalis
Anterior commissure
Ventral pallidum
Organum vasculosum of
the lemina terminalis
Island of Calleja
Olfactory tubercle

 

Figure 3. MALE

Anterior commissure
Organum vasculosum of
the lamina terminalis
Ventral pallidum
Island of calleja
Olfactory tubercle

Figure 4. FEMALE

Cortex

Cingulate cortex
Choroid plexus
Caudate-putamen nucleus
Globus pallidus

 

Figure 5. MALE

Globus pallidus

 

Figure 6.

Figure 7.

Figure 8.

79

Distribution of iron in the forebrain at the
level of the thalamus in a 32 week old female
rat. Perl's-DAB, no counterstain. Stained areas

are accumulations of iron. Magnification X 5.4.

Distribution of iron in the midbrain at the
level of the substantia nigra in a 32 week old
female rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

Distribution of iron in the midbrain at the
level of the substantia nigra in a 32 week old
male rat. Perl's-DAB, no counterstain.
Stained areas are accumulations of iron.

Magnification X 5.4.

80

Figure 6. FEMALE

. Cortex
Cingulate cortex
Dentate gyrus
Hippocampus
Lateral habenular
nucleus
Ventral thalamus
Caudate—putamen
nucleus
Central amygdalar
nucleus
Entopeduncular nucleus
Ventro-medial
hypothalamus

 

Figure 7. FEMALE

Cortex
Superior colliculus
Central grey
Reticular formation
Medial geniculate
nucleus
Dentate gyrus
Hippocampus
Oculomotor nucleus
Substantia nigra
Interpeduncular
nucleus

 

Figure 8. MALE

Superior colliculus
Dentate gyrus
Oculomotor nucleus
Substantia nigra
Interpeduncular
nucleus

 

81

In brain tissue iron occurs in granules, singly or in
clumps; amorphous accumulations; branching fiber-like pro-
cesses; fibers; and on or in glial cells and neurons.
Whether the iron is within the axoplasm or within the myelin
sheath of the fibers cannot be determined from this mater—
ial. When iron is found in association with cell bodies, it
either 1) completely fills the cytoplasm of the cells (com-
monly glial cells), 2) occurs as bouton-like structures
which appear to encrust the surface of the perikaryon and
neuronal processes of nerve cells, or 3) appears as a fine
dusting of small grains within the perikaryon of neurons.

In the following description, the iron accumulating
areas will be organized more with respect to the type of
cellular organization of the iron-containing structures than
to the subdivision of the brain to which the site belongs.

Iron is present in those parts of the brain in which
the blood-brain barrier is absent. Iron-appears here as
extracellular granules andamorphous accumulations, fre-
quently in association with blood-vessels, and also in
branched angular figer-like processes. Sometimes iron is
so highly concentrated that determining its structural
localization is difficult. These areas include the pineal,
the pituitary, area postrema (Figure 9), and the choroid
plexus (Figures 4 and 10).

The circumventricular areas and some peripheral areas
of the brain, some of which are not protected by the blood-

brain barrier, also accumulate iron in granules.

Figure 9. Photomicrograph of iron accumulation in area
postrema and dorsal to the central canal.
Darker staining areas are accumulations of iron.
Perl's-DAB counterstained with thionin.

Magnification X 100.

Figure 10. Photomicrograph of iron accumulation in the
subfornical organ and choroid plexus.
Darker staining areas are accumulations of iron.
Perl's-DAB counterstained with thionin.

Magnification X 100.

83

Figure 9.

Area postrema

Iron accumulations

Central canal.
Vagus nucleus

Hypoglossal
nucleus

 

Figure 10.

Subfornical
organ

Choroid plexus

 

 

 

84

amorphous accumulations and/or fiber-like extensions. These
areas include the subfornical organ (Figure 10), the epen-
dyma of the ventricles of the brain, the lateral edge of the
Optic tract (Figure 11), the organum vasculosum of the
lamina terminalis (Figures 2 and 3), and in many areas of
the ventro-medial hypothalamus (Figure 6). Within the
ventro-medial hypothalamus and median eminence, iron occurs
in the tanycytes (Figures 12, 13 and 14) and in granules,
clumps of granules, and fibers (Figures 12, 13 and 14). The
granules are similar in size and distribution to the gran-
ules stained by the Bargmann modification of the Chrome-
Hematoxylin method for neurosecretion (Pearse, 1961). Iron
also may be seen outlining the capillaries of the median
eminence and lateral hypothalamus.

The rest of the brain is protected by the blood-brain
barrier, and in these areas the sites which accumulate the
most iron are the globus pallidus including the entopedun-
cular nucleus (Figures 1, 2, 4 and 5), ventral pallidum
(Figures 1, 2 and 3), islands of Calleja (Figures 2 and 3),
substantia nigra (reticular zone) (Figures 1, 7 and 8), the
interpeduncular nucleus (Figures 7 and 8), and the deep
nuclei of the cerebellum (Figure 1). By the ventral pal-
lidum, I refer to the precommissural ventral anterior
extension of pallidal tissue into the olfactory tubercle
which is broken up into finger-like extensions by the
medial forebrain bundle and which ends in the islands of

Calleja (see Figures 1, 2, 3 and 15). Even in those

85

Figure 11. Photomicrograph of iron—filled fibers in the
lateral edge of the Optic tract. Iron = brown.
Perl's-DAB counterstained with thionin.

Magnification X 200.

Figure 12. Dark-field photomicrograph of iron-filled
tanycytes and clumps of granules in the
ventro-medial hypothalamus. Iron = white.
Perl's-DAB counterstained with thionin.

Magnification X 120.

 

86

Figure 11.

Iron—filled fibers

 

 

Figure 12.

   
  
 

Third ventricle
:rgzr Iron-filled tanycytes

, Ventro—medial
hypothalamus

1}, Clumps of iron—filled

'.? granules

Figure 13.

Figure 14.

87

Dark-field photomicrograph of iron-filled

structures in the ventro-medial hypothalamus
and arcuate area. Iron = white. Perl's-DAB
counterstained with thionin. Magnification

X 120.

Dark-Field photomicrograph of iron-filled
structures in the ventro-medial hypothalamus
and median eminence. Iron = white. Perl's-DAB
counterstained with thionin. Magnification

X 120.

88

Figure 13.

Iron—filled tanycytes

Third ventricle

Clumps of iron—filled
granules

Ventro—medial
hypothalamus

Arcuate nucleus

Iron—filled fibers

 

Figure 14.

Third ventricle

Iron—filled tanycytes

Iron—filled fibers
Median eminence

Iron—filled granules

 

Iron-filled tanycytes

Figure 15.

Figure 16.

89

Photomicrograph of iron distribution in the
ventral pallidum, islands of Calleja and
Olfactory tubercle. Iron = brown. Perl's-DAB
counterstained with thionin. Coronal view.

Magnification X 82.

Photomicrograph of iron-filled cells and
fibers in the globus pallidus. Iron = brown.
Perl's-DAB counterstained with thionin.

Magnification X 900.

 

90

Figure 15.

Ventral pallidum

Island of Calleja

Olfactory tubercle:
_Polymorph zone

Pyramidal zone

 

Figure 16.

Iron—filled

glial cells

Iron-filled

fine fibers

 

91
conditions in which the iron concentration Of the brain is
low, iron is usually visible in the above areas of the
brain. In all of these areas, iron is localized in glial
cells and in fibers--see globus pallidus (Figure 16),
ventral pallidum (Figure 17), and substantia nigra (retic-
ulata) (Figure 18):' The glial cells are like oligodendro-
cytes in appearance, and iron fills the cytoplasm obliter-
ating internal cellular detail. Some iron-filled cells are
larger than typical glial cells and may be small neurons
(Figure 19). In Perl's stained sections of these areas,
the background neurOpil is a pale blue wash. The DAB
intensification demonstrates that iron in the neuropil
is present in fibers (see Figures 15 through 21). In the
globus pallidus and continuing into the adjacent striatum,
iron is seen in bundles of fine fibers (Figures 1,2 and 4 -
low magnification, Figure 20 - high magnification. In
the ventral pallidum, thick fibers are more common, and
"cascades" of thick fibers are seen running in a dorso-
ventral direction in the plane of frontal sections (Figure~
21). In the substantia nigra (reticulata), the iron—filled
fibers are of medium size, and in the cerebellum, the fibers
are from medium to large in size.

Except for the most medial Olfactory tubercle near the
islands Of Calleja, iron-containing granules are rarely
found in these areas. Within the islands of Calleja, iron
is within fibers in the core and both in and around the

granule cells (Figure 15). Whether the iron is within the

92

Figure 17. Photomicrograph of iron-filled cells and
fibers in the ventral pallidum. Iron = brown.
Perl's-DAB counterstained with thionin.

Magnification X 476.

Figure 18. Photomicrograph of iron-filled cells and
fibers in the substantia nigra. Iron = brown.
Perl's-DAB counterstained with thionin.

Magnification X 476.

Figure 17.

Iron—filled

glial cells

  
 
  
 

-’Varicose fiber
Iron—filled

fibers

Figure 18.

'Iron—filled

glial cells

\

Iron—filled

fibers.

Figure 19.

Figure 20.

94

Photomicrograph of small neuron-like iron-
filled cells in the globus pallidus.
Iron = brown. Perl's-DAB counterstained with

thionin. Magnification X 900.

Dark-field photomicrograph of bundles of fine
iron-filled fibers in the globus pallidus.
Iron = white. Perl's-DAB counterstained with

thionin. Magnification X 1000.

95

Figure 19.

 

Small neuron—like iron—filled cells

Figure 20.

éé.
a:
.1

 

Bundles of fine iron—filled fibers

96

.GNN x coaumuamacmmz .caaoaeu spas
nonwmumhoucsoo m<mum.Huom .oufisz u OOHH .wcounoc Hmudom HonoumH

Cw no so mmHDuOsuum oxwaucouson Ca some mo smouwouowaouosm nHOfiMIJHmQ .NN ouswfim

.onq x soHumoamacwmz .sacoanu nuns umaamumumussoo
m<n:m.auom .OBOHQ n GOHH .OHOuOOSO Shouommflo paw EDnHHHma

Hmuuco> ozu a“ doom muonfim noHHwMucOufl mo :moanmmo. mo amonwOHOflBouogm .HN ounwfim

mcousos Hauaom Honouma

5H so no mousuosuum oxfialnouson 6H GOHH mponam nOHHHMIGOHH mo .monmommos

97

 

HN unseen

.NN unawam

98
thin rim of cytoplasm of these cells or on the surface
cannot be determined from this material.

A similar distribution of iron in glia and fibers is
seen in many other parts Of the rat brain. However, iron
occurs in a much lower concentration, is frequently present
in the walls of blood vessels, and is apparent in Perl's
stained sections only under Optimal conditions. These areas
include the olfactory bulb where iron occurs in periglomer-
ular areas, in olfactory tracts, and in glial cells among
the granular cells (Figure l), the caudate, putamen, and
accumbens nuclei (Figures 1, 2, 4 and 6), the ventral tier
of thalamic nuclei (Figure 6), the lateral habenular nucleus
(Figure 6), inferior colliculus (Figures 1 and 7), and, to a
lesser degree, superior colliculus (Figures 1 and 7),
reticular formation (Figures 1 and 7), medial geniculate
nucleus (Figure 7), brachium of the inferior colliculus,
lateral lemniscus, oculomotor nucleus (Figures 7 and 8),
facial nucleus (Figure 1), and superior Olive (Figure 1).

In the midbrain central grey, iron is localized in glia,
blood vessels, and clumps of granules (Figure 7).

In other low iron areas, iron is not present in glia
and fibers but within bouton-like encrustations which appear
to be on the surface of the perikaryon and neuronal proces—
ses of nerve cells. This type of iron accumulation occurs
frequently in the lateral septum (Figure 22, see also Figure
2), bed nucleus of the stria terminalis (Figure 23, see also

Figure 2), and occasionally is seen in the ventral pallidum

99

.oom x coaumoamacwmz .caaoasu sues
cosfimumuoucsoo m<mum.HHom .SBOHn u OOHH .mcousoc Hmnfifiamm Hmhuco>

aw no so monouosnum oxflaucouson cw con“ mo SQmeOHOHBOuonm .qm ouswflm

.omm x coHumonecwmz .sacoaru sues emsemumumussoo
m<le.H.H®nm .QUHSQV H COHH .mGOHDGG WHHMCHEHGU GHHum m5“ MO mDmHUDG U03

aw so so mousuosuum oxwauaounon OH coma wo smmeOHOHEOOO£a paOfiMuxumm .mm ouswflm

100

mcousoc Hmvflaama Hmuuco> OH no Go

m0H5u03HUW QMHHIGOUSOQ GH GOHH

.GN magmas

mnousoc mfiHmaHEhou mfiuum mau mo wsoaosn non

CH HO CO MQHDUUDMHM QMHHICOUDOS CH fiOHH

 

.mN unseen

101
(Figure 24), diagonal band of Broca, lateral to the organum
vasculosum of the lamina terminalis in the medial preoptic
areas, lateral to the anterior hippocampus, and in the
cortex layers 2, 3 and 5 (Figures 2, 4, 6 and 7).

In some low-iron areas of the brain, a fine dusting of
iron-containing grains, apparently distributed within the
cytoplasm of neurons, is visible. This type of distribution
is sometimes seen in the paraventricular nucleus, supraoptic
nucleus (Figure 25), suprachiasmatic nucleus (Figure 26),
dentate gyrus of the hippocampus (Figure 1, 6, 7 and 8)
cingulate cortex (Figures 2, 4 and 6), pyramidal cells of
the olfactory tubercle (Figures 2 and 3), and in the central
amygdalar nucleus (Figure 6).

Iron is found in the hippocampus adjacent to the cells
of CA 3 where the granule cells of the dentate gyrus
terminate in the mossy fibers (Figure 6). Here iron is

present in stellate-shaped structures.

Study 1 b. Brain iron and age.

The Perl's reaction is mild to absent in the brain
tissue from rats 12 weeks of age or younger. Whether the
DAB step will permit the localization of iron in the brains
of younger animals is not yet known. The intensity of the
iron stain increases with increasing age of the subjects,
and the oldest animals have the darkest stain. Although the
iron concentration increases with age in the rat brain, the

relative intensity among brain areas and the distribution of

Figure 25.

Figure 26.

102

Dark-field photomicrograph of fine grains of
iron in cells of the supraoptic nucleus.
Iron = white. Perl's-DAB counterstained with

thionin. Magnification X 216.

Dark-field photomicrograph of fine grains of
iron in cells of the suprachiasmatic nucleus.
Iron = white. Perl's—DAB counterstained with

thionin. Magnification X 200.

103

’Figure 25.

 

Iron in supraoptic nucleus neurons

Figure 26.

 

Iron in suprachiasmatic nucleus neurons

104
iron does not appear to change; as animals get older, the
presence of iron in low iron areas becomes evident.

Due to a sex difference, which will be discussed in the
following section, only within sex comparisons of iron con-
centration and age will be made. A quantitative, spectro-
photometric determination of age effects on iron concentra-

tion is the object of a future study.

Study 1 c. Sex difference in brain iron.

The iron reaction in the brains of female rats is
darker than the reaction in the brains of male rats.
Figures 2 and 3, 4 and 5, and 7 and 8 compare the iron
concentration of female and male rats 32 weeks of age in the
high-iron areas of the telencephalon and mesencephalon.
These are: globus pallidus, ventral pallidum, substantia
nigra, and interpeduncular nucleus. Whereas the female
brain produces a more intense iron reaction that the male,
the pattern and distribution of iron appears to be the same
in both sexes. Male siblings raised under the same con-
ditions and sacrificed at the same age will have very
similar reactions. However, female siblings raised under
the same conditions and sacrificed at the same age will have
a range of reaction intensities. Among a group of rats 12-
18 weeks of age, the strongest reactions are from.female
tissue; however, the milder reactions of the females are
only as intense as the male reaction. Among 32 and 43 week

old rats, variability also occurs in the intensity of the

105

stain of the female brain tissue; however, the tissue from
the older animal stains darker than the tissue from the
younger animal (as described above in the effects of age).
Vaginal smears were not taken at sacrifice so that it could
not be determined if the variability among the females was

correlated with the estrous cycle.

Study 2. Brain iron and pregnancy.

The treatments are ranked below in order of increasing
intensity of the Perl's reaction.

1. 18 week Old postpartum females.

2. 32 week old 16 days pregnant females.

3. 18 week old 7 days pregnant females.

The reaction Of the last group in the globus pallidus
and substantia nigra is the most intense reaction observed
in any age or treatment group as stained by the Perl's
method alone. Brain iron apparently increases early in
pregnancy and then decreases during the term of pregnancy.
No obvious differences occur in the distribution of iron in
the brain or in the relative intensities among brain areas

during pregnancy.

Study 3. Ovariectomy and ovariectomy plus estrogen treat-
ment effects on brain iron.
The treatments are ranked below in order of increasing

intensity of the Perl's reaction.

106

1. 18 week Old females ovariectomized at 10 weeks,
and 18 week Old females ovariectomized at 10 weeks
and received an estrogen implant at 17 weeks.

2. 32 week old females ovariectomized at 28 weeks.

3. 32 week Old females ovariectomized at 8 weeks and
received an estrogen implant at 30 weeks.

The 18 week Old females which had been ovariectomized
at 10 weeks of age have a brain iron concentration about the
same as a male or lightly-staining female of the same age.
Ovariectomized females of the same age which had received an
estrogen implant at 17 weeks of age do not appear to differ
from any of the above groups. Thirty-two week old females
which had been ovariectomized at 28 weeks stain about as
darkly as males and lightly-staining females of the same
age. However, the 32 week old ovariectomized females which
had received an estrogen implant at 30 weeks are visibly
darker than any Of the groups mentioned here and nearer in
intensity to that reached by 18 week old rats in the first
week of pregnancy.

Neither ovariectomy nor ovariectomy plus estrogen
implant visibly changes the distribution or relative inten-
sity of brain iron between the high iron concentrating areas
within each treatment group: no area or areas in the brain
are affected to any greater degree by these treatments as

can be determined using the Perl's method.

107
Study 4. Castration and castration plus estrogen treatment
effects on brain iron.

No differences in the intensity of the Perl's reaction
are visible between castrated males and intact males of the
same age or between the castrated males and castrated males
which had received the estrogen implant. There is also no
apparent difference in the patterns of distribution or the
relative intensities between the various high iron areas of

the brain.

108
SPECTROPHOTOMETRY
Study 1 - to determine (1) if the levels of brain iron
change throughout the estrous cycle and (2) if
a sex difference occurs in brain iron levels.

The iron concentration of the pooled globus pallidus
and substantia nigra (GP+SN) fluctuates throughout the
estrous cycle with the highest concentration of iron oc-
curring at proestrus (Figure 27). The lowest concentration
of iron occurs at metestrus with estrus and diestrus having
intermediate levels. Proestrus is significantly greater
than diestrus, estrus, or metestrus and diestrus, estrus
and metestrus do not differ significantly from each other
(Table 4). The control male (GP+SN) iron level falls between
than of control females in metestrus and estrus and differs
significantly only from females in proestrus.

The iron concentration of the cortex does not vary
significantly thrOughout the estrous cycle (Figure 27 and
Table 4), and the cortex iron concentration of the control
male does not differ from the females in any stage of
estrous.

Although the liver iron concentration fluctuates
throughout the estrous cycle (Figure 27), the pattern is not
the same as in the GP+SN. The diestrus liver iron concen-
tration is as high as the proestrus iron level with met-
estrus levels lowest and estrus levels intermediate: none
of these differences are significant. The control male

liver iron concentration is significantly lower than the

Figure 27.

109

Iron concentration of liver, pooled globus
pallidus and substantia nigra, cortex and
serum of control males and females in each
of the four stages of the estrous cycle.

Mean values 1 l S.E.M., see ANOVA tables

for sample sizes.

 

Figure 27
22 24o~
0‘“.
mm
E'- 230-
2.3
lJ-‘I
3.30
C235 220'-
2?.
:gw 210—45
E
'5 a.
f—\
g zoo-35
25
15

 

nge/mliS.E.M.
w
o

110

Liver

 

  

Globus Pallidus +
Substantia Nigra

 

L l l 1

Control Diestrus Proestrus Estrus Metestrus

d 9 9 Q 9

111

Table 4. Analysis of Variance and Tukey's Test of Iron

Measures of Estrous Cycle and Sex Difference Data

 

 

 

 

Source DF ‘MS F P
GP+SN

Total 34

Between 4 768.17 6.74 <0.01
Within 30 113.99

CORTEX

Total 33

Between 4 16.91 0.44 NS
Within 29 38.36

LIVER

Total 39

Between 4 34,142.49 99.51 <0.001
Within 35 343.12

SERUM

Total 34

Between 4 0.83 1.60 NS
Within 30 0.52

Tukey's Test*

GP+SN

 

METESTRUS MALE ESTRUS DIESTRUS PROESTRUS
15.40i2.49 20.14i1.60 20.85:2.6l 24.71i5.10 42.50i6.31

 

 

LIVER
MALE METESTRUS ESTRUS DIESTRUS PROESTRUS
75.46i5.36 205.15i7.39 215.27i7.67 228.55i6.11 230.98i5.09

 

*The means i l S.E. subtended by the same line do not differ
from each other at p=0.05.

112

females in every stage of estrus: the level is only about
1/3 that of the females (i.e., 75.46 pg Fe/g wet weight) and
has not been included in Figure 27 because of the magnitude
of the difference.

Although the serum iron concentration fluctuates
slightly throughout the estrous cycle, with the highest
level occurring at estrus (Figure 27), none of the dif-
ferences are significant. The control male serum iron
level, although lower, does not differ significantly from

the females in any stage of estrus.

113
Study 2 - to determine if brain iron levels change during

pregnancy.

A significant rise in the iron concentration of GP+SN
occurs from day 4 to day 8 of pregnancy (Figure 28 and Table
5). By day 12, the iron concentration of GP+SN has fallen
to about day 4 levels where it is maintained, with slight
but not significant decreases, until day 20 of pregnancy.

At 15 to 24 hours postpartum (PP), the iron level is slight-
ly, but not significantly, increased.

Fluctuations in the iron concentration of the cortex
follow a similar pattern (Figure 28); however, the changes
are not significant (Table 5).

Liver iron concentration decreases throughout pregnancy
(Figure 28 and Table 5). Regression analysis demonstrates
that the curve has a significant negative slope with a
significant non-linear component, a variation in the rate of
decrease (prediction equation, Y= 7.85X + 355.63). Liver
iron concentration is greater at day 4 of pregnancy than on
any other day of pregnancy tested. At PP, the iron concen—
tration of the liver is even less than at 20 days, dropping
to about 1/2 of the concentration of day 4 of pregnancy.

The pattern of liver iron change during pregnancy differs
from that seen in GP+SN where an initial rise is followed by

a drop after which levels do not change until PP.

114

Figure 28. Iron concentration of liver, pooled globus
pallidus and substantia nigra, cortex and
serum every four days throughout pregnancy
and 15—24 hours postpartum. Mean values i
l S.E.M., see ANOVA tables for sample

sizes.

 

MTfiOTAL NON-HEME IRON

 

Figure 28

LIJ

US 300

E

I 250

S2

LIJ

3 200

I-

§

2’ 150

0|

3.
100
S
u!
(D
E!
E
B
U.
U)
3.

 

115

 

  
 
  

 

 

T
"'45
Globus
Pallidus-+-
L 35 Substantia
nigra
-25
. Cortex
15.,
41
3 W
Serum
2
1
o' 4 8‘ 12 16 20 15-24 m.
post partum

DAY OF PREGNANCY

116
Table 5. Analysis of Variance with Regression of Iron

Measures of Pregnancy Data

 

 

 

 

Source DF MS F P
GP+SN

Total 35

Between 5 129.67 1.99 NS
Regression l 86.50 1.33 NS
Residual 4 140.46 2.16 NS
Q 1. day 4 vs day 8 1 407.75 6.27 <0.05
Q 2. day 4 vs day 12 l 32.37 0.49 NS
Q 3. day 4 vs day 16 l 4.85 0.07 NS
Q 4. day 4 vs day 20 1 6.97 0.10 NS
Q 5. day 4 vs PP l 13.37 0.20 NS
Within 30 65.01

CORTEX

Total 35

Between 5 90.79 2.08 NS
Regression l 17.50 0.40 NS
Residual 4 109.11 2.50 NS
Q 1. day 4 vs day 8 1 38.05 0.87 NS
Q 2. day 4 vs day 12 1 82.42 1.88 NS
Q 3. day 4 vs day 16 1 65.66 1.50 NS
Q 4. day 4 vs day 20 l 60.66 1.39 NS
Q 5. day 4 vs PP l 16.42 0.38 NS
Within 30 43.66

LIVER

Total 35

Between 5 20,237.97 9.41 <0.001
Regression l 69,384.30 32.26 <0.001
Residual 4 7,951.31 3.70 <0.05
Q 1. day 4 vs day 8 1 10,103.02 4.70 <0.05
Q 2. day 4 vs day 12 1 9,610.11 4.47 <0.05
Q 3. day 4 vs day 16 l 36,695.97 17.06 <0.001
Q 4. day 4 vs day 20 l 45,660.47 21.23 <0.001
Q 5. day 4 vs PF 1 79,059.58 36.76 <0.001
Within 30 2,150.39

Table 5 continued.

117

 

 

 

Source DF MS P
SERUM

Total 35

Between 5 2.87 11.96 <0.001
Regression 1 2.85 11.87 <0.001
Residual 4 2.87 11.97 <0.001
Q 1. day 4 vs day 8 l 0.24 1.00 NS
Q 2. day 4 vs day 12 1 3.16 13.17 <0.005
Q 3. day 4 vs day 16 l 0.00 0.00 NS
Q 4. day 4 vs day 20 1 9.36 38.92 <0.001
Q 5. day 4 vs PP 1 0.34 1.42 NS
Within 30 0.24

118

Regression analysis of the serum iron concentration
demonstrates a slight but significant negative slope. On
days 12 and 20, a dip in serum iron concentration is ob-
served; both are significantly different from the day 4 iron
concentration, but the other days of pregnancy do not differ
significantly from day 4. The pattern of change of serum
iron concentration differs from both the GP+SN and the

liver.

119

Study 3 - to determine (1) if gonadectomy affects brain iron
levels and (2) if 3 weeks estrogen treatment in-
creases or decreases iron levels compared with
control and gonadectomized rats.

Gonadectomy at 4 weeks of age has different effects on
the iron concentration of the GP+SN in males than in females
(Figure 29 and Table 6). Whereas ovariectomy causes no
change in GP+SN iron concentration compared with control
females in estrus, castration causes a significant increase
in GP+SN iron compared with control males. Three weeks of
estrogen treatment of gonadectomized males and females
causes a slight, but not significant, decrease in the GP+SN
iron levels compared with gonadectomized males and females.
The GP+SN iron concentration of both castrated males and
castrated males with estrogen implants are significantly
greater than the control male level. Among the females,
neither ovariectomy nor ovariectomy with estrogen implant
has significantly different GP+SN iron levels than the
control female in estrus.

The iron concentration of the cortex with gonadectomy
and gonadectomy with estrogen treatment differs from that of
the GP+SN, and females respond differently than males
(Figure 30 and Table 6). Castration causes a slight, but
not significant, increase in cortex iron compared with
control males, but ovariectomy causes a significant decrease
in cortex iron compared with females in estrus. Castration

plus estrogen treatment causes a significant drop in cortex

120

.mmNHm mHQEMm pom moanwu

<>oz< mom ..z.m.m H a mmDHm> cam: .AHmm+m x>ov muamHQEw cmwOHumm
nuw3 mmHmEmm wmuflaouomwum>o pom .Ao x>ov mmHmBmw pmNHEODomHHm>o

.Aa mDMHmmV mnhumo a“ moawamm .AHmm+% Hm<uv mucmaaaw amwouumm

LowB moama pmumnummo .Ab Hm<ov mmHmE pmumnummo .Ab Hzouv wmama Houucoo

mo muwflc mwucmumnsm new mspwaamm manon meoom mo aowumhucmocoo couH

.mm mnswam

l2!

Emu
+ ,
0 X>O o X>O 0 mnmhmw

992 $2335
+ m:n___mn_ 250.0

hwm
+
b ...m<o b ...m<o b #200

  

o F
m P
ON
mm
on

mm

w a s FlHDIElM 13M 6
/5" NOHI awaH-NON 1v101

mm 2.6.“.

122
Table 6. Analysis of Variance with Orthogonal Contrasts of
Iron Measures of Control Males (CM), Castrated Males (CAST),
Castrated Males with Estrogen Implants (CAST+EST), Intact
Females in Estrus (ESTF), Ovariectomized Females (OVX), and

Ovariectomized Females with Estrogen Implants (0VX+EST)

 

 

 

Source DF MS F P
GP+SN

Total 35

Between 5 191.56 2.24 NS

Q 1. Males vs Females 1 510.31 5.96 <0.05
Q 2. CM vs CAST, CAST+EST 1 409.32 4.78 <0.05
Q 3. CAST vs CAST+EST 1 5.45 0.06 NS
Q 4. ESTF vs ovx, 0VX+EST 1 8.43 0.10 NS

Q 5. OVX vs 0VX+EST 1 24.31 0.28 NS
Within 30 85.57

CORTEX

Total 35

Between 5 52.97 2.79 <0.05
Q 1. Males vs Females 1 141.69 7.48 <0.025
Q 2. CM vs CAST, CAST+EST 1 0.67 0.04 NS

Q 3. CAST vs CAST+EST 1 41.85 2.21 NS

Q 4. ESTF vs OVX, 0VX+EST 1 77.12 4.07 <0.10
Q 5. OVX vs 0VX+EST 1 3.55 0.19 NS
Within 30 18.94

LIVER

Total 47

Between 5 21,670.92 71.85 <0.001
Q l. Males vs Females l 55,027.24 182.45 <0.001
Q 2. CM vs CAST, CAST+EST l 23,787.48 78.87 <0.001
Q 3. CAST vs CAST+EST l 15,345.64 50.88 <0.001
Q 4. ESTF vs OVX, 0VX+EST 1 9,121.26 30.24 <0.001
Q 5. OVX vs 0VX+EST 1 5,073.00 16.82 <0.001
Within 42 301.61

123

Table 6 continued.

 

 

 

Source DF MS F P
SERUM

Total 53

Between 5 6.61 16.95 <0.001
Q 1. Males vs Females l 1.65 4.23 <0.05
Q 2. CM vs CAST, CAST+EST 1 0.73 1.87 NS

Q 3. CAST vs CAST+EST l 15.40 39.49 <0.001
Q 4. ESTF vs OVX, 0VX+EST 1 0.00 0.00 NS

Q 5. OVX vs 0VX+EST l 15.25 39.10 <0.001
Within 48 0.39

BODY WEIGHT

Total 47

Between 5 38,914.27 78.38 <0.001
Q 1. Males vs Females 1 110,688.02 222.94 <0.001
Q 2. CM vs CAST, CAST+EST l 39,963.02 80.49 <0.001
Q 3. CAST vs CAST+EST 1 12,376.56 24.93 <0.001
Q 4. ESTF vs OVX, 0VX+EST 1 31,518.75 63.48 <0.001
Q 5. OVX vs 0VX+EST l 25.00 0.05 NS
Within 42 496.50

124

.mmmwm mHmEmm How moanmu <>oz< mom

..z.w.m H H mosfim> com: .Aemm+ m x>ov mucmamafi ammonumo nuflB moamamm
meHEOuomHHm>o paw .Am x>ov mmHmEmm pmNHEouomflHm>o.Ao mDMHmmv mdhumw CH
moamfimm .AHmm+ ©Hm<ov mucmamaw amwouumo suwB moama wmumnumwo .Ab Hm<ov

mmHmE UmumuumMU .A.%Hzoov mmHmE HOHucoo mo xmuuoo mo GOMHMHucmoGoo SOHH

.om musmfim

Hmm wa .
+ . +

O X>O O X>O o msmhmw b #95 b .543 bFZOU

 

 

[25

OF

 

 

mp

                         

   

O

)
O

'
o o
’0

o

 

v V V
3030?.
020202

     

I},
N
was LHHoIaM 13M 6
/6r' NOHI awaH-NON 1v101

xmtoo
on 252..

 

126
iron compared with both control males and castrated males
but, in females, ovariectomy plus estrogen treatment causes
a slight, but not significant, rise in cortex iron compared
with ovariectomy. The iron level in the cortex of the
ovariectomy plus estrogen treatment group remains less than
that of females in estrus.

The liver iron concentrations of the male and female
treatment groups differ (Figure 31 and Table 6). Whereas
castration causes a significant increase in liver iron
concentration compared with control males, ovariectomy
causes a decrease compared with females in estrus. Both
gonadectomized groups respond the same way to estrogen
treatment, however, with a significant rise in liver iron
values. Even with gonadectomy occurring as young as 4 weeks
of age, the liver iron concentration of the ovariectomized
females still exceeds that of the castrated males.

The serum iron concentration of the male and female
treatment groups differs significantly (Figure 32 and Table
6). Castration has no effect on serum iron level compared
with the control males; however, ovariectomy causes a drop
in iron concentration compared with females in estrus.
Estrogen treatment of 3 weeks significantly increases serum
iron concentration in both castration plus estrogen and
ovariectomy plus estrogen groups compared with castrated
males and ovariectomized females. This response is similar
to the response of liver iron concentration to estrogen

treatment .

127

.moNHm maaamm How moanmu <>oz< mom

..z.m.m Ha mosam> com: .Ahmw+m x>ov mucmamafl cowouumm sows moamamm
anHEouomHHm>o paw .Am x>ov mwamamw vmnwaouomflum>o .Am mbmemmv msnumw CH
mwamamm .Ahmm+% Hm<ov mucmamaw Gmwouumm LDHB moama pmumuummo .Am Hm<ov

.mmHmE vmumhummo .Aw Hzoov moama Houucoo mo Hm>wa mo cowDMHuamocoo GOHH

.Hm Quasar

l28

Hmm
+

O X>O 0 X>O 0 m3

0. ‘ .. ‘ ‘ .a
ti? .‘f0
.00. 000
000 .000
.000 000
000 .000
.000 000
000 .000
.00. 000
000 .000
.00. 000
000 .000
.00. 000
$3? .fiqfi
000 .000
.00. 000
000 .00.
.00. 000
000 .00.
.00. 000
000 .000
.00. 000
000 0000
.00. 000
000 .00.
.00. 000
o 0 0 c 0 0 .
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. .000
000 000.
.00. .000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00. 000
000 000.
.00.... 90.0 o.
. 0 0 . .0» r..
000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

000

.00.

0.0

.00.

.00. .0.

0.0

.m>_._

mhmm

+
b .540 b .540 b #200

 

000

O.
”

O
:°.°°
’3.

e’

O
’0

0’...
0’03.
0 o

O
0
0

O.
o.
’0

O
02

0
A

0
0

O
.0

0'...
o 0
0° 0’.
o. ‘0‘.

O
’0’...
.0...0 .9

O

0000
0

00
00
0"

o
o:

00
’0
0

O
"
a

O
0’.
o o
O ..

O
o
’0
o

c’
0’ .
o

000

.00

0 1’
J’

0
0
v'

0
0

00
’1

0
0
%:

v

0

 

 

on

om

o:

omp

om:

CD

09

EN

was L-HHDEIM 13M 6
/5fi NOHI awaH-NON 1v101

.n .52...

129

.mmNHm onBmm pom mmHnmu <>oz< mom

.z.m.m H H mmDHm> cam: .AHmm+@ x>ov muamHaaH nowonumm auH3 mmHmamm
meHEouomHnm>o paw .Am N>ov mmHmEmm pmNHEouomHHm>o .Am mDMHmmV msuumm GH
monamm .AHmm+% Hm<ov mucmHmSH ammonumm fiuHB mmHma wmumuumwo .Ab Hm<00

mmHmE pmumhummo .A% H2000 mmHmE Honufioo mo asnwm mo GOHumHucmocoo aouH

.Nm muswfim

I30

#ww
+

o X>0 o x>0 0 m:m#mm

#mm
+
b #90 b#m<0 b #200

 

was l-T-IUJ/NOHI 6"

«m 33E

131

The body weights of the male and female treatment
groups differ significantly (Figure 33 and Table 6).
Castration causes a drop in body weight compared with
control males, but ovariectomy causes an increase in body
weight compared with females in estrus. Estrogen treatment
of 3 weeks causes a further drop in both the castrated plus
estrogen and ovariectomy plus estrogen groups compared with
castrated males and ovariectomized females. Estrogen
treatment has the opposite effect on body weight compared to

that on liver and serum iron concentration.

132

.mwNHm mHmEmm pom mmHnmu <>oz<

mom ..Z.m.m H « moDHm> Gmoz .AHmm+m x>ov mucmHaBH Gmwouumm nuH3
memEmm womwaouoowhm>o paw .Am x>ov mmeEmm pmNHaouomHHm>o .Am mDMHmmv
msnumm SH monEmm .AHmm+© Hm<ov wucmHmfiH cowouumm SuHS memB Umumnummo

.Am Em<ov mmHma pmumuummo .A0 #2000 mmHmB Honuaoo mo ustmB knom

.mm muswnm

I33

#mm #mw
+ +
O X>0 o x>0 o m3m#mm b #m<0 h #m<0 b #200

cop

com

com

  

Ems; 28m

  

'W'EI'S l-T-SWVHE)

mm 95?...

DISCUSSION

Details of the localization of iron in the rat brain
made possible by the DAB intensification of the Perl's
method, the fluctuation of brain iron during the estrous
cycle, the changes during pregnancy, the effects of gonad-
ectomy, and the effects of estrogen treatment, discussed
below, are reported here for the first time. I propose that
this new information, when considered together, suggests:
(l) a role for the presence of iron in the brain, and (2) an
interpretation of its pattern of distribution.

The pattern of distribution of iron in the rat brain
reported here is similar to the distribution which occurs in
the human brain (see literature review). The presence and
extent of accumulation of iron appears to be a character-
istic feature of each brain area. Although iron is required
for the metabolism of every cell, the distribution and
concentration of iron in discrete areas and structures of
the brain suggest that its accumulation occurs either as a
requirement or as a consequence of some specific metabolic
process or processes. Since iron is frequently associated
with large molecules and since most of the brain is pro-
tected by the blood-brain barrier, special mechanisms
transport iron and associated large molecules (e.g., trans-
ferrin) across the basement membranes of capillary walls and
into cells. Such an active uptake further suggests that

iron is an important requirement in some neural metabolic

134

135

process. The uptake of iron in brain areas not protected by
the blood-brain barrier, however, could be considered a
more passive accumulation.
Myelin

The localization of iron in fibers and oligodencrocytes
in many of the high iron areas of the brain suggests that
the iron might be associated with a particular myelin prod-
uced by iron-containing glial cells. In the brain subcel-
lular fractionation studies of Rajan et a1. (1976), a
large portion of brain iron occurs in the myelin fraction.
Also, the persistent deficiency in brain iron following
an iron deficient diet in the studies of Dallman et a1,(l975)
occurs when iron deficiency overlaps with the periOd of
myelination of the central nervous system. Also, in the
male rat, there is a suggestion of an association of myelin
and iron since both myelin (see Norton, 1976) and brain iron
Dallman and Spirito, 1977) are found to have very low turn-
over rates. The association of iron with myelin, however,
does not explain the presence of iron in the brain where it
is not within glial cells but in granules or on or in the
perikarya of nerve cells.
Monoamines

Monoamines, such as dOpamine, norepinephrine, and
serotOnin, require iron for many aspects of their normal
metabolism (see literature review). Although the monoamines
occur in some high-iron areas (Table 7), the distribution

and relatvie abundance of a monoamine and iron in the rat

136

<m<0
<m<0

<m<0

<m<0
<m<0
<m<0
<m¢0

cHaououmm
aHGououmm

chououmm

chououmm
cwcououom

cHnououmm
cHaouonm

cHaououmm

maHusawchoHoz
mcHusmmaHthoz
wcHhsmmchmHoz
mawufimmaHmmuoz

ocwusmmGHmouoz

mcHufimmcHamuoz

kmhw Hmuucmo swannpwz

coHumahom HmHsoHumm

«Haconmn HmnmuwH

mcHEmaom abumwuum
maHBmenwmsv whomnoo

mSEmHmSu HwHucm>

nfinn spouomMHo

 

mumnHm pfim mHHw SH ufimwmum u couH 30H

HmHoDc umHHmanmo moon
mSmHch HmHDoGDUmQHmucH
AnumHSUHumHv mumwn mHucmuman
aanHmo mo mwamHmH

ESpHHHmm Hmuucm>

mnefiflflma manofiw

 

mnmnHm paw «Haw CH ucommnm u COHH ame

mcHEmmoo mumson¢
mocmcwam cmemz
mmuhothB

mnEmHMSuom%£ HmHUmEuouucm>

HH>O

uomnu oHumo mo mwum HmumumH
mmmum umHDoHHucm>Humm

awwuo HmoHaHomnnm

maona pHouono

mfimnumom mmu<

 

mommmUOHm oxHHuHmnHm paw «mcowumHDBSUUm moosmhoam «mmHDamHm SH ucmmmua n couH nwflm

.aflmum umm

mfiu mo mmmh< waHumHuamocoo COHH CH k<m<0 paw mocHEmoaoz mo COHanHHumHQ m£H .# mHan

137

.meH .umpsmmm ram mmHuumommm .ucwumm mwmmH .suom

paw BOOHm .Hmmooo unmmH .muooz paw mmcom “HnoH .upmumumwab "Bonm mH paw GoHumuucmocoo
stfi ou mumuopoa wGHGHmucoo mmmum mmUSHocH <m<0 cam moaHBmocoE mnu mo COHanHHumHn«

chououmm
chouonm
chOuoumm

chououmm

chououmm

mcHufimwcHQmuoz
mcHHLQmCHQmHoz
mcHusmmchmHoz
mGHunmmchmHoz

oannamcHawHoz

mcHnnamaHmmHoz
mGHufimmcHQmHoz

mcHBmmon MHmpw%Em Hmuufioo
maHmeon xmuuoo mudewcwo
Emumhm Hmnwm hmmoe cam mnuhw mumuamn

mamamoommHa HoHHouG<

Ammmhm HmHSUHMqu>HHomv oHumomhm Hmwpmz

xmuuoo

mHHmcHaumu MHHum mo mamHoDc pom

mcHEmmon Enummm HwnmumH
mSoHoDG UHumBmMH£UMHmsm

mSoHoDG HMHnoHHucm>mumm

msmHoac oHumomuadm

 

moonamc mo mmnmxHHmm CH no Go ucmmmum u GouH 30H

.vmnGHucoo # mHan

138

brain are not highly correlated, and monoamines are con-
centrated in many sites where iron is not.
U-aminobutyric acid

i-aminobutyric acid (GABA) is concentrated in those
areas of the brain in which iron is abundant and localized
in glia and fibers (Tab1e7). Further, a glutamic acid
decarboxylase (GAD), although different from the neuronal
CAD, is present in glial cells in the brain (see Cooper,
Bloom and Roth, 1978). However, the corpora quadrigemina,
containing only moderate amounts of iron, are among the
regions of the rat brain containing the highest concentra-
tion of GABA (see Cooper, Bloom and Roth, 1978). Although
the presence of iron in glia and fibers, as occurs in the
globus pallidus and substantia nigra (reticulata), may be
related to the metabolism of GABA, this relationship does
not exPlain the presence of iron in the other high-iron
areas of the brain.
Peptides

There is a correlation between the distribution of iron
in the brain and the known neuroactive peptides. This
correspondence is especially evident in the high4ironwareas
of the brain (Table 8), and concentrations of known peptides
occur only where iron is present. Except for the deep
cerebellar nuclei and lateral islands of Calleja, all high-
iron regions of the brain are associated with at least one
of the known peptides (Table 8). -Enkepha1in is found‘in the

highest concentration in the globus pallidus (Hong et a1.,

139

cHHmsmmxam
cHHmnaoxcm m mocmumnnm aHmcmuOHDoz aHumumoumBom

GHHmzmmxcm zouw HmHquo aHmanHE

aHstmmxcm noHumBHom HwHSUHumm
m mocmumnsm mHSGmnmn Hmnmumq
chnmuowwa< aDuHmHum
GHHmsmmxcm mcHaowHHpmsv muoauoo

qfiH .>< ..c pmHDOHommmmumav GHHmsmmxam

A.fi umHnoHommmmummv

aHmmmHQOmm> mnamHNSu Hmuuco>

caamnamxcm mmmn gasp spouommao

 

muonwm paw mHHmo Hmwdw CH ucmmmumlr coHH 30H

.Hmth HmanmHm CH CHHmsaoxcm HmHosc HMHHmanmo moon

GHHmnamxam m monmumnsm msoHonfi HmHSUcnnmmnmucH

AmHHmHmumH mumav GHHmsmmxcm m mocmumnnm meHG mHucmquSm
mama mmmHHmo mo muamHmH

aHHm3mmxam . BfipHHHmm Hmnuao>

m mocmumnsm caflmnamxcm maufiaama manofio

 

mHoAHm paw mHHmo HmHHw aH ucmmoum n couH anm
SHHmsmoxcm GHuMumoumaom mmmH mumaou<
monoahom waHmmmHmm cHHmnmmxcm aHooumxo chmmHQme> chcmuOHwa< moamcHam smwpmz

mmmH mmu%o%cma

aHanopSMnm chcouOHsz cHHmsmoxcm

m moamumnam cHumumoumaom

cHHmfiamxcm cHumumoumSom mmmH mMH
GHHmfimmxcm m monmumnsm :HumumoumEom mmH
anflmnamxcm

cwoouhxo

cwmanOch< mSEmHmnuom%£ Hmwpoauouuzm>

aHmcouoawc< HH>0

uomhu owumo mo mwpo HmuoumH
aHmcmuowwG< wwwnm HmHSUHHuam>HHmm
GHmawuonn< cmwno HmoHchomnnm
GHmmoHQOmm> mnmea wHOHoao
:HmcmuOHwn< mamuumom mmh<

 

mGOHumHSBDoom muonauoam paw mommmoOHm mxHHuumnHm paw mmHscmmw aH ucmmoum I GOHH mwwm

.chHm umm ofiu mo mmmu< wcHumHunmocou GOHH paw xmmpwummm mo aowusnwhumwn m:# .m mHnma

.HmoH ..Hm um mans “ommfl ..Hm um maaaflflnm
uome ..Hm um NDHBOUmoZ mommH .HmpHHfio cam Hmpham mommH ..Hm um cmahm>HHm uwan ..Hm um
cmucwo MmmmH .uHmmxmm paw mva “wan ..Hm um Bouowoo mwan .auom paw Boon .Hmaooo ”Bonn
mH pom aoHumHucmocoo st5 on mumumwoa wGHchuGoo mmmum moanoGH mopHummm mo coqunHHuman

fifiHmsmmxam nHmcmuowa<

140

Axmuuoo Hmucoumv GH

cHHmsmmxam
aafimsamxam

GHHmsmmxcm
unflmsamxcm mama .mma

Hmsmmxcm m mofimumnam

GHHmsmmxnm m woamumnsm cHumumoumEom

chcmuonsmz GHHmsmmxcm m moGMumnsm mMH
cHumumoumaom mmH
aHHmsamxcm mme

CHHmsmmxcm

cHUOumxo chmmHQOmm>
:Hoouhxo chmoHQOmw>
aHoou%xo chmmHQOmm>
cHoouhxo chmoHQOmm>

mHmpwhfim HmHquo
xmuuoo muwHchHo
amumhm HmnHm
xmmoa ram monkw mumucmo
mamamooamws HoHHmun¢
AmconmH HMHDUHHuCo>V
owumowhm HwHUmZ
xmuuoo
mHchHaHmu
mHHum mo manonc com
Enuamm HmumumH
mDmHoDc UHumEmMHnomumsm
msoHoaa umHnoHHucm>mem
mSmHonc owumomHQSm

 

msousoa mo mmwmxHuwm CH 90 co ucommum.u GoHH 30H

.vmficHucoo m mHQMH

141
1977), the brain region with the highest concentrations of
iron as measured by Hallgren and Sourander (1958). Enkeph-
alin and opiate binding sites have similar distribution
patterns in the rat but have different relative concentra—
tions (Atweh, 1977a,b,c); both occur in areas in which iron
has been localized in the present study. Substance P is
abundant in the iron rich substantia nigra (reticulata)
and interpeduncular nucleus (see Cooper, Bloom and Roth,
1978; Elde and Hkaelt, 1978 for review). The hypophyseal
hormone releasing factors and hypothalamic neurohypophyseal
hormones are present in abundance in the circumventricular.
and ventro-medial hypothalamic areas of the brain. These
are regions in which the blood-brain barrier is absent and
much iron accumulates.

The association of iron with peptides is further
evidenced by the eXistence of several known sites in which
a peptide and iron are not only localized in the same
general region but appear to be within the same cytoarchi-
tecturally distinct structures. The presence of enkephalin
within fine fibers of the globus pallidus has been deter—
mines with immunocytochemical methods (Sar et al.,1978;
Jacobowitz et a1.,1979. Although, as mentioned in the
Results section, it cannot be determined from the present
material whether the iron is within the axoplasm or myelin
of iron-filled fibers, iron is present in fine fibers in
the globus pallidus (Figure 20). Also, Jacobowitz et a1.,

describe large enkephalinergic axons seen ventral to the

142
anterior commissure projecting rostrally towards the olfac-
tory tubercle. These large enkephalinergic fibers are
similar in both appearance and distribution to the iron-
filled "cascades" of large fibers found in the ventral
pallidum and around the islands of Calleja of the olfactory
tubercle (Figure 21).

The localiZation of enkephalin within the granule cells
of the dentate gyrus and continuing to the mossy fiber
terminations adjacent to the pyramidal cells of the hippo-
campal region CA3 and CA4 (Gall et a1., 1981), corresponds
exactly to the localization of iron within this region in
the rat (Figure 6), and applies equally well to the local-
ization of zinc (McLardy, 1962).

The substance identified by neurosecretory stains,
and more recently with immunocytochemistry for neurOphysin
(Elde and Hkaelt, 1978), has been localized in cells of
the supraoptic nucleus, paraventricular nucleus, and to
a lesser extent, the suprachiasmatic nucleus. This subs-
tance is also seen as beaded fibers in the hypothalamic-
hypophyseal tract and in clumps of granules in the ventro-
medial hypothalamus, arcuate area, median eminence, and in
periventricular areas. Iron is found within the supra-
Optic, paraventricular, and suprachiasmatic nuclei and
in fibers, clumps of granules in the ventro-medial hypo-
thalamus, arcuate area, and the median eminence. Further-
more, I have observed that the iron-containing granules

are similar in size and distribution to the neurosecretory

143
granules as seen with the classic neurosecretory stains
(unpublished observations).

Although the distribution of luteinizing hormone
releasing hormone (LHRH) overlaps with the presence of
many other peptides in the iron-rich areas of the organum
vasculosum of the lamina terminalis (OVLT), ventro-medial
hypothalamic, and periventricular areas, LHRH is the only
peptide localized within the tanycytes of the third
ventricle in the rodent brain (Zimmerman et a1., 1974;
King et a1., 1981). This study shows that the tanycytes
of the third ventricle of the rat are iron-filled (Figures
12,13 and 14).

The association of iron with the peptides of the
nervous system suggests that iron might function in some
capacity in the metabolism, transport, and/or storage
of peptides, and the further association of iron with LHRH
suggests a role for iron in neuroendocrine regulation.

Iron occurs in a variety of structures such as in
glial cells and in or on neurons, as well as within fibers,
fiber-like processes, and granules. The presence of iron
in these different structures could be related to its
association with different neurologically active molecules
or in different aspects of the metabolism of a single
substance. Iron within the cytOplasm of neurons may be
associated with the synthesis of a peptide or other sub—
stance. However, since peptides are believed to be formed

from larger molecules during passage along the axon

144

(Marks, 1978), it is interesting that iron is frequently
in high concentration in fibers but only seldom, and
at low concentration in perikarya. This suggests that
the presence of iron in fibers could be related to either
a carrier molecule or to a peptidase., The iron in bouton-
1ike structures could be related to peptides in terminals,
either involved with the storage of the peptide or with a
peptidase. Sullivan et a1, (1979) have recently described
an enkephalinase which is specific for enkephalin and has
a regional distribution similar to the opiate receptor
binding sites and therefore to iron. Enkephalinase is a
metalloenzyme which is inhibited by the iron chelators
1,10 phenanthroline and ethylenediaminetetraacetic acid
(EDTA). It appears likely that the metal in the enkephalin-
ase is iron and therefore that the distribution pattern
of iron is related, at least in part, to the distribution
of enkephalinase.
Iron in Neuroendocrine Function

Estrous cycle and pregnancy

VThe suggestion that iron is involved in neuroendocrine
functions is further supported by the results of this study
which demonstrate that brain iron levels vary in conjunction
with changes in ovarian hormone concentrations. During
proestrus, when pooled globus pallidus plus substantia
nigra (GP+SN) iron concentration is highest, plasma con-

centrations of estrogen and progeSterone are at their peak

145

(Butcher et al.,1974). Although estrogen levels remain
low after day 4 of pregnancy, progesterone increases until
about day 14, after which levels drop (Sato and Henkin,
1973). Iron levels in the GP+SN also increase during the
first half of pregnancy. Although the brain iron level
drops slightly toward the end of pregnancy, at least 12-24
hours post partum the GP+SN concentration is actually raised,
representing, perhaps, the proestrus rise in iron preceeding
a post partum estrus. Cortex iron concentrations do not
change significantly either during the estrous cycle or
during pregnancy; however, the pattern of change of the
cortex during pregnancy is similar to that of the GP+SN.
Iron is localized in glial cells and fibers in the GP+SN,
whereas in the cortex iron is localized in the bouton-like
structures on or in neurons. Perhaps the presence of iron
in different structures reflects a different metabolic
role and thus a different response to hormones.

Sex difference

There is a sex difference in the iron concentration
of the GP+SN of male and proestrus female rats. During the
other stages of the estrous cycle, the female GP+SN iron
levels are within the male range. The iron concentration
of the male GP+SN and the level to which the female GP+SN
returns after a proestrus rise could represent a "baseline"
iron concentration which is required for the optimal func-
tioning of these areas in a 12 week old rat. The baseline

level for older animals of either sex is not known. In the

146

cortex of 12 week old rats, iron concentration is the same
for both sexes and during all stages of estrus. However,
the histochemical demonstration of iron in the brain tissue
of 32 week old rats demonstrates that the female has more
cOrtex iron than the male of the same age (Figures 2,3,4,
5,7 and 8). Thus, with increasing age, male and female
brain iron concentrations may differ even during non-pro-
estrus stages.

Since in the rat the percentage of total brain iron
in the storage form, ferritin, increases with age (Dallman
et a1., 1975), the increase in histochemically identifiable
iron with age seen in this study may reflect an increase
in storage iron only and not in non-ferritin forms of
iron.

Although the accumulation of iron in the brain appears
to be influenced by ovarian hormones, the brain does not
respond the same way as the liver. A rise and fall in
liver iron concentration occurs during the estrous cycle,
but liver iron rises during diestrus, before GP+SN levels.
In this study, liver iron values did not increase through
the first third of pregnancy, as do the GP+SN iron levels,
but dropped continuously from 4 days pregnant through
to 12-24 hours post partum. Also, the sex difference in
liver iron concentration is much more pronounced than the
sex difference in brain iron. Whereas the GP+SN iron dif-
feredonly between proestrus females and males of the same

age, the liver iron of the male is only about 1/3 that of

147

of the concentration of the female at any estrous stage.
This difference in response between the brain and the liver
could be due to a difference in the forms of iron present
in these tissues fCabout 60-70% of the liver iron is in
association with ferritin, whereas only about 15-25%
of brain iron can be shown to be in this form; see liter-
ature review), or the iron-accumulating tissues are respond-
ing to different hormones. The estrogen rise during the
estrous cycle occurs about 12 hours before progesterone
(Butcher, et a1.,1974); perhaps the liver is responding
to estrogen and the brain to progesterone or to some temp-
oral or concentration relationship between the two ovarian
hormones. During pregnancy, the GP+SN iron levels, as
shown here, more closely match the pattern of the plasma
progesterone concentration (Sato and Henkin, 1973) than
do the liver iron values (Figure 28).

Ovariectomy, castration and estrogen implantation

Ovariectomy does not change the iron concentration
of the GP+SN compared with that of non-proestrus females.
The iron concentration of this part of the brain remains
at the "baseline" level, the maintenance of which, evidently,
is not dependant upon the presence of ovarian hormones.
In the spectrophotometric study, estrogen implants also
have no significant effect on the iron concentrations of
the GP+SN. In the histochemical study, when ovariectomy
was performed at a young age, estrogen treatment had no

visible effect on the brain iron levels. Only when

148
ovariectomy ocCured in an adult, followed shortly by ,
estrogen treatment, did the hormone appear to increase
GP+SN iron. Either estrogen alone does not increase brain
iron, or ovariectomy after a long period of time affects
the iron-accumulating abilities of the GP+SN. Cortex iron,
however, decreases after ovariectomy-and does rise slightly,
but not significantly, with estrogen treatment which Sug-
gests that the maintenance of cortex iron concentration
may be influenced by ovarian hormones in a manner differ-
ent from that which influences the GP+SN.

Testicular hormones apparently suppress the iron-
accumulating abilities of the GP+SN since an increase in
brain iron is seen in those areas with castration. Iron
levels in the cortex are apparently unaffected by castra-
tion. The "baseline" GP+SN iron level in the normal male
is apparently meaintained by suppressing iron accumulation.
Estrogen treatment had no significant effect on the iron
concentration of the GP+SN in the castrated male.

Both ovariectomy and castration have profound effects
on the iron accumulation of the liver, which suggests that
the ovarian hormones are necessary in order to maintain
the high liver iron concentrations in the female and that
testicular hormones suppress liver iron accumulation in the
male. Estrogen implants caused significant increases in
liver iron in both sexes. Thus, both male and female liver
tissues were able to respond to estrogen administration, and

the hormone had similar effects in both sexes.

149

Estrogen stimulated ferritin synthesis (Bj¢rklid and
Helgeland, 1970) which results in the increase in iron
concentration seen in the liver of estrogen treated males
and females in this study. However, if the brain is also
accumulating iron in the form of ferritin, one would expect
a similar response. Estrogen treatment also stimulates the
synthesis of ceruloplasmin (Planas, 1973), a ferroxidase.
Increase cerulOplasmin may act to move iron from tissues
into the blood. This might explain the slight, but not
significant, decreases in brain iron concentration seen
with estrogen treatment. The rise in GP+SN iron concentra-
tion during proestrus may be due to progesterone alone or
to some temporal or concentration relationship between est-
rogen and progesterone as noted above.'

Even though gonadectomy was performed at four weeks of
age in this study, well before sexual maturity in the rat,
the male and female brain tissues responded differently to
gonadectomy. Thus, by 4 weeks of age, a sex difference in
the brain iron accumulating abilities had already occurred.
Perhaps this sex difference in the brain is another central
nervous system function subject to the organizational
effects of early gonadal hormone exposure (MacLusky and
Naftolin, 1981).

Globus Pallidus and NeuroendocrinezFunctiOn,

The globus pallidus and substantia nigra are generally

thought to be part of the motor systems of the brain.

Neither of these regions are loci for estrogen receptors

150

nor are they generally included in the group of brain
regions thought to be involved in reproductive functions.
Only two studies were found that link the globus pallidus
to sex differences or sexual behavior. There is a sex
difference in the effects of globus pallidus lesions on
weight gain in the rat (Hahn and Lenard, 1977), and vagino-
cervical stimulation incudes an increase in 2-deoxy-D-g1u-
cose utilization in the globus pallidus (Allen et a1., 1981).
Vaginocervical stimulation affects a number of physiological
processes involved with reproductive events including pro-
gesterone secretion, sperm transport, sexual receptivity,
locomotion, and perception of pain in female rats (see
Allen et at., 1981). Brain areas showing increased meta-
bolism following vaginocervical stimulation that have been
implicated in the control of reproduction include: globus
pallidus, medial preOptic areas, mesencephalic reticular
formation, bed nucleus of the stria terminalis, and dorsal
raphe.
Iron in the Medial Preoptic Area

Electrochemical stimulation with stainless steel elec-
trodes or the application of iron salts to the medial
preoptic area of the brain induces ovulation (Everett and
Radford, 1961; Dyer and Burnet, 1976; van der Schoot et al.,
1978; Columbo and Saporta, 1980). These procedures cause
irritation and/or lesioning in the site of the iron deposi-
tion. Recently, Columbo and Saporta (1980) have demonstr-

ated that the iron deposition causes increased 2-deoxy-D-

151
glucose uptake and suggest that the increased metabolic
rate produced by iron is the ovulation-inducing stimulus.
Interpretation of Iron Fluctuation in the Brain

The fact that endogenous iron in the medial preoptic
area can induce ovulation suggests that the naturally occur-
ring proestrus peak in the iron concentration of the globus
pallidus and ventral pallidum, areas immediately adjacent
to the preoptic region might somehow be involved in the
normal ovulatory stimulus.

The distribution of iron described here suggests an
association of iron with peptide metabolism and, in some
areas, specifically with enkephalinase.‘ Because endogenous
opioid peptides have been implicated in the regulation of
the proestrus luteinizing surge of the rat (Pang et a1.,
1977; Muraki et a1., 1979) and met-enkephalin in the
hypothalamus fluctuates during the rat estrous cycle (Kuman
et al.,1979), the activity of the associated enkephalinases
might also be expected to fluctuate with changing events
in the estrous cycle. Perhaps the iron accumulation in the
globus pallidus during proestrus is related to a change in
peptide metabolism which, in turn, influences reproductive
functions.

The fluctuations of iron seen in the high-iron areas
of the GP+SN may also reflect a similar change occurring
in the high-iron ventro-medial hypothalamic areas, thus
influencing LHRH directly through the association of iron

with LHRH in the organum vasculosum of the lamina terminalis

152

tanycytes, and arcuate and median eminence areas.

The monoamines, involved in gonadotropin release
(McCann and Ross, 1975) and sexual behavior (Meyerson
et a1., 1973) are regulated by iron-containing enzymes
(see literature review). This represents another route
through which changes in iron availability can affect
reproductive functions.

The results of this study suggest, firstly, that
iron in involved in neuroendocrine regulation and reprod-
uctive functions and secondly, that the distribution
pattern of iron in the brain may be related to the associa-
tion of iron with neuroactive peptides. These results
empnasize the importance of iron for the normal metabolism
of the brain and behavior. Iron deficiency, a common
world-wide nutritional disorder, has effects on reproduct-
ive physiology. Perhaps the amenorrhea and menstrual
irregularity associated with iron deficiency anemia in
women is due to insufficient iron present to permit adequ—
ate functioning of the gonadotrOpin release mechanism rat-

her than being due to anemia per se.

153

Why the Brain has so much Iron.

Prefaced with a reminder that very little is
known about the metabolism of iron in the brain, the
following is an outline of how the available facts can
be interpreted in order to explain how it is that the
brain accumulates so.much iron.

Brain iron can be subdivided into three compartments:
myelin iron, non-myelin non-ferritin iron, and ferritin
iron.

The first compartment, myelin iron, increases in
total amount during the myelination period of brain growth.
However, the amount of iron in this form will level off-
as myelination slows down and will remain at a steady
level after about 6 month of age in the rat.

The second compartment, non-myelin non-ferritin iron,
includes all the iron enzymes and those forms of iron
involved in the dynamics of neuronal and glial metabolism.
This compartment could remain at a relatively constant
amount after brain development ceases, however, fluctuations
in amounts of iron could take place as increased accumu-
lation or turnover of specific iron molecules occurs in
response to specific metabolic needs.

The third compartment, ferritin iron, is likely the
compartment of brain iron which increases with the increas-
ing age of the animal. In the liver, ferritin iron in-
creases with the age of the animal (see literature review).

Also, Dallman et al.(1974), report that the brain tissue

154
from older rats has a higher percentage of ferritin than
that found in the tissue of young rats. The accumulation
of iron in ferritin is not necessarily a direct function
of the immediate needs of the brain for it is not likely
that the older rat brain requires more iron in order to
function. Perhaps those parts of the brain which have the
facility to store iron, and other metals (e.g. globus
pallidus and substantia nigra) accumulate metals without
a mechanism for regulating the amounts of metal taken up.
Metals remain stored because there is no ready mechanism
for their removal.

Thus, the increase in iron in the brain with age can
be explained during development by the increases in myelin
and nonémyelin non-ferritin iron. After the brain ceases
to grow the ferritin iron compartment continues to increase

as is seen in the liver.

155

Future Research

Two main hypotheses developed from this study: (1)
the distribution of iron is related to the metabolism of
peptides, and (2) iron is involved in neuroendocrine
regulation.

Further research with a view to demonstrating the
validity of these hypotheses is outlined below.

A relationship of iron and peptides can be better
understood by answering the following questions:

- Do both iron and a peptide occur in the same
structure? Do, for example, both iron and enkephalin
occur in the same fibers within the globus pallidus and
ventral pallidum? Do both iron and neurophysin occur in
the same granules in the ventro-medial hypothalamus?

- During the estrous cycle do the levels of enkephalin
change, or does the activity of enkephalinase change with
the fluctuations of iron concentration? Does an iron
chelator (e.g. Desferal) in the globus pallidus effect
the amount of enkephalin or the activity of enkephalinase?
Does the inhibition of enkephalinase effect the concentra-
tion of iron?

A relationship of iron to neuroendocrine regulation
would be better understood by answering the following
questions:

- Does iron concentration fluctuate during the
estrous cycle in the medial preoptic area, organum

vasculosum of the lamina terminalis, ventro-medial

156

hypothalamus and median eminence?

- Does an iron chelator in the third ventricle, medial
preoptic area or globus pallidus prevent or in any measur-
able way effect ovulation?

Also, a role for the proestrus rise in iron concent-
ration in the globus pallidus in the ovulatory process
could be demonstrated if : (l) electrochemical stimulation
of the globus pallidus caused ovulation, or if (2) an
iron chelator in the globus pallidus prevented ovulation.

The techniques required to answer the above questions

are presently available.

157
SUMMARY

The results of the histochemical localization of iron
in the rat brain and the spectrophotometric measurement of
iron concentration in the pooled globus pallidus and sub-
stantia nigra and the cortex as well as in samples of the
serum and liver have demonstrated that:

- Brain iron is unevenly distributed in the rat brain
in a pattern similar to that found in the human
brain.

- Brain iron occurs in different structures in
different parts of the brain. In the high-iron
areas of globus pallidus, ventral pallidum, substantia
nigra (reticulata) and dentate nucleus iron occurs
in glial cells and fibers; in the circumventricular
organs iron occurs in granules, tanycytes, fiber-like
processes and amorphous accumulations. In lower
iron areas such as the lateral septum, bed nucleus
of the stria terminalis and cortex iron occurs in
bouton-like structures on or in the perikarya of
neurons. In the supraoptic, paraventricular and
suprachiasmatic nuclei iron occurs as a fine dusting
of grains within the perikarya of cells.

- The distribution of iron in the rat brain correlates
to some extent with the distribution of neuroactive
peptides; in come cases, iron and a peptide appear

to be localized in the same structure.

158
- Brain iron increases with age in the rat brain.

- Brain iron fluctuates during the estrous cycle in
the iron-rich areas, rising to the highest levels
during proestrus.

- Brain iron concentration increases during the
first third of pregnancy and then falls but is not
depleted by pregnancy.

- There is a sex difference in iron content in the
high-iron areas of the brain between intact males
and females in proestrus.

- Ovariectomy and castration have different effects
on brain iron levels. GP+SN iron concentration
increases with castration but remains the same with
ovariectomy.

- Estrogen treatment of rats, gonadectomized for 5
weeks has little effect on the iron accumulation
in the high-iron regions.

These results suggest that:

- The pattern of iron distribution may be related to
a participation of iron in the metabolism of pep-
tides; specifically, the distribution of iron, in
part, appears to be related to enkephalinase which
some evidence suggests is an iron enzyme.

- The iron-accumulating regions of the brain are
influenced by ovarian hormones; and progesterone
seems to be the more effective hormone.

- The sex difference in iron metabolism is another

159

CNS factor influenced by early hormone exposure.
Iron plays a role in neuroendocrine regulation

and may be involved in the process of ovulation
either at the level of the pre0ptic, at the inter-
action of iron and LHRH, or through its involvement

in monoamine metabolism.

BIBLIOGRAPHY

BIBLIOGRAPHY

Arai, Y. and R.A. Gorski, 1968. Critical exposure time
for androgenization of the rat hypothalamus determin-
ed by antiandrogen injection. Proc. Soc. Exp. Biol.
Med. 121: 590-593.

Ashkenazi, R., D. Ben Shachar and M.B.H. Youdim. 1980.
Iron deficiency and the dopamine receptor in the
caudate in the rat. Neurosci. Abst. 6: 599.

Allen, T.O., N.T. Adler, J.H. Greenberg and M. Reivich.
1981. Vaginocervical stimulation selectively increase-
es metabolic activity in the rat brain. Science
211: 1070-1072.

Atweh, S.F.‘and J.M. Kuhar. 1977. Autoradiographic
localization of opiate receptors in rat brain.
1. Spinal cord and lower medulla. Brain Res. 124:
53-67.

Atweh, S.F. and M.J. Kuhar. 1977. Autoradiographic
localization of opiate receptors in rat brain.
11. The brain stem. Brain Res. 129: 1-12.

Atweh, S.F. and J.M. Kuhar. 1977. Autoradiographic
localization of opiate receptors in rat brain.
111. The telencephalon. Brain Res. 134: 393-405.

Beaton, G.H., 1974. Epidemiology of iron deficiency.
In: Iron in Biochemistry and Medicine. Eds. A. Jacobs
and M. Worwood. Academic Press. London. 477-525.

 

Bjorklid, E. and L. Helgeland. 1970. Sex difference in
the ferritin content of rat liver. Biochim. Biophys.
Acta. 221: 583-592.

Bliss, C.I. 1967. Statistics in Biology. McGraw-Hill
New York. 1: pp 558.

 

Brannan, T.S., H.S. Maker and I.P. Raes. 1981. Regional
distribution of catalase in the adult rat brain.
J. Neurochem. 36: 307—309.

'160

161

Bray, R.C. 1975. Molybdenum iron-sulfer flavin hydroxyla-
ses and related enzymes. In: The Enzymes Ed. P.D.
Boyer. 42: 299-417. Academic press. N.Y.

 

Butcher. R.L.,W.E. Collins and N.W. Fugo. 1974. Plasma
concentration of LH,FSH, prolactin, progesterone
and estradiol 17 beta throughout the 4 day estrous
cycle of the rat. Endo. 94: 1074—1708.

Cammermeyer. J. 1947. Deposition of iron in paraventricu-
lar areas of the human brain in hemachromatosis.
J. Neuropathol. Exptl. Neurol. 4: 111-127.

Cantwell, R.J. 1974. The long-term neurological sequelae
of anemia in infancy. Pediatr. Res. g: 342.

Charlton. R.W., D.M. Hawkins. W.O. Mavor and T.H. Bothwell.
1970. Hepatic storage iron concentration in different
population groups. Int. Clin. Nutr. 24; 358—371.

Charlton. R.W., P. Jacobs, J.D. Torrance and T.H. Bothwell.
1965. The role of the intestinal mucosa in iron
absorption. J. Clin. Invest. 44: 543-554.

Cheney, B.,A. Lothe, E.H. Morgan, S.K. Sood and C.A. Finch.
1967. Internal iron exchange in the rat. Am. J.
Physiol. 212: 376-380.

Columbo J.A. and S. Saporta. 1980. Increased local uptake
of 2-deoxyg1ucose after electrochemical or direct
depostion of iron into the rat brain. Exper. Neurol.
19: 417-437.

Cook, J.D., C. Hershko and C.A. Finch. 1973. Storage
iron kinetics. V. Iron exchange in the rat. Br. J.
Haem. 45: 695-705.

Cooper, J.R., F.E. Bloom and R.H. Roth. 1978. The
Biochemical Basis 24 Neuropharmacology. 3rd. Edition.
Oxford University Press. N.Y. pp 327.

 

 

Courville, C.B.,R.E. Nusbaum and E.M. Butt. 1963.
Changes in trace metals in brain in Huntington's
chorea. Arch. Neurol. (Chic.) 4: 481-489.

Cumings. N.J., 1948. The copper and iron content of
brain and liver in the normal and in hepato-lenticular
degeneration. Brain. 24: 410-415.

Cumings, J.N. 1968. Trace metals in the brain and Wilson's
disease. J. Clin. Path. 24: 1-7.

162

Dallman, P.R. and H.C. Schwartz. 1965. Distribution of
cytochrome C and myoglobin in rats with dietary
iron deficiency. Ped. §§: 677-685.

Dallman, P.R., M.A. Siimes and E.C. Manies. 1975. Brain
iron: persistent deficiency following short-term
iron deprivation in the young rat. Br. J. of Haematol.
34: 209-215.

Dallman, P.R. and R.A. Spirito. 1977. Brain iron in the
rat: Extremely slow turnover in normal rats may

explain long-lasting effects of early iron deficiency.
J. Nutr. 107: 1075-1081.

Davison, A.N. and A. Peters. 1970. Myelination. Spring-
field, Ill. Thomas.

 

Deizel, P.B. 1955. Iron in the brain: A chemical and
histochemical examination. In: Biochemistry 2: Egg
Developing Nervous System. Ed. H. Waelsch. Academic
Press, London. 145-152.

 

 

Dogterom, J.F., G.M. Snijdewint and R.M. Buijs. 1978.
The distribution of vasopressin and oxytocin in the
rat brain. Neurosci. Lettr. 9: 341-346.

Dooling, E.C., W.C. Schoene and E.P. Richardson. 1974.
Hallervorden-Spatz syndrome. Arch. Neurol. 49: 70-83.

Douglas, W.W., 1974. Histamine and antihistamines; 5-
hydroxytriptamine and antagonists. In: The Pharma-
cological Basis 9: Therapeutics. Eds. L.ST—Coodman
and A. Gilman. Ed. 5. MacMillan Publishing Co.
N.Y. 613—622.

 

 

Drysdale, J.W., 1976. Ferritin phenotypes: structure
and metabolism. Ciba Symposium 51 on Iron Metabolism
Elsevier, Amsterdam.

Dubrunquez, R. and J. Lederer. 1963. Le controle
hormonal du metabloisme du fer. XIX. Action recip-
roque de la surrenalectomie et de l'ovariectomy sur
le fer tissulaire de la ratte. Ann. d'Endo. 24: 709.

Dukes, P.P. and E. Goldwasser. 1961. Inhibition of
erythropoiesis by estrogens. Endo. 44: 21-29.

Dyer, R.C. and R. Burnet. 1976. Effects of ferrous iron
on preoptic area neurons and luteinizing hormone
secretion in the rat. J. Endocr. 69: 247-254.

163

Edgerton, V.R.,S.L. Bryant and C.A. Gillespie. 1972.
Iron deficiency anemia and physical performance
and activity of rats. J. Nutr. 102: 381.

Elde, R. and T. kufelt. 1978. Distribution of hypo-
thalamic hormones and other peptides in the brain.
In: Frontiers in Neuroendocrinology. 5. Eds. W.F.

Ganong and L. Martini. Raven Press N.Y. 1-33.

 

 

Everett, J.W. and H.M. Radford. 1961. Irritative depos-
its from stainless steel electrodes in the preoptic
rat brain causing release of pituitary gonadotropin.
Proc. Soc. Exper. Biol. and Med. lgg: 604-609.

Fairbanks, V.F., J.L. Fahey and E. Beutler. 1971 Clin-
ical Disorders 94 Iron Metabolism. 2nd Edition.
Grune and Stratton, N.Y., pp 486.

 

 

Fielding. J. and R. Ryall. 1970. The action of reducing
agents on the dissociation of iron-transferrin with
special reference to thioglycollate. _Clin. Chem.
Acta. gg; 423=429.

Flerko, B. 1966. Control of gonadotropin secretion in
the female. In: Neuroendocrinolggy. Eds. L. Martini
and W.F. Ganong. Academic Press N.Y. 4: 613-688.

 

G. Frederick Smith Chemical Company. 1980. ‘The Iron
Reagents. G. Frederick Chemical Company, Columbus OH.

Freeman, W. 1930. Deficiency of catalytic iron in the
brain in schizophrenia. AMA Arch. Neurol. Psychiat.
24: 300-310.

Frieden. E. 1974. The evolution of metals as essential
elements. In: Protein-Metal Interactions. Ed.
M. Friedman. Plenum Press. N.Y. Adv. Exptl.
Med. Biol. 4g: 1-32.

 

Frieden, E. and S. Osaki. 1974. Ferroxidases and ferri-
reductases: their role in iron metabolism. In:
Protein-Metal Interactions. Ed. M. Friedman.

Plenum Press. N.Y. Adv. Exptl. Med. Biol. 4g: 33-60.

 

Gall, C., N. Brecha, H.J. Karten and K. Chang. 1981.
Localization of enkephalin-like immunoreactivity
to identified axonal and neuronal populations of
the rat hippocampus. J. Comp. Neurol. 444:
335-350.

164

Ganong, W.F. 1975. The role of catecholamines and
acetylcholine in the regulation of endocrine
function. Life Sci. 22: 1401—1414.

Ganten, D., K. Fuxe, M.I. Phillips, J.R.E. Mann and U.
Ganten, 1978. The brain isorenin-angiotensin
system: biochemistry, localization, and possible
role in drinking and blood pressure regulation.

In: Frontiers in Neuroendocrinology Eds. W.F. Ganong

and L. Martini. Raven Press. New York. 2: 61-101.

 

 

Garby, L. 1973. Iron deficiency: definition and preva-
lance. In: Clinics 42 Haematology. Ed. S.T.
Collender. W.B. Saunders. London. 2: 245-257.

 

Gill, J.L. 1978. Design and analysis of Experiments
in the animal and medical sciences. The Iowa State
University Press. Ames, Iowa. 3 Volumes pp 883.

Gitlin, D. and A. Cruchaud. 1962. On the kinetics of iron
absorption in mice. J. Clin. Invest. 44; 344-350.

Glover, J. and A. Jacobs. 1972. Activity pattern of
iron deficient rats. Brit. Med. J. 2: 627-628.

Goldberg, Wm. and N. Allen. 1980. Nonspecific accumulat-
ion of metals in the globus pallidus in Hallervorden-
Spatz disease. Trans. of Amer. Neuro. Assoc. 222:
106-108.

Goodman, L. 1953. Alzheimer's disease, a clinico-pathalo-
gic analysis of twenty-three cases with a theory on
pathogenesis. J. Nerv. Ment. Dis. 117: 97-103.

Guizzetti, P. 1915. Principali resultati dell'
applicazione grossolona a fresco delle reazioni
istochimiche del ferro sul sistema nervoso centrale
del l'womo e di alaini mammiferi domestici. Riv.
Pat. Nerv. Ment. 29: 103-117.

Hahn, Z. and L. Lenard. 1977. Elimination of the sex-
dependence of short term body weight changes after
bilateral pallidal lesion by neonatal castrationp
Acta Physiol. Acad. Sci. Hung. 22: 229-231.

Hallgren, B. and P. Sourander. 1958. The effect of age on
the nonhaem iron in the human brain. J. Neurochem.
3: 41-51.

Hallgren, B. and P. Sourander. 1960. The non-haemin‘iron
in the cerebral cortex in Alzheimer's disease.
J. Neurochem. 4: 307-310.

165

Halliwell, B. 1979. Oxygen-free-radicals in living
systems: Dangerous but useful? In: Strategies 2:
Microbial Life 43 Extreme Environments Ed. M. Shilo.
Dahlem Konferenzen. Berlin. 195-221.

 

 

HANES. 1974. Preliminary findings of the first health and
nutrition examination survey, U.S., 1971-1972; Dietary
intake and biochemical findings, U.S. Dept. of Health,
Education and Welfare Publication NO(HRA) 74-1219-1.

Hanig, R.C. and M.H. Aprison. 1967. Determination of
calcium, copper, iron, magnesium, manganese, potassium
sodium, zinc and chloride concentration in several
brain areas. Anal. Biochem. 21: 169-177.

Harrison, P.M., R.J. Hoare, T.C. Hoy and I.G. Macara. 1974.
Ferritin and haemosiderin: structure and function.

In: Iron 44 Biochemistry and Medicine. Eds. A. Jacobs
amd M. Worwoodi ’Academic Press. EOndon. 73-109.

 

Harrison, W.W., M.G. Netsky and M.D. Brown. 1968. Trace
elements in human brain: Copper, zinc, iron and
magnesium. Clin. Chem. Acta. 22: 55-61.

Hershko, C. and L. Eilon. 1974. The-effect of sex dif-
ference on iron exchange in the rat. Br. J. Haemat.
22: 471-481.

Hershko, C., H. Cohen and G. Zajicek. 1976. Iron mobil-
ization in the pregnant rat. Brit. J. Haematol. 22:
505-516.

Hong, J. S., H.Y.T. Yang, W. Fratta and E. Costa. 1977.
Determination of methionine enkephalin in discrete
regions of rat brain. Brain Res. 134: 383-386.

Hsiang, N.S. 1941. Brain pathology of Kaschin-Becks
disease. J. Oriental Med. 22: 119-162.

Huebers, H. and W. Rummel. 1975. Iron binding proteins:
mediators in iron absorption. In: Intestinal
Permeation Excerpta Medica, Amsterdam.

 

 

Humason, G.L. 1979. Animal Tissue Techniques. 4th Ed.
W.H. Freeman and Co. San Francisco.

 

Jacobowitz, D.M., M.A. Silver and W.G. Soden. 1979.
Mapping of leu—enkephalin-containing axons and cell
bodies of the rat forebrain. In: Endorphins 44 Mental
Health Research. Ed. E. Usdin. MacMillen. 62-73.

 

166

Jacobs, A. and M. Worwood. 1974. Iron 22 Biochemistry
and Medicine. Academic Press, London.

 

 

Jones, B.E. and R.Y. Moore. 1977. Ascending projections
of the locus coeruleus in the rat. ll. Autoradio-
graphic study. Brain Res. 127: 23-53.

Kaldor, I. and M. Powell. 1957. Studies on intermediary
iron metabolism. The influence of age and sex on the
storage of supplemental dietary iron in the rat.
Aust. J. of Exp. Biol. 22: 123

Kalra, P.S., S.P. Kalra, L. Krulick, C.P. Fawcett and
S.M. McCann. 1972. Involvement of norepinephrine in
transmission of the stimulatory influence of progester-
one on gonadotropin.release. Endo. 20: 1168-1176.

Kalra, P.S. and S.M. McCann. 1973. Involvement of
catecholamines in feedback mechanisms. Prog. Brain
Res. 22: 185-198.

Keen, C.L. and L.S. Hurley. 1980. Developmental changes in
concentrations of iron, copper and zinc in mouse
tissues. Mech. of Ageing and Devel. 22: 161-176.

Kennedy, B.J. and A.S. Gilbertsen. 1957. Increased
erythropoiesis induced by androgenic-hormone therapy.
New Eng. J. Med. 22: 123.

Kessner, J. and L. Kalk. 1973. Strategy for evaluating
health services. Institute of Medicine, National
Academy of Sciences, Washington. D.C.

King, J.C., S.A. Tobet, F.L. Snavely and A.A. Arimura.
1981. The LHRH system in normal and neonatally

androgenized female rats. Peptides, 2: Suppl. 1.:
85-100.

Kuhn, D.M., B. Ruskin and W. Lovenberg. 1980. Tryptophan
hydroxylase. J. Biol. Chem. 255: 4137-4143.

Kuman, M.S.A., C.L. Chen and T.F. Muther. 1979. Changes
in the pituitary and hypothalamic content of methio-
nine-enkephalin during the estrous cycle of rats.
Life Sci. 22: 1687-1696.

Linder, M.C. and H.N. Munro. 1973. Metabolic and chemical
features of ferritins, a series of iron inducible
tissue proteins. Am. J. of Path. 22: 263-278.

167

Linder, M.C.,J.R. Moor, L.E. Scott and H.N. Munro. 1973.
Mechanism of sex difference in rat tissue iron stores.
Biochem. Biophys. Acta. 297: 70-80.

Lintzel, W., J. Richenberger and E. Schairer. 1944. Uber

den eisenstoff wechsel des neugeborenen und des Sfiug-
lings. Z. Gesamte Exp. Med. 113: 591-612.

Lisk, R.D. and M.A. Barfield. 1975. Sites and mechanisms
of steroid effects on behavior. In: Anatomical Neuro-
endocrinology. Ed. W.E. Stumpf and L.D. Grant. S
Karger, Basel. 232-244.

 

MacDonald, R.A. 1964. Hemochromatosis and Hemosiderosis.
Springfield, Ill. Charles C. Thomas.

Mackler, B., R. Person, L.R. Miller, A.P. Indamar and C.A.
Finch. 1978. Iron deficiency in the rat: Biochemical
studies of brain metabolism. Pediat. Res. 12: 217-
220. '__

MacLusky, N,J. and F. Naftolin, 1981. Sexual different-
iation. of the central nervous system. Science 211:
1294-1303.

Manis, J. G. and D. Schacter. 1962. An active transport
of iron: effects of oral iron and pregnancy. Amer.
J. Physiol. 203: 81-87.

Marks, N. 1978. Biotransformation and degradation of
corticotropins, lipotropins and hypothalamic peptides.
In: Frontiers in Neuroendocrinology. Eds. W.F.
Ganong and L.MT—Martihi. Raven Press. New York.
329-379.

 

 

McCance, R.A. and E.M. Widdowson. 1937. Absorption and
excretion of iron. Lancet 2: 680-684.

McCance, R.A. and E.M. Widdowson. 1951. The metabolism
of iron during suckling, J. Physiol. 112: 450-458.

McCann, S.M. and R.L. Moss. 1975. Putative neurotrans-
mitters involved in dischargeing gonadotropin-
releasing neurohormones and the action of LH-
releasing hormone on the CNS. Life Sci. 22: 833-
852.

McEwen, B.S. 1971. Neural gonadal steroid actions.
Science. 211: 1303-1311.

168

McLardy, T. 1962. Zinc enzymes and the hippocampal
mossy fibre system. Nature 194: 300-302.

Meyer, A. 1958. Hallervorden-Spatz syndrome. In:
Neuropathology. Eds. W. Blackwood. W.H. McMenemey
A. Meyer and R.M. Norman. E. Arnold. London. pp 525.

 

Meyerson. B.J.,M. Eliasson. L. Lindstrdm, A. Michanek
and A.C. deerlund. 1973. Monoamines and female
sexual behavior. In: Psychopharmacology, Sexual
Disorders and Drug Abuse. Ed. T.A. Ban. North-
Holland. Amsterdam. 463-472.

 

 

Moskowitz, A.D., J.C. Liebeskind and L.L. Butcher. 1980.
Distribution of enkephalin-like immunoreactivity in
the mouse brain: strain comparisons. Neurosci.
Abst. 2: 616. '

Muller, M. 1922. Uber physiologisches Vorkommen von
Eisen im Zentralnervensystem. Z. ges. Neurol.
Psychiat. 21: 519-535.

Muraki. T.,H. Nakadate, Y. Tokunage. R. Kato and T. Makin.
1979. Effect of narcotic analgesics and naloxone on
proestrus surges of LH,FSH, and prolactin in rats.
Neuroendo. 22: 241-247.

Murray, M.J. and N. Stein. 1968. The effect of adminis-
tered oestrogens and androgens on the absorption of
iron by rats. Brit. J. Haemat. 24: 407-409.

Musil, A.,H. Bertha, W. Haas and O. Waurschinek. 1962.
Untersuchungen fiber die regionale kationenvertulung
im menchlichen gehirn. Mh. Chemie 22L 536-540.

Nguyen-Legros, J.,J. Bizot, M. Bolesse et J. -P.Pulicani.
1980. "Noir de diaminobenzedine”: une nouvelle met-
hode histochimique de revelation du fer exogene.
Histochem. 22: 239-244.

Norton. W.T. 1976. Formation, structure and biochem—
istry of meylin. InzNeurochemistry. Eds. Siegel,
G.J., R.W. Albers. R. Katzman, and W.B. Agranoff.
Little, Brdwn and Co. Boston. 74-103.

 

Norton, W.T. and S.E. Poduslo. 1973. Myelination in rat
brain. Changes in myelin composition during matura-
tion. J. Neurochem. 22: 759-762.

Noyes, W.D., H.H. Bothwell and C.A. Finch. 1964. The
role of reticulo-endothelial cells in iron metabolism.
Brit. J. Haematol. 2: 43-55.

169

Nutrition Canada. 1973. Nutrition Canada National Survey.
Canadian Dept. of National Health and Welfare. Publ.
No. H 58-36-1973.

Oski. F.A. 1979. The nonhematologic manifestations of
iron deficiency. Am. J. Dis. Child. 133: 315-321.

Otis, L. and M.C. Smith. 1940. Further evidence of sex
variation in the utilization of iron by anemic rats.
Science 22: 146-148.

Palkovitz, M. 1973. Isolated removal of hypothalamic or
other brain nuclei of the rat. Brain Res. 22: 449-450.

Pang. C.N.,E. Zimmerman and C.H. Sawyer. 1977. Morphine
inhibition of the preovulatory surges of plasma LH
and FSH in the rat. Endo. 101: 1726-1732.

Parent. A., L. Descarries and A. Beaudet. 1981. Organ-
ization of ascending serotonin systems in the adult
rat brain. A radioautographic §tudy after intra-
ventricular administration of ( H) 5-hydroxytryptamine.
Neurosci. 2; 115-138.

Pearse, A.G.E. 1961 Histochemistry. 2nd. Ed. LIttle,
Brown and Co. Boston.

 

Phillips, H.S.,B.T. Ho, F. Hostetter and G.P. Kozlowski.
1980. Distribution of luteinizing hormone releasing
hormone (LHRH) in olfactory pathways of the hamster.
Anat. Rec. ‘222: 148A.

Planas. J.’ 1973. The serum ferroxidase activity and the
iron mobilization by estrogens. Rev. esp. Fisiol.
22: 293-300.

Planas. J. 1976. The serum ferroxidase system and the
effect of estrogen on plasma iron. Rev. esp. Fisiol.
22: 115-122.

Politt, E. and R.L. Leibel. 1976. Iron deficiency and
behavior. J. Pediatr. 22: 372-382.

Pollycove. M. 1972. Hemochromatosis. In: The Metabolic
, Basis 22 Inherited Disease. Eds. Stanbury, J.,B.
Wyngaarden and D.S. Freidrickson. McGraw-Hill, N.Y.
1051-1081.

Rafaelsen. O.J. and B. Kofod. 1969. Iron. In: Handbook
22 Neurochggistry. 22’ Ed. A. Lajtha. Plenum Press.
N.Y. 261-271.

 

170

Rajan, K.S., R.W. Colburn and J.M. Davis. 1971. Metal
chelates in the storage and transport of neurotrans
mitters. J. Neurochem.22: 345-364.

Rajan, K.S., R.W. Colburn and J.M. Davis. 1976.
Distribution of metal ions in the subcellular fract
ions of several rat brain areas. Life Sci. 22; 423-
432.

Rubin, J.J. and L.J. Willmore. 1980. Prevention of iron-
induced epileptiform discharges in rats by treatment
of antiperoxidants. Exp. Neurol. 21: 472-480.

Sandler, M. and G.L. Gessa. 1975. Sexual Behavior -
Pharmacology and Biochemistry. Raven Press. New York.

 

 

Sar, M., W.E. Stumpf, R.J. Miller, K. Chang and P.
Cratrecasas. 1978. Immunohistochemical localization

of enkephalin in rat brain and spinal cord. J. Comp.
Neurol. 182: 17-38.

Sato, N. and R.I. Henkin. 1973. Pituitary-gonadal regula—
tion of copper and zinc metabo.ism in the female rat.
Am. J. of Physiol. 225: 508—512.

Sheldon, J.H. 1927. The iron content of the tissues in

haemochromatosis, with special reference to the brain.
Quart. J. Med. 22: 123-137.

Sherman, A.R. 1978. Lipogenesis in iron-deficient adult
rats. Lipids 22: 473-478.

Silverman, A.J., L.C. Krey and E.A. Zimmerman. 1979.
A comparative study of the luteinizing hormone rel-
easing hormome (LHRH) neuronal networks in mammals.
Biol. of Reprod. 22: 98—110.

Smith, M.C. and L. Otis. 1937. Sex variation in the util-
ization of iron by anemic rats.
Science, 22: 125-126.

Snyder, S.H, and S.R. Childers, 1979, Opiate receptors
and opioid peptides. Ann. Rev. Neurosci. 2: 35-64.

Sourkes, T.L. 1972, Psychopharmacology, In: Bgsig
Neurochemistry. Eds. Albers. R,W. G,T. Seigel,
R.R. Katzman and B.W, Agranoff. Little Brown
Boston. 581-606,

 

171

Spatz, H. 1922. Uber des eisennachweis im gehirn,
besonders in zentren des extra—pyramidalmotorischen
systems. Z. ges. Neurol. Psychiat. 21: 261-390.

Spatz, H. 1922. Uber stoffwechseleigentumlichkeiten
in den stammganglien. Z. ges. Neurol. Psychiat. 28:
641-650.

Spatz, H. and A. Metz. 1926. Untersuchungen Uber stoff

transport im nervensystems. Z. ges. Neurol. Psychiat.
100: 428-449.

Steenbock. H., J. Semb and E.C. Van Donk. 1936.. Sexual
differentiation in the storage of iron in the rat.
(Abst) J. Biol. Chem. 114.Ci.

Strassmann, G. 1945. Hemosiderin and tissue iron in the
brain, its relationship, occurrence and importance.
J. Neuropathol. Exptl. Neurol. 4: 393-401.

Subcommittee on Iron. 1979. Iron. University Park Press
Baltimore.

Sullivan, 8., J.D. Raese, H. Akil, D. Blacker and J.D.
Barchas. 1979. Enkephalinase: Preliminary character-
ization and effect of phosphorylation. Neurosci.
Abst. 2: 541.

Sundermann, A. and G. Kempf. 1961. Uber den physiolo~
geschen eisengehalt einiger stammhirnganglien und

siene abhflngigkeit vom lekensalter. Z. Alternsforsch
22: 97-105J

Symes, A.L., K. Missala and T.L. Sourkes. 1971 Iron and
riboflavinsdependant metabolism of a monoamine in the
rat 22 vivo.‘ Science. Wash. 214: 153-155.

Szanto. J. and F. Gallijas. 1966. A study of iron
metaolism in neuropsychiatric patients. Arch. Neurol.
24: 438-442.

Tamir, H. A. Klein.and M.M. Rapport. 1976. Serotonin
‘ binding protein: enhancement of binding by Fe+f and
inhibition of binding by drugs. J. Neurochem. _2‘_§_:
871—878.

Ten-state Nutritional Survey, 1968 k970. Volume IV -

Biochemical, U.S. Department of Health, Education and
Welfare, Publication No(HSM) 72—8132.

172

Thompson, B.E. and L.S. Valberg. 1980. Passage of iron
out of the intestinal mucosa of the rat. Can. J.
Physio. Pharmacol. 22: 129—133.

Tietz, N.W. 1976. Fundamentals 22 Clinical Chemistry
2nd. Ed. W.B. Saunders Co. Philadelphia{

 

 

Tingey, AjH. 1937. The iron, copper and manganese content
of the human brain. J. Ment. Sci. 22: 451-460.

Turner, C.D. 1961. General Endocrinology. 3rd Ed.
W.B. Saunders Co. Philadelphia.

 

Ungerstedt, U. 1971. Stereotaxic mapping of the monamine
pathways in the rat brain. Acta Physiol. Scand. 227.
(Suppl). 1-48.

van der Schoot, P.,D.W. Lincoln and J.S. Clark. 1978.
Activation of hypothalamic neuronal activity by the

electrolytic deposition of iron into the preoptic
area. J. Endor. 22: 107-120.

Webb, T.E. and F.A. Oski. 1973. The effect of iron
deficiency on scholastic achievement, behavioral
stability and perceptual sensitivity of adolescents.
Pediat. Res. 2: 294-298.

Webb, T.E. and F.A. Oski. 1973. Iron deficiency anemia
and scholastic achievement in young adolescents.
J. Pediatr. 22: 827-831.

Weinberg, J.,P.R. Dallman and S. Levine. 1979. Iron
deficiency during early development in the rat.
Behavioral and physiological consequences. Pharmac.
Biochem. Behav. 22? 493-502.

Weinberg, J.,S. Levine and P.R. Dallman. 1979. Long-
term consequences of early iron deficiency in the
rat. Pharmac. Biochem. Behav. 22; 631 638.

Weinfeld, A. 1964. Storage iron in man. Acta Medica
Scandinavia. 177: Suppl. 427: 1-29.

Wescott. G.B., A. Shoden and C.A. Finch. 1953. A
quantitative fractionation of tissue ferritin and
homosiderin. J. Biol. Chem. 204: 815-821.

W.H.O, Scientific Group on Nutritional Anaemias. 19685
World Health Organization Tech. Rep. Sef, 405.

173

Widdowson, E.M. and R.A. McCance, 1948. Sexual
differences in the storage and metabolism of iron.
Biochem. J. 42: 577-581.

Williamson, A.M. and K.T. Ng. 1980. Behavioral Effects
of iron deficiency in the adult rat. Physiol. Behav.

24: 561-567.

Wollemann, M. 1951. A photometrical method for testing
the presence of iron in the central nervous system.
Acta Morph. 2: 127-132.

Worwood, M. 1977. The climical biochemistry of iron.
Semin. Hematol. 24: 3-30.

Wuth, O. 1923. Uber den eisengehalt des gehirns.
Z. ges. Neruol. Psychiat. 24: 474-477. I

Yasinobu, K.T., H.F. Mower and O. Hayaishi. 1975. Iron
and Copper proteins. Plenum Press. New York.

 

Youdim, M.B.H. and A.R. Green. 1977. Biogenic monoamine
metabolism and functional activity in iron-deficient
rats: behavioral correlates. In: Iron Metabolism,
Ciba symposium. Elsevier Amsterdam. 201.225.

 

Youdim, M.B.H., A.R. Green, M.R. Bloomfield, B.O. Mitchell,
D.J. Heal, D.G. Grahame-Smith, 1980. The effects of
iron deficiency on brain biogenic monoamine biochemis-
try and function in rats. Neuropharm. lg: 259—267.

Young, W.C. 1961. The hormones and mating behavior.
In: Sex and 22terna1 Secretions. ‘
Willians and Wilkins. Baltimore. 1173-1239.

 

Zak, B.,E.S. Baginski and E. Epstein. 1980. Modern iron

ligands useful for the measurement of serum iron.
Ann. Clin. Lab. Sci. 29: 276-289.

Zaleski. S. 1887. Das eisen der organe bei morbus maculo-
sus werlhofii. -
Arch. exptl. Pathol. u. Pharmakol. 22, 77- 96.

Zimmerman, E.A., K.C. Hsu, M. Ferin and G.P. Kozlowski.
1974. Localization of gonadotropin-releasing hormone
(Gn-RH) in the hypothalamus of the mouse by immuno-
peroxidase technique. Endo. 22: 1-8.