Effect of Manipulation of Iron Storage,
Transport, or Availability on Myelin
Composition and Brain Iron Content in
Three Different Animal Models
E. Ortiz,1J.M. Pasquini,1K.Thompson,2B. Felt,3G. Butkus,4J. Beard,5and
1Biological Chemistry Department, School of Pharmacy and Biochemistry, University of Buenos Aires,
Buenos Aires, Argentina
2Harvard School of Public Health, Boston Massachusetts
3Department of Pediatrics, University of Michigan, Ann Arbor, Michigan
4Department of Neural and Behavioral Sciences, Pennsylvania State College of Medicine, Hershey, Pennsylvania
5Department of Nutrition, Pennsylvania State University, State College, Pennsylvania
Several observations suggest that iron is an essential
factor in myelination and oligodendrocyte biology. How-
ever, the specific role of iron in these processes remains
to be elucidated. This role could be as an essential
cofactor in metabolic processes or as a transcriptional or
translational regulator. In this study, we used animals
models each with a unique defect in iron availability,
storage, or transfer to test the hypothesis that disrup-
tions in these mechanisms affect myelinogenesis and
myelin composition. Disruption of iron availability either
by limiting dietary iron or by altering iron storage capacity
resulted in a decrease in myelin proteins and lipids but
not the iron content of myelin. Among the integral myelin
proteins, proteolipid protein was most consistently af-
fected, suggesting that limiting iron to oligodendrocytes
results not only in hypomyelination but also in a decrease
in myelin compaction. Mice deficient in transferrin must
receive transferrin injections beginning at birth to remain
viable, and these mice had increases in all of the myelin
components and in the iron content of the myelin. This
finding indicates that the loss of endogenous iron mobil-
ity in oligodendrocytes could be overcome by application
of exogenous transferrin. Overall, the results of this study
demonstrate how myelin composition can be affected by
loss of iron homeostasis and reveal specific chronic
changes in myelin composition that may affect behavior
and attempts to rescue myelin deficits.
© 2004 Wiley-Liss, Inc.
Key words: iron storage; iron transport; myelin compo-
sition; brain iron content
Iron deficiency is a common nutritional disorder,
especially in children, affecting over 25% people world-
wide (Cook et al., 1994; Beard and Connor, 2003). Iron
deficiency during pregnancy in humans is frequently as-
sociated with prematurity and perinatal mortality (Beard,
2003). Children who experienced iron deficiency suffer
behavioral defects, such as impaired learning and memory
function (Lozoff et al., 1987; Baynes, 1996). Iron defi-
ciency in humans is reportedly associated with a decrease
in auditory and visual evoked potentials attributed to
hypomyelination (Algarin et al., 2003). This attribution in
consistent with magnetic resonance imaging in human
infants indicating that altered iron status is associated with
hypomyelination (Deregnier et al., 2000). A relationship
between iron and brain function has been more directly
assessed in rodent models, in which iron deficiency is
reportedly associated with persistent changes in resting
energy status, neurotransmission, and myelination (Rao et
al., 2003). A direct effect of iron deficiency on myelin has
been shown, which includes a decrease in lipids (Larkin
and Rao, 1990; Oloyede et al., 1992; Kwik-Uribe et al.,
2000) and some proteins (Beard et al., 2003). The litera-
ture on iron deficiency and neural function has recently
been reviewed (Beard and Connor, 2003).
The principal cell type in the brain that stains posi-
tively following iron histochemistry is white matter oli-
godendrocytes (LeVine and Macklin, 1990; Connor and
Menzies, 1996). During development, iron and ferritin are
initially found in microglia in the brain (Connor et al.,
1995; Cheepsunthorn et al., 1998). Subsequently, as my-
elination is initiated, iron, ferritin and transferrin (Tf)
mRNA are found within oligodendrocytes (Bartlett et al.,
1991; Connor et al., 1995; Cheepsunthorn et al., 1998).
*Correspondence to: James R. Connor, PhD, Professor, Department of
Neural and Behavioral Sciences (H109), Penn State Universtiy College of
Medicine, 500 University Dr., Hershey, PA 17033. E-mail: email@example.com
Received 5 February 2004; Revised 7 May 2004; Accepted 12 May 2004
Published online 7 July 2004 in Wiley InterScience (www.
interscience.wiley.com). DOI: 10.1002/jnr.20207
Journal of Neuroscience Research 77:681–689 (2004)
© 2004 Wiley-Liss, Inc.
Furthemore, peak iron uptake into the brain coincides
with the onset of myelination (Crowe and Morgan, 1992).
Functionally, biosynthesis of cholesterol and lipids
that are abundant and key components of myelin requires
iron as a cofactor in key synthetic steps (Pleasure et al.,
1984). Indeed, the demyelination associated with tellu-
rium toxicity is thought to result from blockage of an
iron-requiring step in cholesterol biosynthesis (Wagner-
Recio et al., 1991). Specific iron-requiring enzymes in-
volved in maintaining a high rate of metabolic and
biosynthetic activity, such as glucose-6-phosphate dehy-
drogenase, dioxygenase, succinic dehydrogenase, and
NADH dehydrogenase, as well as the cytochrome oxidase
system, are all elevated in oligodendrocytes relative to
other cells in the brain (Cammer, 1984). Thus, iron could
be involved in myelination, both directly through syn-
thetic pathways and indirectly as an essential cofactor for
the activity of metabolic enzymes in oligodendrocytes.
We hypothesize that the temporal and spatial events
involving iron accumulation by oligodendrocytes during
development are critical to myelinogenesis and myelin
composition and, if disrupted, may affect myelin and
myelin-related functions throughout life. To test this idea,
we disrupted the iron delivery protein Tf and the iron
storage protein ferritin by using genetically modified mu-
tant mouse lines and disrupted the availability of iron by
using a developmental iron-deficient rat model.
MATERIALS AND METHODS
The studies reported here were performed in compliance
with the animal procedures approved by our institutional animal
use committee (protocols 2002-129 and 7623).
Hypotransferrinemic (hpx) mice were used to test the
specific hypothesis that the ability to mobilize iron in oligoden-
drocytes is required for the production of normal myelin. These
animals lack the ability to make Tf because of a splicing defect
in the mRNA (Bernstein, 1987) and as a result have less than 1%
of the normal circulating levels of Tf in plasma. The animals in
this study were 4–5 months of age and came from our colony,
which has been maintained for the past 10 years. The animals
require weekly Tf injections from birth to survive and are
maintained with a weekly intraperitoneal injection of 0.3 ml
human apotransferrin (6 mg/ml) per animal (Buys et al., 1991).
Although the Tf injections permit iron delivery to cells and
specifically the oligodendrocytes (Dickinson and Connor,
1995), they do not replace the endogenously produced Tf in
oligodendrocytes, and there is no known mechanism for remov-
ing Tf from endosomes when it is internalized into the cells.
Thus hpx mice provide a unique opportunity to determine the
role of endogenously produced Tf by oligodendrocytes in my-
elin production. Nonaffected littermates were used as controls.
The hpx/hpx mice are easily identified at birth by their pale
appearance. In total, ten control and five hpx/hpx animals were
Heterozygotes for the H-Ferritin (H-Frt) Null Mutation
The generation of this mouse line has been reported
previously (Thompson et al., 2003). Southern blot analysis of
the litters reveals that only heterozygotes for the mutation sur-
vive to birth (Thompson et al., 2003). To identify the heterozy-
gotic animals, DNA is collected from the tail and digested with
NcoI, and the DNA is subjected to Southern blotting as previ-
ously reported (Thompson et al., 2003). These mice were used
to test the hypothesis that disruptions in intracellular iron storage
capacity will decrease the production of myelin. The animals
used in this study were 6 months of age, and the controls were
unaffected littermates. There were a minimun of three animals
Developmental Iron-Deficient Rat Model
Pregnant Sprague Dawley dams were fed an iron-deficient
diet (4–10 mg/kg iron) or an iron-sufficient diet (40 mg/kg
iron)1beginning at gestational day 5. Diets were prepared by
Harlan Teklad Nutritionals (Madison, WI). Mothers and litters
were maintained on their diets through gestation and lactation.
After postnatal day 20, all animals were fed the iron-sufficient
diet. Pups were weaned at postnatal day 23 and maintained on
the same diet until they were killed at 6 months of age for
myelin analysis. There were six animals per group for the myelin
analyses and three animals per group for the iron analyses.
The animals were anesthetized with sodium pentobarbital
and then killed by decapitation. The brain was excised and the
tissue homogenized at 10% (w/v) in ice-cold 0.32 M sucrose.
An aliquot of the total homogenate was saved, and the remain-
der was fractionated to obtain purified myelin (Norton and
Poduslo, 1973). The purified myelin fraction was used imme-
diately or stored at –20°C for further studies.
Lipid Extraction and Chemical Determination
An aliquot of the purified myelin fraction and an aliquot
of the total homogenate were used to determine total protein.
Bovine serum albumin was used as a standard. The myelin
membranes were extracted with 2:1 chloroform:methanol (v/v)
according to Folch et al. (1957), and the washed total lipid
extract was used for the analysis of lipids. Chloroform:methanol-
soluble proteins (proteolipid proteins) were determined accord-
ing to Lees and Paxman (1972). Lipid phosphorous was deter-
mined according to Chen et al. (1956), galactolipids according
to Hess and Lewin (1965), and cholesterol by using the method
of Searcy and Bergquist (1960). The factors used to calculate
total weights of galactolipids and phospholipids were 4.6 and
24.2, respectively (Lapetina et al., 1968).
Analysis of Fatty Acids
For the quantification of fatty acids, an aliquot of the
extract was dried under a stream of N2and sterified with
methanol:benzene (4:1 v/v) at 100°C for 1 hr. Fatty acid analysis
was performed by gas chromatography on a Shimadzu GC-8a
1The pregnant dams are started on a diet containing 4 mg/kg iron and then
are switched to 10 mg/kg on P7, which is required to maintain the pups.
682Ortiz et al.
gas chromatograph with a DB-23 capillary column (J&W Sci-
entific, Folson, CA) with temperature programming at 5°C/min
between 140°C and 220°C. Fatty acids were identified by their
retention time and by cochromatography with commercial stan-
Polyacrylamide Gel Electrophoresis and
An aliquot of purified myelin (100 ?g protein) was pre-
cipitated with acetone at –20°C and centrifuged, and the pellet
was dried under N2and dissolved in 50 mM Tris-HCl buffer,
pH 6.8, containing 2% sodium dodecyl sulfate (SDS), 0.1%
2-mercaptoetahnol, and 20% glycerol. The proteins were sepa-
rated by electrophoresis with 12.5% acrylamide gels and a dis-
continuous buffer system. The separated proteins were either
stained with Coomasie brilliant blue R250 or electrotransferred
to nitrocellulose membranes for Western blot analysis of myelin
basic protein. Membranes containing the electrotransferred pro-
teins were individually incubated with the following primary
antibodies: rabbit antimyelin basic protein (1/3,000; generously
supplied by Dr. Tony Campagnoni, UCLA) and guinea pig
antiexon II myelin basic protein (1/4,000; generously supplied
by Dr. David Colman) The secondary antibodies where either
anti-rabbit and anti-guinea pig conjugated to peroxidase at a
dilution 1/20,000. Quantitative analysis was carried out using a
Gel-Pro Analyzer (Media Cybernetics Inc.). Tf and ferritin
concentrations were determined in an immunoblot assay with
antibodies to Tf (Chemicon, Temecula, CA) and H-ferritin
(Cheepsunthorn et al., 1998), respectively.
Analysis of Myelin Proteins
Protein concentration was determined, and aliquots of the
different samples (50 ?g protein) were subjected to SDS-PAGE.
Proteins were quantitated by using a Gel Pro System and are
referred to in percentage of the total amount of proteins or as the
amount of protein in each band, taking as 100% the amount of
total myelin proteins present in 1 g fresh tissue.
Myelin was extracted as described above and then sub-
jected to lyophilization. Iron content was determined by atomic
absorption spectrophotometry in a standard matrix.
Statistical analysis between multiple groups was carried
out by ANOVA (one-way repeated) and post hoc analysis with
Student’s t-test for multiple means. All analyses were performed
with Graph Pad Prism 3.0 software.
In the hpx mice, the myelin indices were, in general,
elevated compared with those in control animals. The
total protein concentration in isolated myelin was in-
creased by 12% (P ? 0.05), and the chloroform-methanol
extracted proteins were increased by 35% (P ? 0.01).
Total phospholipids and galactolipids were increased by
30% (P ? 0.05) and 40%% (P ? 0.05), respectively. These
results are summarized in Table I. The protein profiles
obtained by SDS-PAGE of the myelin fraction isolated
from hpx mice and control mice revealed that there was
no increase in the percentage of any of the integral myelin
proteins in the total amount of myelin protein, but, based
on weight, all of the protein except for MAG increased
(Table II). The fatty acid composition of the myelin lipids
was unchanged in the hpx mice compared with the con-
trol group (Table III).
H-Ferritin Heterozygotes for the Null Mutation
Mice that have a decrease in expression of H-ferritin
have less total protein in myelin and chloroform:
methanol-soluble protein than wild-type control mice.
Total phospholipids and galactolipids were also signifi-
cantly decreased. These results are summarized in Table I.
TABLE I. Chemical Composition and Iron Content of Myelin in Hypotransferrinemic Mice (Hpx), H-Ferritin Mutant Mice
and Iron-Deficient (I.D.) Rats†
Hpx Ferritin mutantsI.D. rat
[Fe] brain (whole
brain minus myelin)
2.39 ? 0.052.71 ? 0.10**2.38 ? 0.131.93 ? 0.05*9.78 ? 0.238.25 ? 0.49*
0.81 ? 0.05
2.46 ? 0.12
2.63 ? 0.11
1.24 ? 0.03***
4.01 ? 0.24***
3.68 ? 0.21***
1.17 ? 0.04
2.89 ? 0.14
2.19 ? 0.09
0.91 ? 0.02*
1.97 ? 0.14***
1.89 ? 0.10*
1.43 ? 0.01
3.90 ? 0.17
6.18 ? 0.11
4.06 ? 0.11
1.30 ? 0.06*
3.21 ? 0.21*
5.16 ? 0.27*
2.85 ? 0.20***
126 ? 44.6
26.5 ? 6.0
229 ? 75.3***
164.0 ? 19.3***
112 ? 29.8
19.1 ? 7.6
88.3 ? 21.8
18.6 ? 2.3
82.9 ? 5.3
18.9 ? 1.5
88.8 ? 30.5
20.1 ? 1.4
†Total and chloroform:methanol-soluble proteins and lipids were determined as described in Materials and Methods. Six rats were analyzed per group for
the myelin analyses and three animals per group for the iron analyses. There were three mice per group for the H-ferritin mouse line studies and five per
group for the hpx mouse studies. Results for myelin components are expressed as mg/g fresh tissue and are the mean ? SEM of multiple analyses performed
in triplicate. Iron data are reported as ?g Fe/g tissue. nd, Not determined.
*P ? 0.05.
**P ? 0.005.
***P ? 0.0005.
Iron Manipulation and Myelin Composition683
When the protein profile of the myelin was analyzed by
SDS-PAGE, only proteolipid protein (PLP) was decreased
in the H-Frt mutant mice both as a percentage of the total
protein and in relation to myelin weight (Table IV). The
fatty acid composition of the myelin fraction isolated from
the mutant ferritin mice did not differ from that of normal
mice (Table V).
Subjecting rats to dietary iron deficiency during de-
velopment was associated with less total myelin protein as
well as decreased cholesterol, phospholipids, and galacto-
lipids (Table I). The 30% decrease in cholesterol was the
largest change in the latter measurements. The protein
profile of myelin revealed that PLP and MBP21 declined
as a percentage of the total myelin protein, but all of the
integral myelin proteins except for CNPase and MBP14
decreased when measured against total tissue weight (Ta-
ble VI). The fatty acid composition was similar between
the two groups, with the exception of the 20:0 form
Because iron itself was manipulated in this group, we
determined the relative amounts of Tf and ferritin in the
myelin fraction of the iron-deficient rats. Tf is decreased
by over 50% in the developmentally iron-deficient rats
compared with normal rats (Fig. 1), but the ferritin levels
are similar between the two groups (Fig. 2).
Neither the iron-deficient diet nor the genetic ma-
nipulation of the H-ferritin gene resulted in a statistically
significant difference in the concentration of iron in the
myelin. However, the hpx mice had four times more iron
in the myelin fraction than normal (Table I). The iron
content of the nonmyelin fraction of the brain homoge-
nate was also analyzed. Similarly to the case for the myelin
fraction, there was no difference in the iron concentration
between developmentally iron-deficient and control rats
or H-ferritin-deficient mice and their controls. However,
the hpx mice had over twice as much iron as unaffected
littermates (Table I).
The results of this study demonstrate that disruptions
in iron processing, storage, or availability affect both the
quantity and the quality of myelin. The results also indi-
cate that the myelin composition and quantity are altered
even if the iron content of the myelin acheives normal
levels. The hypermyelination in hypotransferrinemic mice
reveals a rare mechanism for increasing myelin production
and iron content of oligodendrocytes and raises questions
about the role of endogenously produced Tf in oligoden-
A goal of this analysis was to determine whether
there were specific deficits within myelin composition
associated with specific disruptions in iron homeostatsis.
Although the decrease in myelin content associated with
developmental dietary iron deficiency was fairly uniform,
only PLP and MBP21 decreased as a percentage of the
total protein. PLP is required for compaction and forma-
tion of myelin sheaths and may be required for oligoden-
drocyte maturation (Greer and Lees, 2002). MBP21 is part
of a family of intracellular adhesion proteins (Staugaitis et
al., 1996) whose expression pattern is normally unchanged
TABLE II. Distribution of Myelin Proteins in Hypotransferrinemic Mice*
[(%/g total protein)]
[(%/g total protein)]
4.62 ? 0.41
7.81 ? 0.72
23.96 ? 1.94
6.02 ? 0.36
8.31 ? 0.39
5.18 ? 0.43
18.13 ? 1.85
4.12 ? 0.20 (ns)
7.97 ? 0.65 (ns)
22.20 ? 2.09 (ns)
5.67 ? 0.12 (ns)
8.24 ? 0.47 (ns)
5.33 ? 0.42 (ns)
20.70 ? 2.07 (ns)
0.10 ? 0.01
0.18 ? 0.01
0.49 ? 0.04
0.13 ? 0.01
0.20 ? 0.01
0.12 ? 0.01
0.41 ? 0.05
0.11 ? 0.01 (ns)
0.25 ? 0.003(P ?0.01)
0.68 ? 0.06 (P ?0.02)
0.16 ? 0.003(P ?0.04)
0.23 ? 0.01 (P ?0.04)
0.16 ? 0.004(P ?0.02)
0.60 ? 0.05 (P ?0.02)
*Myelin membranes isolated from control (n ? 5) and hypotransferrinemic (n ? 5) mice were separated by electrophoresis in 12.5% acrylamide gels using
a discontinuous buffer system as described. Quantitative analysis was carried out using a Gel-Pro Analyzer. Values are expressed as % protein/g protein
and mg protein/g fresh tissue. ns, Not statistically significant.
TABLE III. Myelin Fatty Acid Composition in
Hpx/? (%) Hpx/hpx (%)
13.5 ? 0.5
21.2 ? 0.3
23.4 ? 0.8
3.64 ? 0.087
0.613 ? 0.006
1.49 ? 0.01
5.69 ? 0.08
7.85 ? 0.25
1.98 ? 0.08
4.52 ? 0.11
3.80 ? 0.19
3.47 ? 0.08
9.33 ? 0.43
14.0 ? 2.2 (ns)
21.4 ? 2.7 (ns)
23.6 ? 3.4 (ns)
3.79 ? 0.18 (ns)
0.631 ? 0.039 (ns)
1.44 ? 0.26 (ns)
6.03 ? 0.51 (ns)
8.07 ? 0.99 (ns)
1.95 ? 0.12 (ns)
4.88 ? 0.34 (ns)
3.87 ? 0.18 (ns)
3.47 ? 0.42 (ns)
9.98 ? 1.22 (ns)
*A total lipid extract was performed on myelin membranes isolated from
hpx/? and hpx/hpx mice. Fatty acid analysis was carried out by gas
chromatography on an aliquot of the lipid extract as described in Materials
and Methods. Results are expressed as the percentage of total fatty acids ?
SEM. ns, No statistically significant difference.
684Ortiz et al.
in myelin with age (Barbarese et al., 1978). MBP tran-
scription is reportedly affected directly by Tf/iron
(Espinosa-Jeffrey et al., 2002). The decrease in both of
these proteins suggests that the compaction of myelin is
insufficient in iron deficiency. Furthermore, our data in-
dicate that attempts to correct developmental iron defi-
ciency by a return to normal dietary iron levels at weaning
is not sufficient to reestablish normal myelin composition,
even if normal amounts of iron in the myelin fraction can
Fatty acid composition was also examined in this
study because of its importance during brain development
and for normal maturation. Iron is a structural component
of both delta 6 and delta 9 desaturase enzymes (Strittmatter
and Enoch, 1978; Okayasu, 1981), and delta 9 has been
shown to be reduced as a result of dietary iron deficiency
(Rao, 1979). Severe dietary iron restriction has produced
significant changes in brain fatty acid composition
(Oloyede et al., 1992; Stangl and Kirchgessner, 1998). A
feeding paradigm in mice that was similar to ours resulted
in a decrease in fatty acid composition (Kwik-Uribe et al.,
2000). In our study, only the 20:0 fatty acid family was
affected. In a more severe model of dietary iron restriction
in rats, the sum of the n-3 fatty acids was decreased, but
not the n-6. The only specific fatty acid family to decrease
was 20:4 (Beard et al., 2003). Because some of the previ-
ous studies were performed with whole-brain homoge-
nates (Larkin and Rao, 1990; Oloyede et al., 1992; Beard
et al., 2003), the decrease in fatty acid composition is not
myelin specific, as shown in our study, but nonetheless
presumably contributes to the overall myelin deficit that is
associated with iron deficiency. Because limiting Tf avail-
ability or iron storage capacity does not effect the fatty acid
composition, the effects of iron availability on fatty acid
composition appear to be related to energy production by
oligodendrocytes rather than a direct effect of iron.
It cannot be ruled out that the decrease in myelin
with iron deficiency or compromised iron storage capacity
could reflect a decrease in the number of oligodendro-
cytes. Recent evidence suggests that iron deficiency can
affect the proliferation of oligodendrocyte precursor cells
and generation of oligodendrocytes (Mayer-Proschel and
Morath, 2002). An initial effect of iron deficiency would
be presumably to lower H-ferritin in the oligodendro-
cytes. H-ferritin mRNA is one of the earliest genes ex-
pressed in oligodendrocytes committed to myelination
(Sanyal et al., 1996). In addition, ferritin protects oligo-
dendrocytes from stress (Qi and Dawson, 1994; Qi et al.,
1995), and H-ferritin has an antiapoptotic effect (Cozzi et
al., 2003). Evidence that increased apoptosis of oligoden-
drocytes will affect myelin was found recently (Atkinson et
al., 2003). Therefore, less H-ferritin production by the
oligodendrocytes could have negatively affected their sur-
vival. Consistent with the possibility that oligodendrocyte
number is decreased by compromising ferritin production
or iron availability is the decrease in Tf in the myelin
fraction measured in the current study. Normally, Tf
production in the brain increases with iron deficiency
(Beard and Connor, 2003) and, in the brain, Tf and Tf
mRNA are dependent on a normal population of oligo-
dendrocytes (Connor et al., 1987; Bartlett et al., 1991). Tf
mRNA is also reportedly decreased in the brains of iron-
deficient rats (Han et al., 2003), which could be consistent
with fewer oligodendrocytes. We directly examined the
effect of loss of H-ferritin on myelin production by intro-
TABLE IV. Distribution of Myelin Proteins in Ferritin Mutant Mice*
[(%/g total protein)]
[(%/g total protein)]
3.28 ? 0.46
11.82 ? 1.92
35.86 ? 2.37
8.00 ? 0.91
3.91 ? 1.33
11.22 ? 2.14
8.39 ? 2.42
3.40 ? 0.66 (ns)
8.67 ? 2.16 (ns)
27.50 ? 2.17 (P ? 0.04)
5.54 ? 1.08 (ns)
8.00 ? 1.74 (ns)
16.60 ? 2.81 (ns)
8.67 ? 1.89 (ns)
0.09 ? 0.01
0.34 ? 0.06
0.91 ? 0.09
0.23 ? 0.03
0.11 ? 0.04
0.32 ? 0.06
0.24 ? 0.07
0.08 ? 0.02 (ns)
0.20 ? 0.05 (ns)
0.49 ? 0.10 (P ? 0.02)
0.16 ? 0.02 (ns)
0.19 ? 0.04 (ns)
0.39 ? 0.07 (ns)
0.20 ? 0.04 (ns)
*These data were collected as described for Table II. Values are expressed as % protein/g protein and mg protein/g fresh tissue. There were three mice
in each group. ns, Not statistically significant.
TABLE V. Myelin Fatty Acid Composition in Ferritin
11.8 ? 0.239
20.3 ? 0.390
25.3 ? 0.521
3.51 ? 0.0579
0.383 ? 0.0167
0.908 ? 0.0458
6.87 ? 0.238
7.93 ? 0.136
1.43 ? 0.0744
4.80 ? 0.301
3.92 ? 0.271
2.45 ? 0.246
10.60 ? 0.601
12.1 ? 0.176 (ns)
20.2 ? 0.353 (ns)
25.4 ? 0.307 (ns)
3.42 ? 0.0780 (ns)
0.376 ? 0.0120 (ns)
0.962 ? 0.0641 (ns)
6.80 ? 0.307 (ns)
7.59 ? 0.0802 (ns)
1.57 ? 0.0825 (ns)
4.81 ? 0.108 (ns)
3.65 ? 0.332 (ns)
2.70 ? 0.217 (ns)
11.1 ? 0.515 (ns)
*Myelin membranes isolated from wild-type mice (n ? 3) and mice
heterozygotic for the null ferritin mutation (n ? 3) were used to obtain a
total lipid extract. Fatty acid analysis was carried out by gas chromatography
in an aliquot of the lipid extract as described in Materials and Methods.
Results are the mean expressed as % total fatty acids ? SEM. ns, No
statistically significant difference.
Iron Manipulation and Myelin Composition 685
ducing a mouse line heterozygotic for the H-ferritin null
mutation. The results from this mouse line were very
similar to those obtained from the iron-deficient mice. In
a previous study, we did not observe any increase in
oxidative stress markers in cells in the white matter, sug-
gesting that oligodendrocytes were not under duress in this
model (Thompson et al., 2003). However, we cannot rule
out that there are fewer oligodendrocytes in this model,
and studies to address this possibility are currently under-
way. Clearly, evaluation of the mechanism by which iron
deficiency or compromised iron storage can affect myeli-
nation must include the effect on oligodendrocyte cell
number and the effect of iron deficiency on Tf production
The relationship between oligodendrocytic matura-
tion and brain iron accumulation has been examined by
using genetically mutant rats and mice. If oligodendrocytes
are present but do not make myelin, brain iron uptake is
unaffected (Gocht et al., 1993), but the iron accumulates
in astrocytes and microglia (Connor and Menzies, 1990).
If oligodendrocytes reach maturity and produce myelin,
even if the myelin is abnormal, iron accumulates in oli-
godendrocytes (LeVine, 1991; Connor et al., 1993). The
inability to make Tf does not preclude oligodendrocytes
from taking up iron (Dickinson and Connor, 1995;
Takeda et al., 1998) or, as shown in the current study,
from synthesizing myelin. Taken together, these studies
provide clear evidence that myelin-producing oligoden-
drocytes are not required for brain iron uptake, nor is iron
accumulation by oligodendrocytes affected by the quality
of myelin produced; however, the amount of iron accu-
mulated by oligodendrocytes does affect the quality and
quantity of myelin produced by these cells.
Although limiting iron availability is associated with
hypomyelination, the results from the hypotransferrinemic
mice indicate that hypermyelination can be induced by Tf
injections, even in the absence of endogenous production
of Tf by oligodendrocytes. The increases in myelin com-
ponents in the hypotransferrinemic mice are consistent
with reports that Tf is important for myelination and
oligodendrocyte survival but raise a number of questions
about the role of endogenously synthesized Tf, especially
given the finding that Tf synthesized in oligodendrocytes
is not secreted (de Arriba Zerpa et al., 2000). Although the
hpx mice receive Tf injections systemically, there is less Tf
in the hpx mouse brain than normal (Dickinson and
TABLE VI. Distribution of Myelin Proteins in Iron-Deficient Rats*
[(%/g total protein)]
[(%/g total protein)]
2.48 ? 0.09
2.45 ? 0.48
39.07 ? 1.41
2.33 ? 0.13
6.70 ? 0.54
25.22 ? 2.06
7.80 ? 1.08
2.34 ? 0.09
2.37 ? 0.19
0.24 ? 0.01
0.24 ? 0.05
3.82 ? 0.14
0.23 ? 0.01
0.65 ? 0.05
2.47 ? 0.20
0.76 ? 0.11
0.19 ? 0.01 (P ? 0.005)
0.20 ? 0.02
2.84 ? 0.07 (P ? 0.0002)
0.12 ? 0.01 (P ? 0.0001)
0.43 ? 0.04 (P ? 0.01)
1.97 ? 0.07 (P ? 0.02)
0.73 ? 0.11
34.36 ? 0.81 (P ? 0.02)
1.44 ? 0.10 (P ? 0.01)
5.17 ? 0.44
23.88 ? 0.89
8.78 ? 1.34
*The data were generated and reported as described for Table II.
TABLE VII. Myelin Fatty Acid Composition in Iron-Deficient
Control (%)I.D. rats (%)
C 18:2 cis
C 18:2 trans
13.0 ? 0.316
21.2 ? 0.181
27.2 ? 0.311
3.95 ? 0.156
0.557 ? 0.0150
1.20 ? 0.0279
3.87 ? 0.112
6.42 ? 0.0800
1.43 ? 0.0744
3.15 ? 0.0445
6.21 ? 0.172
13.8 ? 0.214 (ns)
21.3 ? 0.323 (ns)
27.2 ? 0.740 (ns)
4.33 ? 0.294 (ns)
0.558 ? 0.00924 (ns)
1.08 ? 0.05 (P ? 0.046)
3.49 ? 0.144 (ns)
6.33 ? 0.173 (ns)
1.26 ? 0.127 (ns)
3.11 ? 0.156 (ns)
6.07 ? 0.317 (ns)
*Myelin membranes isolated from control (n ? 6) and iron-deficient (n ?
6) rats were used to obtain a total lipid extract. Fatty acid analysis was carried
out by gas chromatography in an aliquot of the lipid extract as described in
Materials and Methods. Results are the mean expressed as % total fatty
acids ? SEM. ns, No statistically significant difference.
Fig. 1. Transferrin in the myelin fraction of iron-deficient rats (I.D.).
The relative amount of transferrin was determined in the myelin
fraction from control and iron-deficient rats by immunoblot analysis.
The amount of transferrin is decreased by 50% in the iron-deficient
model. *P ? 0.05.
686Ortiz et al.
Connor, 1995). The injected Tf gains access to the brain
and oligodendrocytes (Dickinson and Connor, 1995), but
the exogenous Tf taken into these cells should remain in
the endosomes and not contribute to the intracellular
mobility of the iron within cytoplasm. This statement
assumes that endocytic recycling of the Tf-Tf receptor
complex is similar in oligodendrocytes to that in all other
cells studied (Aisen et al., 2001), but, based on the current
results, this warrants specific investigation. Only the oli-
godendrocytes in the adult hpx mice stain for iron and Tf
(Dickinson and Connor, 1995).
There is a generalized increase in expression of all
myelin constituents in the hpx mice compared with the
control group. The current biochemical analysis of myelin
content does not agree with a previous morphological
analysis that reported a decrease in the thickness of the
corpus callosum in hpx mice as revealed with a luxol fast
blue stain (Dickinson and Connor, 1994). The biochem-
ical analysis used in the current study is a global measure-
ment of myelin content, and the morphological analysis
reported specific regional differences (e.g., specifically at
the level of the corpus callosum, not the entire subcortical
tract). The results for the hpx mice are remarkably similar
to the effects following intracranial injection of Tf on early
postnatal myelin production in normal rats (Escobar Ca-
brera et al., 1994; Marta et al., 2002). The results are also
consistent with the findings from a transgenic line that
expressed the secreted form of Tf in oligodendrocytes
(Saleh et al., 2003) and are consistent with cell culture
studies that have identified the importance of Tf for oli-
godendrocytes survival, differentiation, and growth (Espi-
nosa de los Monteros et al., 1999). How Tf injections
affect myelin related functions remains to be studied, but
improved motor function was found in the transgenic
mice. These studies taken together suggest a powerful
trophic effect of Tf on myelin production in vivo and
suggest that Tf injections could provide a therapeutic
opportunity for hypomyelination and perhaps in some
demyelinating diseases. However, it should not be con-
cluded from these data that the increase in myelin in the
hpx mice translates to increased functional performance. A
single injection of apotransferrin showed ultrastructural
evidence of myelin decompactation in the optic nerve and
corpus callosum (Marta et al., 2003).
The effect of Tf supplementation on myelin amounts
and composition implies that the hypermyelination in hpx
mice is Tf mediated. How oligodendrocytes receive iron
has been under investigation for a number of years. During
development, Tf receptors are detected immunohisto-
chemically on oligodendrocytes but are not present after
myelination begins (Lin and Connor, 1989). Furthermore,
no autoradiographic studies, or immunohistochemical
studies, or in situ hybridization studies have identified Tf
receptors or Tf mRNA in white matter in adults rats,
mice, or humans (Hulet et al., 1999a,b, 2000, 2002; Han
et al., 2003). Tf receptor expression in white matter can-
not be detected in iron-deficienct rats (Han et al., 2003) or
in hypotransferrinemia (Dickinson and Connor, 1998).
We have reported that a protein that binds ferritin is
present specifically in white matter tracts in rodent and
human brains (Hulet et al., 1999a,b). The binding of
ferritin follows the temporal onset and spatial pattern of
myelinogenesis in the mouse brain (Hulet et al., 2002).
Thus, it is possible that the effects of exogenous Tf injec-
tions on myelin in the whole-animal model are induced
early in development, because Tf injections begin shortly
alter birth. In support of this idea is the finding that
intracranial injections at postnatal day 20 do not induce an
increase in myelin as was seen following similar injections
at postnatal day 3 (Escobar Cabrera et al., 1994, 2000), but
the increase in myelin composition associated with the
postnatal day 3 injection persists until at least 60 days of
age. Consequently, systemic Tf injections may be a viable
option for therapeutic intervention for iron deficiency,
but the efficacy may be time sensitive.
This work was supported by a program project grant
from the National Institutes of Health (NICHD grant
39386 to Betsy Lozoff, MD) to the group in Brain and
Behavior in Early Iron Deficiency and by a Fogarty In-
ternational Research Collaboration Award (1 R03
TW6288-01A1 to J.R.C.) and by the University of Bue-
nos Aires (grant B 092 to J.M.P.).
Aisen P, Enns C, Wessling-Resnick M. 2001. Chemistry and biology of
eukaryotic iron metabolism. Int J Biochem Cell Biol 33:940–959.
Algarin C, Peirano P, Garrido M, Pizarro F, Lozoff B. 2003. Iron deficiency
anemia in infancy: long-lasting effects on auditory and visual system
functioning. Pediatr Res 53:217–223.
Atkinson S, Li YQ, Wong CS. 2003. Changes in oligodendrocytes and
myelin gene expression after radiation in the rodent spinal cord. Int J
Radiat Oncol Biol Phys 57:1093–1100.
Fig. 2. Ferritin in the myelin fraction of iron-deficient rats (I.D.). The
relative amount of ferritin was determined in the myelin fraction from
control and iron-deficient rats by immunoblot analysis. The amounts of
ferritin in the various myelin fractions were not statistically significantly
Iron Manipulation and Myelin Composition687
Barbarese E, Carson JH, Braun PE. 1978. Accumulation of the four myelin
basic proteins in mouse brain during development. J Neurochem 31:779–
Bartlett WP, Li XS, Connor JR. 1991. Expression of transferrin mRNA in
the CNS of normal and jimpy mice. J Neurochem 57:318–322.
Baynes RD. 1996. Assessment of iron status. Clin Biochem 29:209–215.
Beard J. 2003. Iron deficiency alters brain development and functioning. J
Beard JL, Connor JR. 2003. Iron status and neural functioning. Annu Rev
Beard J, Wiesinger J, Connor J. 2003. Pre- and postweaning iron deficiency
alters myelination in Sprague-Dawley rats. Dev Neurosci 25:308–315.
Bernstein SE. 1987. Hereditary hypotransferrinemia with hemosiderosis, a
murine disorder resembling human atransferrinemia. J Lab Clin Med
Buys SS, Martin CB, Eldridge M, Kushner JP, Kaplan J. 1991. Iron
absorption in hypotransferrinemic mice. Blood 78:3288–3290.
Cammer W. 1984. Oligodendrocyte associated enzymes. In: Norton WT,
editor. Oligodendroglia. New York: Plenum Press. p 199–232.
Cheepsunthorn P, Palmer C, Connor JR. 1998. Cellular distribution of
ferritin subunits in postnatal rat brain. J Comp Neurol 400:73–86.
Chen PS Jr, Toribara T. 1956. Microdetermination of phosphorus. Anal
Connor JR, Menzies SL. 1990. Altered cellular distribution of iron in the
CNS of myelin deficient rats. Neuroscience 34:265–271.
Connor JR, Menzies SL. 1996. Relationship of iron to oligodendrocytes
and myelination. Glia 17:83–93.
Connor JR, Phillips TM, Lakshman MR, Barron KD, Fine RE, Csiza CK.
1987. Regional variation in the levels of transferrin in the CNS of normal
and myelin-deficient rats. J Neurochem 49:1523–1529.
Connor JR, Menzies SL, Roskams AJ, Williams ME. 1993. Transferrin and
iron in the shiverer mouse brain. J Neurosci Res 36:501–507.
Connor JR, Pavlick G, Karli D, Menzies SL, Palmer C. 1995. A histo-
chemical study of iron-positive cells in the developing rat brain. J Comp
Cook JD, Skikne BS, Baynes RD. 1994. Iron deficiency: the global
perspective. Adv Exp Med Biol 356:219–228.
Cozzi A, Levi S, Corsi B, Santambrogio P, Campanella A, Gerardi G,
Arosio P. 2003. Role of iron and ferritin in TNFalpha-induced apoptosis
in HeLa cells. FEBS Lett 537:187–192.
Crowe A, Morgan EH. 1992. Iron and transferrin uptake by brain and
cerebrospinal fluid in the rat. Brain Res 592:8–16.
de Arriba Zerpa GA, Saleh MC, Fernandez PM, Guillou F, Espinosa de los
Monteros A, de Vellis J, Zakin MM, Baron B. 2000. Alternative splicing
prevents transferrin secretion during differentiation of a human oligoden-
drocyte cell line. J Neurosci Res 61:388–395.
Deregnier R, Nelson C, Thomas K, Wewerka S, Georgieff M. 2000.
Neurophysiologic evaluation of auditory recognition memory in healthy
newborn infants and infants of diabetic mothers. J Pediatr 137:777–784.
Dickinson TK, Connor JR. 1994. Histological analysis of selected brain
regions of hypotransferrinemic mice. Brain Res 635:169–178.
Dickinson TK, Connor JR. 1995. Cellular distribution of iron, transferrin,
and ferritin in the hypotransferrinemic (Hp) mouse brain. J Comp Neurol
Dickinson TK, Connor JR. 1998. Immunohistochemical analysis of trans-
ferrin receptor: regional and cellular distribution in the hypotransferrine-
mic (hpx) mouse brain. Brain Res 801:171–181.
Escobar Cabrera OE, Bongarzone ER, Soto EF, Pasquini JM. 1994. Single
intracerebral injection of apotransferrin in young rats induces increased
myelination. Dev Neurosci 16:248–254.
Escobar Cabrera OE, Soto EF, Pasquini JM. 2000. Myelin membranes
isolated from rats intracranially injected with apotransferrin are more
susceptible to in vitro peroxidation. Neurochem Res 25:87–93.
Espinosa de los Monteros A, Kumar S, Zhao P, Huang CJ, Nazarian R, Pan
T, Scully S, Chang R, de Vellis J. 1999. Transferrin is an essential factor
for myelination. Neurochem Res 24:235–248.
Espinosa-Jeffrey A, Kumar S, Zhao PM, Awosika O, Agbo C, Huang A,
Chang R, De Vellis J. 2002. Transferrin regulates transcription of the
MBP gene and its action synergizes with IGF-1 to enhance myelinogen-
esis in the md rat. Dev Neurosci 24:227–241.
Folch PJ, Lees M, Sloane-Stanley G. 1957. A simple method for isolation
and purification of total lipids from animal tissues. J Biol Chem 226:499–
Gocht A, Keith AB, Candy JM, Morris CM. 1993. Iron uptake in the brain
of the myelin deficient rat. Neurosci Lett 154:187–190.
Greer JM, Lees MB. 2002. Myelin proteolipid protein—the first 50 years.
Int J Biochem Cell Biol 34:211–215.
Han J, Day JR, Connor JR, Beard JL. 2003. Gene expression of transferrin
and transferrin receptor in brains of control vs. iron-deficient rats. Nutr
Hess HH, Lewin E. 1965. Microassay of biochemical structural components
in nervous tissue II. Method for cerebrosides, proteolipid proteins and
residue proteins. J Neurochem 12:205–211.
Hulet SW, Powers S, Connor JR. 1999a. Distribution of transferrin and
ferritin binding in normal and multiple sclerotic human brains. J Neurol
Hulet SW, Hess EJ, Debinski W, Arosio P, Bruce K, Powers S, Connor JR.
1999b. Characterization and distribution of ferritin binding sites in the
adult mouse brain. J Neurochem 72:868–874.
Hulet SW, Heyliger SO, Powers S, Connor JR. 2000. Oligodendrocyte
progenitor cells internalize ferritin via clathrin-dependent receptor medi-
ated endocytosis. J Neurosci Res 61:52–60.
Hulet SW, Menzies S, Connor JR. 2002. Ferritin binding in the developing
mouse brain follows a pattern similar to myelination and is unaffected by
the jimpy mutation. Dev Neurosci 24:208–213.
Kwik-Uribe C, Gietzen D, German J, Golub M, Keen C. 2000. Chronic
marginal iron intake during early development in mice results in persistent
changes in dopamine metabolism and myelin composition. J Nutr 130:
Lapetina E, Soto E, De Robertis E. 1968. Lipids and proteolipids in isolated
subcellular membranes of rat brain cortex. J Neurochem 15:437–445.
Larkin E, Rao G. 1990. Importance of fetal and neonatal iron: Adequacy for
normal development of central nervous system. In: Dobbing J, editor.
Brain, behaviour and iron in the infant diet. London: Springer-Verlag. p
Lees MB, Paxman S. 1972. Modification of the Lowry procedure for the
analysis of proteolipid proteins. Anal Biochem 47:184–192.
LeVine SM. 1991. Oligodendrocytes and myelin sheaths in normal, quak-
ing and shiverer brains are enriched in iron. J Neurosci Res 29:413–419.
LeVine SM, Macklin WB. 1990. Iron-enriched oligodendrocytes: a reex-
amination of their spatial distribution. J Neurosci Res 26:508–512.
Lin HH, Connor JR. 1989. The development of the transferrin-transferrin
receptor system in relation to astrocytes, MBP and galactocerebroside in
normal and myelin-deficient rat optic nerves. Brain Res Dev Brain Res
Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez
E, Jimenez R, Mora LA, Gomez I, Krauskoph D. 1987. Iron deficiency
anemia and iron therapy effects on infant developmental test performance.
Marta CB, Ortiz EH, Hallak M, Baron B, Guillou F, Zakin MM, Soto EF,
Pasquini JM. 2002. Changes in the expression of cytoskeletal proteins in
the CNS of transferrin transgenic mice. Dev Neurosci 24:242–251.
Marta C, Paez P, Lopez M, Pellegrino de Iraldi A, Soto EF, Pasquini JM.
2003. Morphological changes of myelin sheats in rats intracranially in-
jected with apotransferrin. Neurochem Res 28:101–110.
Mayer-Proschel M, Morath DJ. 2002. Iron deficiency during embryogen-
esis and consequences for oligodendrocyte generation in vivo. Dev Neu-
688Ortiz et al.
Norton WT, Poduslo S. 1973. Myelination in rat brain: method of myelin Download full-text
isolation. J Neurochem 21:749–757.
Okayasu T. 1981. Purification and characterization of linoleoyl-CoA de-
saturase from rat liver microsomes [author’s translation]. Hokkaido Igaku
Oloyede OB, Folayan AT, Odutuga AA. 1992. Effects of low-iron status
and deficiency of essential fatty acids on some biochemical constituents of
rat brain. Biochem Int 27:913–922.
Pleasure D, Kim SU, Silberberg DH. 1984. In vitro studies of oligoden-
droglial lipid metabolism. In: Norton W, editor. Oliogodenroglia. New
York: Plenum Press. p 175–198.
Qi Y, Dawson G. 1994. Hypoxia specifically and reversibly induces the
synthesis of ferritin in oligodendrocytes and human oligodendrogliomas.
J Neurochem 63:1485–1490.
Qi Y, Jamindar TM, Dawson G. 1995. Hypoxia alters iron homeostasis and
induces ferritin synthesis in oligodendrocytes. J Neurochem 64:2458–
Rao PS. 1979. Fatty acid composition of cerebrosides and phospholipids in
brains of undernourished rats. Nutr Metab 23:136–144.
Rao R, Tkac I, Townsend E, Gruetter R, Georgieff M. 2003. Perinatal
iron deficiency alters the neurochemical profile of the developing rat
hippocampus. J Nutr 133:3215–3221.
Saleh MC, Espinosa de los Monteros A, de Arriba Zerpa GA, Fontaine I,
Piaud O, Djordjijevic D, Baroukh N, Garcia Otin AL, Ortiz E, Lewis S,
Fiette L, Santambrogio P, Belzung C, Connor JR, de Vellis J, Pasquini
JM, Zakin MM, Baron B, Guillou F. 2003. Myelination and motor
coordination are increased in transferrin transgenic mice. J Neurosci Res
Sanyal B, Polak PE, Szuchet S. 1996. Differential expression of the heavy-
chain ferritin gene in non-adhered and adhered oligodendrocytes. J Neu-
rosci Res 46:187–197.
Searcy R, Bergquist LM. 1960. A new color reaction for the quantitation
of serum cholesterol. Clin Chim Acta 5:192–199.
Stangl GI, Kirchgessner M. 1998. Different degrees of moderate iron
deficiency modulate lipid metabolism of rats. Lipids 33:889–895.
Staugaitis SM, Colman DR, Pedraza L. 1996. Membrane adhesion and
other functions for the myelin basic proteins. Bioessays 18:13–18.
Strittmatter P, Enoch HG. 1978. Purification of stearyl-CoA desaturase
from liver. Methods Enzymol 52:188–193.
Takeda A, Devenyi A, Connor JR. 1998. Evidence for non-transferrin-
mediated uptake and release of iron and manganese in glial cell cultures
from hypotransferrinemic mice. J Neurosci Res 51:454–462.
Thompson K, Menzies S, Muckenthaler M, Torti FM, Wood T, Torti SV,
Hentze MW, Beard J, Connor J. 2003. Mouse brains deficient in
H-ferritin have normal iron concentration but a protein profile of iron
deficiency and increased evidence of oxidative stress. J Neurosci Res
Wagner-Recio M, Toews AD, Morell P. 1991. Tellurium blocks choler-
sterol synthesis by inhibiting squlaene metabolism: preferential vulnera-
bility to this metabolic block leads to peripheral nervous system demy-
elination. J Neurochemistry 57:1891.
Iron Manipulation and Myelin Composition 689